www.fgks.org   »   [go: up one dir, main page]
Next Issue
Volume 16, June-1
Previous Issue
Volume 16, May-1
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
8.3
4.2
Journals
Remote Sensing
Volume 16
Issue 10
remotesensing-logo

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Remote Sens., Volume 16, Issue 10 (May-2 2024) – 171 articles

Cover Story (view full-size image): Satellite-acquired short-wave infrared (SWIR) imagery has long been used to identity different types of rocks and minerals, but technological constraints have largely prevented the acquisition of high-resolution imagery required for most archaeological investigations. This experimental study employs a new drone-deployed, hyperspectral SWIR sensor in an effort to locate and characterize archaeological artifacts. Producing imagery at a 4cm spatial resolution across 326 spectral bands in SWIR wavelengths, we employ a supervised classification algorithm to locate and identify different artifact types, including ceramics, lithics, and metals. The results showcase the potential of this emerging technology to transform approaches to the discovery, mapping, and interpretation of the surface archaeological record and broader cultural landscapes. View this paper
  • Issues are regarded as officially published after their release is announced to the table of contents alert mailing list.
  • You may sign up for e-mail alerts to receive table of contents of newly released issues.
  • PDF is the official format for papers published in both, html and pdf forms. To view the papers in pdf format, click on the "PDF Full-text" link, and use the free Adobe Reader to open them.
Order results
Result details
Section
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
27 pages, 9304 KiB  
Article
KNN Local Linear Regression for Demarcating River Cross-Sections with Point Cloud Data from UAV Photogrammetry URiver-X
by Taesam Lee, Seonghyeon Hwang and Vijay P. Singh
Remote Sens. 2024, 16(10), 1820; https://doi.org/10.3390/rs16101820 - 20 May 2024
Viewed by 483
Abstract
Aerial surveying with unmanned aerial vehicles (UAVs) has been popularly employed in river management and flood monitoring. One of the major processes in UAV aerial surveying for river applications is to demarcate the cross-section of a river. From the photo images of aerial [...] Read more.
Aerial surveying with unmanned aerial vehicles (UAVs) has been popularly employed in river management and flood monitoring. One of the major processes in UAV aerial surveying for river applications is to demarcate the cross-section of a river. From the photo images of aerial surveying, a point cloud dataset can be abstracted with the structure from the motion technique. To accurately demarcate the cross-section from the cloud points, an appropriate delineation technique is required to reproduce the characteristics of natural and manmade channels, including abrupt changes, bumps and lined shapes. Therefore, a nonparametric estimation technique, called the K-nearest neighbor local linear regression (KLR) model, was tested in the current study to demarcate the cross-section of a river with a point cloud dataset from aerial surveying. The proposed technique was tested with synthetically simulated trapezoidal, U-shape and V-shape channels. In addition, the proposed KLR model was compared with the traditional polynomial regression model and another nonparametric technique, locally weighted scatterplot smoothing (LOWESS). The experimental study was performed with the river experiment center in Andong, South Korea. Furthermore, the KLR model was applied to two real case studies in the Migok-cheon stream on Hapcheon-gun and Pori-cheon stream on Yecheon-gun and compared to the other models. With the extensive applications to the feasible river channels, the results indicated that the proposed KLR model can be a suitable alternative for demarcating the cross-section of a river with point cloud data from UAV aerial surveying by reproducing the critical characteristics of natural and manmade channels, including abrupt changes and small bumps as well as different shapes. Finally, the limitation of the UAV-driven demarcation approach was also discussed due to the penetrability of RGB sensors to water. Full article
(This article belongs to the Special Issue Intelligent Processing and Application of UAV Remote Sensing Image Data)
►▼ Show Figures

Figure 1

Figure 1
<p>Example of the distance measurement: (<b>a</b>) aerial photo with a selected cross-section (two red dots, L and N and thick red line); (<b>b</b>) magnified photo of the panel (<b>a</b>) with assisted 3D axis (x, y and z) and the selected point (M); (<b>c</b>) emphasized triangle with the points of NML. Note that (1) the cross-section can be defined with the x-axis by connecting points N and L with the line; (2) point M is the example point that contains the red line at the panel (<b>a</b>), which is a group of points in reality; and (3) the actual distance of M from N in the x-axis is represented as <span class="html-italic">k</span>, which can be designated as N to the point that meets line NL perpendicularly from M. The aerial images were taken from the authors.</p> Full article ">Figure 2
<p>Assumed synthetic trapezoidal channel (not a real one) to test the KLR model (thick black dotted line) for the simulated point cloud data with Equation (S6) to a total of 322 points (red dots) with different portions of the number of neighbors (<math display="inline"><semantics> <mrow> <mi>k</mi> <mo>=</mo> <mi>a</mi> <msqrt> <mi>n</mi> </msqrt> </mrow> </semantics></math>, here <span class="html-italic">a</span> = 1, 2, 3 and 4 at each row panels). Note that (1) the trapezoidal sections are consistent with a 4 m top both sides and a 6 m base width as well as a 1:1 side slope with a 6 m height; (2) the number of points for the channel was divided at each 0.1 m to a total of 161 points (blue line); (3) two times the divided data are simulated with Equation (S6) to a total of 322 points (red dots) shown in the left panels (<b>a</b>) while five times larger point data than the points in the left panels (<b>a</b>) were simulated and shown in the right panels (<b>b</b>); and (4) the elevation of the bottom channel was assumed to be 18 m.</p> Full article ">Figure 3
<p>Root mean square error (RMSE) between the KLR estimate with different multipliers (<span class="html-italic">a</span>) of the number of neighbors (<math display="inline"><semantics> <mrow> <mi>k</mi> <mo>=</mo> <mi>a</mi> <msqrt> <mi>n</mi> </msqrt> </mrow> </semantics></math>) and the original trapezoid points for the case of 2 times the original points (<b>a</b>) and 10 times (<b>b</b>); with the original trapezoid points for all of the cases between 1 and 12 times the original points (<b>c</b>); as well as the optimum multiplier (<b>d</b>) with the RMSE value at the top panel for each multiple simulation. Note that increasing the number of multiple simulations indicates that the number of overlapped photos increases and the cloud points are multiplied.</p> Full article ">Figure 4
<p>Polynomial regression (top panels, (<b>a-1</b>,<b>b-1</b>)) with the black dashed line with circles and the black dotted line with triangle markers for PolyFit2 and PolyFit4, respectively (see Equations (2) and (4)) and LOWESS (bottom panels, (<b>a-2</b>,<b>b-2</b>)) were fitted to the stochastically simulated point cloud data (red circles) of two times the divided points (322 points) in the left panels ((<b>a-1</b>,<b>a-2</b>)) and five times larger than those points shown in the right panels ((<b>b-1</b>,<b>b-2</b>)) for the synthetic trapezoidal channel points (blue line).</p> Full article ">Figure 5
<p>Different KNN-based methods to estimate the synthetic trapezoidal channel as (<b>a</b>) KNN, (<b>b</b>) KNN1 and (<b>c</b>) KLR models. Note that (1) the KNN model was reproduced from the original paper of Lall and Sharma [<a href="#B35-remotesensing-16-01820" class="html-bibr">35</a>]; and (2) the KNN1 model (i.e., k = 1) indicates that the closet point was used to demarcate the channel.</p> Full article ">Figure 6
<p>Synthetic U-shape river cross-section (blue solid line with cross markers) and the simulated point could data (red circles) of 10 times the synthetic channel with Equation (S6) as well as the fitted estimates to KLR (<b>a</b>), LOWESS (<b>b</b>) and PolyFit (<b>c</b>). Note that (1) the U-shape river cross-section was designed with the power function as in Equations (12) and (13) and the U-shape was synthetically built following the reference of Neal, Odoni, Trigg, Freer, Garcia-Pintado, Mason, Wood and Bates [<a href="#B41-remotesensing-16-01820" class="html-bibr">41</a>] and the section was divided into 262 points; (2) the V-shape river cross-section was designed with the height of 4 m and top width of 16 m and the section was divided into 121 points.</p> Full article ">Figure 7
<p>Location of the River Experiment Center (REC) at the top panels and the selected channels of the ground surveying for the straight river with I-shape (I<sub>1</sub>, … I<sub>6</sub>), the meandering river with S-shape (S<sub>1</sub>, … S<sub>7</sub>) and the steep river with U-shape (U<sub>1</sub>, …U<sub>7</sub>) rivers. The aerial image is taken and produced by the authors and no copyright is required.</p> Full article ">Figure 8
<p>Point cloud data (red circles) for the channel I<sub>1</sub> (1) and I<sub>5</sub> (2) of the REC site and model-fitted line (black dashed line) with KLR (<b>a</b>), LOWESS (<b>b</b>) and PolyFit (<b>c</b>) as well as the ground surveying.</p> Full article ">Figure 9
<p>Point cloud data (red circles) for the channel U<sub>1</sub> (1) and U<sub>2</sub> (2) of the REC and model-fitted line (black dashed line) with KLR (<b>a</b>), LOWESS (<b>b</b>) and PolyFit (<b>c</b>) as well as the ground surveying.</p> Full article ">Figure 10
<p>Point cloud data (red circles) for the channels S<sub>3</sub> (1) and S<sub>6</sub> (2) of the REC and model-fitted line (black dashed line) with KLR (<b>a</b>), LOWESS (<b>b</b>) and PolyFit (<b>c</b>) as well as the ground surveying.</p> Full article ">Figure 11
<p>Study area of the applied stream in the panel (<b>a</b>), Migok-cheon in South Korea, located in the province of Hapcheon-gun and locations of four tested sites in the panel (<b>b</b>) in the Migok-cheon stream. Note that the other four panels surrounding the left-top panel magnify each tested site by showing the point clouds of the observed data taken from the UAV photogrammetry. The aerial images were taken from the authors.</p> Full article ">Figure 12
<p>Point cloud data (red circles) and model-fitted line (black dashed line) with KLR (<b>a</b>), LOWESS (<b>b</b>) and PolyFit (<b>c</b>) as well as the observed surveying for Site-1 (left panels) and Site-2 (right panels). Note that (1) the observed line was drawn from the previous surveying in BRTMA [<a href="#B47-remotesensing-16-01820" class="html-bibr">47</a>]; and (2) the detailed information including the map is attached in the <a href="#app1-remotesensing-16-01820" class="html-app">Supplementary Materials (Figures S1 and S2</a> as well as <a href="#remotesensing-16-01820-t001" class="html-table">Table 1</a>).</p> Full article ">Figure 13
<p>Point cloud data (red circles) and model-fitted line (black dashed line) with KLR (<b>a</b>), LOWESS (<b>b</b>) and PolyFit (<b>c</b>) as well as the observed surveying for Site-3 (left panels) and Site-4 (right panels). Note that (1) the observed line was drawn from the previous surveying in BRTMA [<a href="#B47-remotesensing-16-01820" class="html-bibr">47</a>]; and (2) the detailed information including the map is attached in <a href="#app1-remotesensing-16-01820" class="html-app">Supplementary Materials (Figures S1 and S2</a> as well as <a href="#remotesensing-16-01820-t001" class="html-table">Table 1</a>).</p> Full article ">
19 pages, 10961 KiB  
Article
Revealing the Hidden Consequences of Increased Soil Moisture Storage in Greening Drylands
by Yu Wang, Tian Han, Yuze Yang, Yue Hai, Zhi Wen, Ruonan Li and Hua Zheng
Remote Sens. 2024, 16(10), 1819; https://doi.org/10.3390/rs16101819 - 20 May 2024
Viewed by 421
Abstract
Vegetation primarily draws water from soil moisture (SM), with restoration in drylands often reducing SM storage (SMS). However, anomalies have been detected in the Beijing–Tianjin Sand Source Region (BTSSR) of China via the Global Land Data Assimilation System (GLDAS) and Gravity Recovery and [...] Read more.
Vegetation primarily draws water from soil moisture (SM), with restoration in drylands often reducing SM storage (SMS). However, anomalies have been detected in the Beijing–Tianjin Sand Source Region (BTSSR) of China via the Global Land Data Assimilation System (GLDAS) and Gravity Recovery and Climate Experiment (GRACE). This study quantified the sources of increased SMS in drylands to elucidate the effects of vegetation restoration on SMS. The results indicated the following: (1) In vegetated drylands, 46.2% experienced a significant increase in SMS while 53.8% remained stable; both were positively correlated with the normalised difference vegetation index (NDVI). (2) The increase in SMS was accompanied by a decrease in groundwater storage (GWS), as indicated by the significant correlation coefficients of −0.710 and −0.569 for SMS and GWS, respectively. Furthermore, GWS served as the primary source of water for vegetation. (3) The results of the redundancy analysis (RDA) indicated that the initial vegetation, the driver of the observed trend of increased SMS and decreased GWS, accounted for 50.3% of the variability in water storage. Therefore, to sustain dryland ecosystems, we recommend that future vegetation restoration projects give due consideration to the water balance while concurrently strengthening the dynamic monitoring of SMS and GWS. Full article
(This article belongs to the Section Ecological Remote Sensing)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>(<b>a</b>) Location of the BTSSR and spatial pattern of various land uses/covers. (<b>b</b>) Investment and the area restored in the BTSSR since 2001.</p> Full article ">Figure 2
<p>Framework of the study.</p> Full article ">Figure 3
<p>GWSA vs. the measured depth of the groundwater table at (<b>a</b>) Inner Mongolia Station (NMG) and (<b>b</b>) Erdos Station (ESD). ‘*’ indicates significance at the <span class="html-italic">p</span>-value &lt; 0.05 level.</p> Full article ">Figure 4
<p>(<b>a</b>) Combination of the NDVI and SMS trends in the BTSSR from 2001 to 2019. In the legend, ‘+’ represents a significant increase, ‘o’ denotes no significant change, and ‘−’ indicates a significant decrease; the values in the colour blocks refer to the proportion of vegetation areas occupied at the 95% confidence level, with × for nonexistent. The NDVI and SMS trends in the (<b>b</b>) ++ area (Type 1) and (<b>c</b>) +o area (Type 2) and their correlation coefficients. The green dotted line represents the trend-fit line for NDVI, and the blue dotted line represents the trend-fit line for SMS. ‘*’ indicates significance at the <span class="html-italic">p</span>-value &lt; 0.05 level.</p> Full article ">Figure 5
<p>Water sources for SMS in Type 1 areas and Type 2 areas. Dark green represents Type 1 areas, light green represents Type 2 areas, and the sum of the shares of the two water sources in each type area is 100%.</p> Full article ">Figure 6
<p>Temporal trends of hydrological components in (<b>a</b>) Type 1 areas and (<b>b</b>) Type 2 areas. ‘*’ indicates significance at the <span class="html-italic">p</span>-value &lt; 0.05 level, while ‘ns’ indicates not significant. The blue, orange, red and black dotted line represents the trend-fit line for PRE, AET, GWSA and QS, respectively.</p> Full article ">Figure 7
<p>(<b>a</b>) Spatial distribution and (<b>b</b>) trend area proportion of groundwater storage anomaly temporal change trends. ‘*’ indicates significance at the <span class="html-italic">p</span>-value &lt; 0.05 level.</p> Full article ">Figure 8
<p>(<b>a</b>) RDA plot showing the relationship between the water storage trend (dark blue arrows) and environmental variable trend (red arrows) in drylands with restored vegetation. Different colours and shapes are used to distinguish between types: dark green spheres indicate Type 1 areas, and light green diamonds indicate Type 2 areas. RDA 1 and RDA 2 represent the first and second axes, respectively. The values in parentheses indicate the proportion (%) of significantly bounded variation accounted for by each RDA component. (<b>b</b>) Relationship between the initial NDVI and the SMS and GWSA trends. ‘*’ indicates significance at the <span class="html-italic">p</span>-value &lt; 0.05 level.</p> Full article ">Figure 9
<p>Changes in the GWSA-DSI in greened drylands, with the coloured areas representing the occurrence of groundwater drought.</p> Full article ">
19 pages, 9502 KiB  
Article
Statistical Analysis of Multi-Year South China Sea Eddies and Exploration of Eddy Classification
by Yang Jin, Meibing Jin, Dongxiao Wang and Changming Dong
Remote Sens. 2024, 16(10), 1818; https://doi.org/10.3390/rs16101818 - 20 May 2024
Viewed by 399
Abstract
Mesoscale eddies are structures of seawater motion with horizontal scales of tens to hundreds of kilometers, impact depths of tens to hundreds of meters, and time scales of days to months. This study presents a statistical analysis of mesoscale eddies in the South [...] Read more.
Mesoscale eddies are structures of seawater motion with horizontal scales of tens to hundreds of kilometers, impact depths of tens to hundreds of meters, and time scales of days to months. This study presents a statistical analysis of mesoscale eddies in the South China Sea (SCS) from 1993 to 2021 based on eddies extracted from satellite remote sensing data using the vector geometry eddy detection method. On average, about 230 eddies with a wide spatial and temporal distribution are observed each year, and the numbers of CEs (52.2%) and AEs (47.8%) are almost similar, with a significant correlation in spatial distribution. In this article, eddies with a lifetime of at least 28 days (17% of the number of total eddies) are referred to as strong eddies (SEs). The SEs in the SCS that persist for several years in similar months and locations, such as the well-known dipole eddies consisting of CEs and AEs offshore eastern Vietnam, are defined as persistent strong eddies (PSEs). SEs and PSEs affect the thermohaline structure, current field, and material and energy transport in the upper ocean. This paper is important as it names the SEs and PSEs, and the naming of eddies can facilitate research on specific major eddies and improve public understanding of mesoscale eddies as important oceanic phenomena. Full article
(This article belongs to the Special Issue Recent Advances on Oceanic Mesoscale Eddies II)
►▼ Show Figures

Figure 1

Figure 1
<p>The bathymetry of the SCS derived from ETOPO1 (shaded: ocean depth; unit: m).</p> Full article ">Figure 2
<p>The SCS surface current in (<b>a</b>) the summer (June–August) and (<b>b</b>) winter (December, January–February) averaged during the 1993–2021 period (unit: m/s). The SCS surface wind speed in (<b>c</b>) the summer and (<b>d</b>) winter averaged during the 1993–2021 period (unit: m/s).</p> Full article ">Figure 3
<p>(<b>a</b>) The interannual variation in eddy numbers. (<b>b</b>) The seasonal variation in eddy numbers on average in a year during the 1993–2021 period. The blue (red) dashed lines indicate the trend of change in the number of CEs (AEs) on the interannual scale.</p> Full article ">Figure 4
<p>The eddy number per year (lifetime ≥ 7 days) averaged during the 1993–2021 period for (<b>a</b>) CEs and (<b>b</b>) AEs.</p> Full article ">Figure 5
<p>The distribution of eddy trajectories with lifetimes of at least 7 days during the 1993–2021 period (a sub image is drawn every two years). The trajectories of CEs (AEs) are shown in solid blue (red) lines, “*” is the position of eddy generation, and “o” is the position of eddy extinction.</p> Full article ">Figure 6
<p>The number of eddies with different (<b>a</b>) lifetimes (days ≥ 7), (<b>b</b>) radius values (km), (<b>c</b>) EKE values (m<sup>2</sup>/s<sup>2</sup>), (<b>d</b>) SLA deviations (m), and (<b>e</b>) maximum current speeds (m/s) during the 1993–2021 period.</p> Full article ">Figure 7
<p>Eddy (<b>a</b>) radius (km), (<b>b</b>) EKE (m<sup>2</sup>/s<sup>2</sup>), (<b>c</b>) SLA deviation (m), and (<b>d</b>) maximum current speed of eddy (m/s) based on different eddy lifetimes.</p> Full article ">Figure 8
<p>SE number per year (lifetime ≥ 28 days) averaged during 1993–2021 period for (<b>a</b>) CE and (<b>b</b>) AE.</p> Full article ">Figure 9
<p>The distribution of SE trajectories during the 1993–2021 period. The trajectories of CEs (AEs) are shown in solid blue (red) lines, “*” is the position of eddy generation, and “o” is the position of eddy extinction.</p> Full article ">Figure 10
<p>Dipole eddy distribution averaged from 1993 to 2021 (No dipole structure is detected in 1998, 2006, 2010, or 2015). Colored shading and black vectors are SLA and geostrophic current anomalies, respectively.</p> Full article ">Figure 11
<p>The numbers of eddies with different (<b>a</b>) vorticity (s<sup>−1</sup>) and (<b>b</b>) SST anomalies (°C) during the 1993–2021 period.</p> Full article ">Figure 12
<p>(<b>a</b>) Spatial distribution of pressure (Pa; shaded) and wind speed (m/s; black vectors) during Typhoon “Lekima” (red circle) on 6 August 2019. (<b>b</b>) Distribution of SCS eddies on 6 August 2019. CEs (AEs) are shown in black (blue) solid lines, background colors represent SLA (m), and black vectors indicate surface current field (m/s).</p> Full article ">Figure 13
<p>The distribution of surface eddies in the SCS from 2017 to 2021. The black box in (<b>a</b>–<b>c</b>) shows the “Chelsea (C20210110-10°N, 113°E)-P” distribution from 2019 to 2021. The black box in (<b>d</b>–<b>h</b>) shows the “Erin (C20210628-13°N, 113°E)-P” and “Eunice (A20210721-12°N, 111°E)-P” distributions from 2017 to 2021. The colored shading and black vectors represent the SLA and geostrophic current anomalies, respectively. The title of each subgraph is the date of the eddy distribution map. The CEs (AEs) are shown in blue (red) solid lines.</p> Full article ">
28 pages, 9420 KiB  
Article
Coastline Automatic Extraction from Medium-Resolution Satellite Images Using Principal Component Analysis (PCA)-Based Approach
by Claudio Parente, Emanuele Alcaras and Francesco Giuseppe Figliomeni
Remote Sens. 2024, 16(10), 1817; https://doi.org/10.3390/rs16101817 - 20 May 2024
Viewed by 555
Abstract
In recent decades several methods have been developed to extract coastlines from remotely sensed images. In fact, this is one of the principal fields of remote sensing research that continues to receive attention, as testified by the thousands of scientific articles present in [...] Read more.
In recent decades several methods have been developed to extract coastlines from remotely sensed images. In fact, this is one of the principal fields of remote sensing research that continues to receive attention, as testified by the thousands of scientific articles present in the main databases, such as SCOPUS, WoS, etc. The main issue is to automatize the whole process or at least a great part of it, so as to minimize the human error connected to photointerpretation and identification of training sites to support the classification of objects (basically soil and water) present in the observed scene. This article proposes a new fully automatic methodological approach for coastline extraction: it is based on the unsupervised classification of the most decorrelated fictitious band derived from Principal Component Analysis (PCA) applied to the satellite images. The experiments are carried out on datasets characterized by images with different geometric resolution, i.e., Landsat 9 Operational Land Imager (OLI) multispectral images (pixel size: 30 m), a Sentinel-2 dataset including blue, green, red and Near Infrared (NIR) bands (pixel size: 10 m) and a Sentinel-2 dataset including red edge, narrow NIR and Short-Wave Infrared (SWIR) bands (pixel size: 20 m). The results are very encouraging, given that the comparison between each extracted coastline and the corresponding real one generates, in all cases, residues that present a Root Mean Squared Error (RMSE) lower than the pixel size of the considered dataset. In addition, the PCA results are better than those achieved with Normalized Difference Water Index (NDWI) and Modified NDWI (MNDWI) applications. Full article
(This article belongs to the Special Issue Coastal and Littoral Observation Using Remote Sensing)
►▼ Show Figures

Figure 1

Figure 1
<p>Geolocalization of the study areas: the rectangles delimit the three study areas located respectively in Campania (red), Sardinia (yellow) and Sicily (green); the map is in equirectangular projection and WGS 84 ellipsoidal coordinates (EPSG: 4326).</p> Full article ">Figure 2
<p>The 3 study areas in false color compositions of the Landsat 9 OLI images in UTM/WGS 84 plane coordinates (EPSG: 32633): Campania on the <b>left</b>, Sardinia on the <b>right</b> and Sicily on the <b>bottom</b>.</p> Full article ">Figure 3
<p>The 3 study areas, in false color compositions of the Sentinel-2A images in UTM/WGS 84 plane coordinates (EPSG: 32633): Campania on the <b>left</b>, Sardinia on the <b>right</b> and Sicily on the <b>bottom</b>.</p> Full article ">Figure 4
<p>Workflow of the adopted approach.</p> Full article ">Figure 5
<p>Landsat 9 OLI synthetic bands for Campania study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 6
<p>Landsat 9 OLI synthetic bands for Sardinia study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 7
<p>Landsat 9 OLI synthetic bands for Sicily study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 8
<p>Sentinel-2 (20 m) synthetic bands for Campania study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 9
<p>Sentinel-2 (20 m) synthetic bands for Sardinia study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 10
<p>Sentinel-2 (20 m) synthetic bands for Sicily study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 11
<p>Sentinel-2 (10 m) synthetic bands for Campania study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 12
<p>Sentinel-2 (10 m) synthetic bands for Sardinia study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 13
<p>Sentinel-2 (10 m) synthetic bands, Sicily study area: PCA-1 on the <b>left</b>, NDWI on the <b>right</b> and MNDWI on the <b>bottom</b>.</p> Full article ">Figure 14
<p>Geolocation of the selected zones to show details of K-means results: Port of Naples (Zone A) on the <b>left</b> and coastal area of San Teodoro (Zone B) on the <b>right</b>.</p> Full article ">Figure 15
<p>K-means clustering in Zone A (Port of Naples) for synthetic images derived by L9: results for PCA-1 image on the <b>left</b> and for NDWI image on the <b>right</b>.</p> Full article ">Figure 16
<p>K-means clustering in Zone B (coastal area of San Teodoro) for synthetic images derived by –m: results for PCA-1 image (on the <b>left</b>) and for NDWI image (on the <b>right</b>).</p> Full article ">Figure 17
<p>Geolocation of the selected zones to show details of automatically vectorized coastlines obtained through PCA and through NDWI: Torre del Greco (Zone C) on the <b>left</b> and San Giovanni (Zone D) on the <b>right</b>.</p> Full article ">Figure 18
<p>Details of automatically vectorized coastlines extracted from Landsat 9 OLI images in the area of Torre del Greco (Zone C): on the <b>left</b> the coastline from PCA (in green), on the <b>right</b> the coastline from NDWI (in blue); in both cases the reference coastline resulting from RGB visual interpretation and manual vectorization (in red) is reported for comparison.</p> Full article ">Figure 19
<p>Details of automatically vectorized coastlines extracted from S2–20 m in the area of San Giovanni (Zone D): on the <b>left</b> the coastline from PCA (in green), on the <b>right</b> the coastline from NDWI (in blue); in both cases the reference coastline resulting from RGB visual interpretation and manual vectorization (in red) is reported for comparison.</p> Full article ">Figure 20
<p>Details of automatically vectorized coastlines extracted from –m in the area of San Giovanni (Zone D): on the <b>left</b> the coastline from PCA (in green), on the <b>right</b> the coastline from NDWI (in blue); in both cases the reference coastline resulting from RGB visual interpretation and manual vectorization (in red) is reported for comparison.</p> Full article ">
16 pages, 4039 KiB  
Article
A Soil Moisture and Vegetation-Based Susceptibility Mapping Approach to Wildfire Events in Greece
by Kyriakos Chaleplis, Avery Walters, Bin Fang, Venkataraman Lakshmi and Alexandra Gemitzi
Remote Sens. 2024, 16(10), 1816; https://doi.org/10.3390/rs16101816 - 20 May 2024
Viewed by 601
Abstract
Wildfires in Mediterranean areas are becoming more frequent, and the fire season is extending toward the spring and autumn months. These alarming findings indicate an urgent need to develop fire susceptibility methods capable of identifying areas vulnerable to wildfires. The present work aims [...] Read more.
Wildfires in Mediterranean areas are becoming more frequent, and the fire season is extending toward the spring and autumn months. These alarming findings indicate an urgent need to develop fire susceptibility methods capable of identifying areas vulnerable to wildfires. The present work aims to uncover possible soil moisture and vegetation condition precursory signals of the largest and most devastating wildfires in Greece that occurred in 2021, 2022, and 2023. Therefore, the time series of two remotely sensed datasets–MAP L4 Soil Moisture (SM) and Landsat 8 NDVI, which represent vegetation and soil moisture conditions—were examined before five destructive wildfires in Greece during the study period. The results of the analysis highlighted specific properties indicative of fire-susceptible areas. NDVI in all fire-affected areas ranged from 0.13 to 0.35, while mean monthly soil moisture showed negative anomalies in the spring periods preceding fires. Accordingly, fire susceptibility maps were developed, verifying the usefulness of remotely sensed information related to soil moisture and NDVI. This information should be used to enhance fire models and identify areas at risk of wildfires in the near future. Full article
(This article belongs to the Special Issue Satellite Soil Moisture Estimation, Assessment, and Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Location map of the wildfires analyzed in the present work. Numbers correspond to those indicated in <a href="#remotesensing-16-01816-t001" class="html-table">Table 1</a>. Red designated areas correspond to wildfires that are used for the method development (Fires 1–5), while yellow is used for verification (Fires 6–11) only.</p> Full article ">Figure 2
<p>Fire susceptibility map flow chart.</p> Full article ">Figure 3
<p>Time series of soil moisture and (<b>a</b>) number of fires, (<b>b</b>) burned area, in Greece (2015–2023).</p> Full article ">Figure 4
<p>Comparison of in situ SM data and SMAP L4 surface SM and root zone SM in the Rhodope area (NE Greece) during the study period.</p> Full article ">Figure 5
<p>Time series graphs of NDVI and SMAP L4 surface SM anomalies for five wildfires during 2021–2023 in Greece. Arrows indicate the date of wildfire occurrence.</p> Full article ">Figure 6
<p>Spatial distribution of NDVI (<b>a</b>–<b>c</b>) and SM anomalies (<b>d</b>–<b>f</b>) satisfying the fire susceptibility criteria in the broader area of Greece for 2021, 2022, and 2023.</p> Full article ">Figure 7
<p>Spatial distribution of fire susceptibility (<b>a</b>–<b>c</b>), VDI at the NUTS3 level (<b>d</b>–<b>f</b>), and cross-tabulation results with fire susceptibility mapping of the present work (<b>g</b>–<b>i</b>) for Greece during 2021, 2022, and 2023.</p> Full article ">
20 pages, 14379 KiB  
Article
Integrating Climate and Satellite Data for Multi-Temporal Pre-Harvest Prediction of Head Rice Yield in Australia
by Allister Clarke, Darren Yates, Christopher Blanchard, Md. Zahidul Islam, Russell Ford, Sabih-Ur Rehman and Robert Paul Walsh
Remote Sens. 2024, 16(10), 1815; https://doi.org/10.3390/rs16101815 - 20 May 2024
Viewed by 515
Abstract
Precise and prompt predictions of crop yields are crucial for optimising farm management, post-harvest operations, and marketing strategies within the agricultural sector. While various machine learning approaches have been employed to forecast crop yield, their application to grain quality, particularly head rice yield [...] Read more.
Precise and prompt predictions of crop yields are crucial for optimising farm management, post-harvest operations, and marketing strategies within the agricultural sector. While various machine learning approaches have been employed to forecast crop yield, their application to grain quality, particularly head rice yield (HRY), is less explored. This research collated crop-level HRY data across four seasons (2017/18–2020/21) from Australia’s rice-growing region. Models were developed using the XGBoost algorithm trained at varying time steps up to 16 weeks pre-harvest. The study compared the accuracy of models trained on datasets with climate data alone or paired with vegetative indices using two- and four-week aggregations. The results suggest that model accuracy increases as the harvest date approaches. The dataset combining climate and vegetative indices aggregated over two weeks surpassed industry benchmarks early in the season, achieving the highest accuracy two weeks before harvest (LCCC = 0.65; RMSE = 6.43). The analysis revealed that HRY correlates strongly with agroclimatic conditions nearer harvest, with the significance of vegetative indices-based features increasing as the season progresses. These features, indicative of crop and grain maturity, could aid growers in determining optimal harvest timing. This investigation offers valuable insights into grain quality forecasting, presenting a model adaptable to other regions with accessible climate and satellite data, consequently enhancing farm- and industry-level decision-making. Full article
(This article belongs to the Topic Geospatial Digital Innovations for Smart Agriculture and Forestry)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Location of the rice growing regions where crop records were collated in this study. The field boundaries included in this study are coloured to highlight the rice-growing regions of the industry: MIA—Murrumbidgee Irrigation Area; CIA—Coleambally Irrigation Area; EMV—Eastern Murray Valley; WMV—Western Murray Valley. Australian Grain Storage (AGS) receival site locations are shown in orange. State reporting regions were taken from the Australian Bureau of Statistics—Statistical Areas Level 4 (SA4). Datum GDA2020 (print in colour.)</p> Full article ">Figure 2
<p>HRY distribution across four rice growing seasons in the Riverina region of Australia. The histograms represent the frequency of HRY records, with each colour corresponding to a different season, as indicated. The dataset average HRY (58.7%) is displayed as the dashed vertical line.</p> Full article ">Figure 3
<p>Methods workflow diagram illustrating the replicated data processing and model development steps for the two-week (2 WK) and four-week (4 WK) dataset construction methods.</p> Full article ">Figure 4
<p>Comparison of predictive model accuracies for HRY over two-week intervals leading up to harvest delivery. Model performance is evaluated using LCCC and RMSE, with intervals spanning from 14 weeks before harvest to the point of delivery.</p> Full article ">Figure 5
<p>Comparison of predictive model accuracies for HRY over four-week intervals leading up to harvest delivery. Model performance is evaluated using LCCC and RMSE, with intervals spanning from 12 weeks before harvest to the point of delivery.</p> Full article ">Figure 6
<p>Comparison of season-by-industry level prediction RMSE (HRY%) in the four-week and two-week data aggregation methods.</p> Full article ">Figure 7
<p>Nested donut charts categorise the total mean SHAP of each model at the first level by the two-week time interval and delivery stand (categories shown in colour) and, secondly, by the type of data the stage-dependent features were calculated from. MET = meteorological data; RS = remote sensing; CROP = crop management.</p> Full article ">Figure 8
<p>Correlation between satellite-derived vegetation indices and harvest grain moisture percentage, stratified by HRY percentage. (<b>a</b>) Correlation between average EVI 2–4 weeks pre-harvest and harvest grain moisture percentage, (<b>b</b>) Correlation between average NDVI 0–2 weeks pre-harvest and harvest grain moisture percentage. Data points are colour-coded according to HRY ranging from lower (green/yellow) to higher (purple/blue) yields. The linear fit lines indicate the trend and strength of the relationship between these vegetation indices and the grain moisture content at harvest.</p> Full article ">Figure 9
<p>SHAP value partial dependence plots of 2–4 wPH average EVI and 0–2 wPH average NDVI and HRY. The points represent each instance in the dataset, while the point colour indicates the feature value from low (yellow) to high (purple). The red lines in the plots indicate the smoothed LOESS relationship between the explanatory variables (on the x-axis) and their respective SHAP values (on the y-axis) (print in colour).</p> Full article ">Figure 10
<p>Temporal comparison of the two-week climate + GIS model prediction in the CY21 season compared with the observed industry benchmark provided by SunRice. Prediction time points are shown on the x-axis, from 14 weeks before harvest to the grain elevator (delivery). The vertical dashed black line indicates the period from which the model was more accurate than the industry benchmark.</p> Full article ">
19 pages, 3360 KiB  
Article
A Multi-Satellite Space Environment Risk Prediction and Real-Time Warning System for Satellite Safety Management
by Ning Kang, Liguo Zhang, Weiguo Zong, Pan Huang, Yuqiang Zhang, Chen Zhou, Jian Qiao and Bingsen Xue
Remote Sens. 2024, 16(10), 1814; https://doi.org/10.3390/rs16101814 - 20 May 2024
Viewed by 403
Abstract
In response to the need for a space security situation assessment during orbit, the multi-satellite space environmental risk prediction and early warning system is based on the detection results of the space weather payload of the Fengyun 4A and 4B satellites, as well [...] Read more.
In response to the need for a space security situation assessment during orbit, the multi-satellite space environmental risk prediction and early warning system is based on the detection results of the space weather payload of the Fengyun 4A and 4B satellites, as well as the prediction results of the National Space Weather Center, for the first time. By comprehensively utilizing some open-source data, it is the first time that we have achieved a 24 h advanced prediction of the space environment high-energy proton, low-energy particle, and high-energy electron risks for the safety of the Fengyun-series high-orbit satellites, and a real-time warning of satellite single-event upset, surface charging, and deep charging risks. The automation system has preliminarily achieved an intelligent space risk assessment for the safety of multiple stationary meteorological satellites, effectively improving the application efficiency of the space environmental data and the products of Fengyun-series satellites. The business status is stable in operation, and the resulting error between the predicted results of various risk indices and the measured data was less than one level. The warning accuracy was better than 90%. This article uses the system for the first time, to use Fengyun satellite data to, accurately and in a timely manner, predict and warn us about the low-energy particles and surface charging high-risk levels of the Fengyun 4A and 4B satellites during the typical space weather event on 21 April 2023, in response to the impact of complex spatial environmental factors on the safety of Fengyun-series high-orbit satellites. The construction and operation of a multi-satellite space environmental risk prediction and early warning system can provide a reference for the safety work of subsequent satellite ground system operations. Full article
(This article belongs to the Special Issue Using Remote Sensing Satellites to Explore the Electromagnetic Environment and Natural Hazard Disturbances in Space)
►▼ Show Figures

Figure 1

Figure 1
<p>Flow chart of spatial environment comprehensive risk index.</p> Full article ">Figure 2
<p>Flow chart of multi-satellite space environmental risk prediction and real-time warning system.</p> Full article ">Figure 3
<p>Demonstration of the interface for multi-satellite space environmental risk prediction and real-time warning system.</p> Full article ">Figure 4
<p>Variation of solar wind speed index during on 21 April 2023.</p> Full article ">Figure 5
<p>The 24 h time evolution of FY-4A space environment prediction and real-time warning risk level on 21 April 2023.</p> Full article ">Figure 6
<p>The 24 h time evolution of FY-4B space environment prediction and real-time warning risk level on 21 April 2023.</p> Full article ">Figure 7
<p>The 24 h time evolution of risk levels for comprehensive prediction and real-time warning of space environment on 21 April 2023.</p> Full article ">Figure 8
<p>Changes in surface differential charging potential of three probes A/B/C for FY-4A satellite space weather load and surface charging real-time risk level switch on 21 April 2023.</p> Full article ">Figure 9
<p>Changes in surface differential charging potential of three probes A/B/C for FY-4B satellite space weather load and surface charging real-time risk level switch on 21 April 2023.</p> Full article ">
21 pages, 4981 KiB  
Article
A Segmented Sliding Window Reference Signal Reconstruction Method Based on Fuzzy C-Means
by Haobo Liang, Yuan Feng, Yushi Zhang, Xingshuai Qiao, Zhi Wang and Tao Shan
Remote Sens. 2024, 16(10), 1813; https://doi.org/10.3390/rs16101813 - 20 May 2024
Viewed by 457
Abstract
Reference signal reconstruction serves as a crucial technique for suppressing multipath interference and noise in the reference channel of passive radar. Aiming at the challenge of detecting Low-Slow-Small (LSS) targets using Digital Terrestrial Multimedia Broadcasting (DTMB) signals, this article proposes a novel segmented [...] Read more.
Reference signal reconstruction serves as a crucial technique for suppressing multipath interference and noise in the reference channel of passive radar. Aiming at the challenge of detecting Low-Slow-Small (LSS) targets using Digital Terrestrial Multimedia Broadcasting (DTMB) signals, this article proposes a novel segmented sliding window reference signal reconstruction method based on Fuzzy C-Means (FCM). By partitioning the reference signals based on the structure of DTMB signal frames, this approach compensates for frequency offset and sample rate deviation individually for each segment. Additionally, FCM clustering is utilized for symbol mapping reconstruction. Both simulation and experimental results show that the proposed method significantly suppresses constellation diagram divergence and phase rotation, increases the adaptive cancellation gain and signal-to-noise ratio (SNR), and in the meantime reduces the computation cost. Full article
(This article belongs to the Topic Radar Signal and Data Processing with Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Frame structure of DTMB signal.</p> Full article ">Figure 2
<p>PN420 linear feedback shift register.</p> Full article ">Figure 3
<p>The digital television transmitter block diagram.</p> Full article ">Figure 4
<p>Reconstruction scheme at the reception.</p> Full article ">Figure 5
<p>Principle of channel equalization.</p> Full article ">Figure 6
<p>Comparison between this article’s blocking and traditional blocking.</p> Full article ">Figure 7
<p>Frame body constellation diagram.</p> Full article ">Figure 8
<p>The process of the segmented sliding window reconstruction method based on FCM.</p> Full article ">Figure 9
<p>Structure of the 2N + 1 order transversal filter.</p> Full article ">Figure 10
<p>The BER versus SNR for different reconstruction methods. (<b>a</b>) 4QAM; (<b>b</b>) 16QAM; (<b>c</b>) 32QAM; (<b>d</b>) 64QAM.</p> Full article ">Figure 10 Cont.
<p>The BER versus SNR for different reconstruction methods. (<b>a</b>) 4QAM; (<b>b</b>) 16QAM; (<b>c</b>) 32QAM; (<b>d</b>) 64QAM.</p> Full article ">Figure 11
<p>(<b>a</b>) The cancellation gain versus block length; (<b>b</b>) the algorithm running time changes with the block length.</p> Full article ">Figure 12
<p>The 900-th frame constellation diagram. (<b>a</b>) Traditional unpartitioned reconstruction method; (<b>b</b>) segmented sliding window reconstruction method based on FCM.</p> Full article ">Figure 13
<p>NLMS adaptive cancellation gain using PN420 DTMB signal.</p> Full article ">Figure 14
<p>Comparison of LSS target detection effects using PN420 DTMB signal. (<b>a</b>) Traditional unpartitioned reconstruction method; (<b>b</b>) segmented sliding window reconstruction method based on FCM.</p> Full article ">Figure 15
<p>The 864-th frame constellation diagram. (<b>a</b>) Traditional unpartitioned reconstruction method; (<b>b</b>) segmented sliding window reconstruction method based on FCM.</p> Full article ">Figure 16
<p>NLMS adaptive cancellation gain using PN595 DTMB signal.</p> Full article ">Figure 17
<p>Comparison of LSS target detection effects using PN595 DTMB signal. (<b>a</b>) Traditional unpartitioned reconstruction method; (<b>b</b>) segmented sliding window reconstruction method based on FCM.</p> Full article ">
18 pages, 4896 KiB  
Article
Global Inversion of Lunar Surface Oxides by Adding Chang’e-5 Samples
by Shuangshuang Wu, Jianping Chen, Chenli Xue, Yiwen Pan and Cheng Zhang
Remote Sens. 2024, 16(10), 1812; https://doi.org/10.3390/rs16101812 - 20 May 2024
Viewed by 419
Abstract
The chemical distribution on the lunar surface results from the combined effects of both endogenic and exogenic geological processes. Exploring global maps of chemical composition helps to gain insights into the compositional variation among three major geological units, unraveling the geological evolution of [...] Read more.
The chemical distribution on the lunar surface results from the combined effects of both endogenic and exogenic geological processes. Exploring global maps of chemical composition helps to gain insights into the compositional variation among three major geological units, unraveling the geological evolution of the Moon. The existing oxide abundance maps were obtained from spectral images of remote sensing and geochemical data from samples returned by Apollo and Luna, missing the chemical characteristics of the Moon’s late critical period. In this study, by adding geochemical data from Chang’e (CE)-5 lunar samples, we construct inversion models between the Christiansen feature (CF) and oxide abundance of lunar samples using the particle swarm optimization–extreme gradient boosting (PSO-XGBoost) algorithm. Then, new global oxide maps (Al2O3, CaO, FeO, and MgO) and Mg# with the resolution of 32 pixels/degree (ppd) were produced, which reduced the space weathering effect to some extent. The PSO-XGBoost models were compared with partial least square regression (PLSR) models and four previous results, indicating that PSO-XGBoost models possess the capability to effectively describe nonlinear relationships between CF and oxide abundance. Furthermore, the average contents of our results and the Diviner results for 21 major maria demonstrate high correlations, with R2 of 0.95, 0.82, 0.95, and 0.86, respectively. In addition, a new Mg# map was generated, which reveals different magmatic evolutionary processes in the three geologic units. Full article
(This article belongs to the Special Issue Planetary Geologic Mapping and Remote Sensing II)
►▼ Show Figures

Figure 1

Figure 1
<p>The corrected CF map derived by Lucey et al. [<a href="#B32-remotesensing-16-01812" class="html-bibr">32</a>].</p> Full article ">Figure 2
<p>Linear or nonlinear relationships between (<b>a</b>) Al<sub>2</sub>O<sub>3</sub>, (<b>b</b>) CaO, (<b>c</b>) FeO, and (<b>d</b>) MgO and the CF values for 49 lunar sampling points.</p> Full article ">Figure 3
<p>Scatter plots of the measured and predicted values for (<b>a</b>) Al<sub>2</sub>O<sub>3</sub>, (<b>b</b>) CaO, (<b>c</b>) FeO, and (<b>d</b>) MgO from PSO-XGBoost models. Error bars represent 95% confidence intervals around the oxides’ predicted values.</p> Full article ">Figure 4
<p>Scatter plots of the measured and predicted values for (<b>a</b>) Al<sub>2</sub>O<sub>3</sub>, (<b>b</b>) CaO, (<b>c</b>) FeO, and (<b>d</b>) MgO from PLSR models.</p> Full article ">Figure 5
<p>Global distribution maps of (<b>a</b>) Al<sub>2</sub>O<sub>3</sub>, (<b>b</b>) CaO, (<b>c</b>) FeO and (<b>d</b>) MgO abundances.</p> Full article ">Figure 6
<p>The LRO WAC and compositional distributions of three interesting regions including the Kepler crater (8.1°N, 38.0°W), Mare Moscoviense (148°E, 27°N), and the Helmet dome (16.6°S, 31.4°W).</p> Full article ">Figure 7
<p>Average abundances of four oxides in the (<b>a</b>) Moon (70°N/S), (<b>b</b>) maria, (<b>c</b>) highlands, and (<b>d</b>) SPA basin, according to five sets of results.</p> Full article ">Figure 8
<p>Sketch map showing the locations of the three geological units (maria are marked in blue, highlands are marked in orange, and the SPA basin is identified in green).</p> Full article ">Figure 9
<p>Scatter plot with error bars for (<b>a</b>) Al<sub>2</sub>O<sub>3</sub> (<b>b</b>) CaO (<b>c</b>) FeO and (<b>d</b>) MgO abundances from Ma et al. [<a href="#B28-remotesensing-16-01812" class="html-bibr">28</a>] and this study, for 21 major maria. Error bar indicates the standard deviation.</p> Full article ">Figure 10
<p>Mg# map across the Moon (70°N/S).</p> Full article ">
17 pages, 6680 KiB  
Article
Monitoring of Low Chl-a Concentration in Hulun Lake Based on Fusion of Remote Sensing Satellite and Ground Observation Data
by Siyuan Zhang, Yinglan A, Libo Wang, Yuntao Wang, Xiaojing Zhang, Yi Zhu and Guangwen Ma
Remote Sens. 2024, 16(10), 1811; https://doi.org/10.3390/rs16101811 - 20 May 2024
Viewed by 563
Abstract
China’s northern Hulun Lake is a significant body of water internationally. The issue of eutrophication has gained prominence in recent years. The achievement of precise chlorophyll-a (Chl-a) monitoring is crucial for safeguarding Hulun Lake’s ecosystem. The machine learning-based remote sensing inversion method has [...] Read more.
China’s northern Hulun Lake is a significant body of water internationally. The issue of eutrophication has gained prominence in recent years. The achievement of precise chlorophyll-a (Chl-a) monitoring is crucial for safeguarding Hulun Lake’s ecosystem. The machine learning-based remote sensing inversion method has been shown to be effective in capturing the intricate relationship between independent and dependent variables; however, it lacks a priori knowledge and is limited by the quality of remote sensing data sources. The relationship between independent and dependent variables can be more accurately simulated with the use of suitable auxiliary variables. Therefore, three machine learning models—random forest (RF), adaptive boosting (AdaBoost), and extreme gradient boosting (XGBoost)—were established in this study using meteorological observation parameters as auxiliary variables combined with Sentinel-2 satellite image remote sensing band combinations as independent variables and measured Chl-a data as dependent variables. The estimation effects before and after the fusion of meteorological ground observation data were compared, and the best model was used to estimate the spatial–temporal variation trend of Chl-a in the regional water body. The results show that (1) the addition of meteorological parameters as auxiliary variables improved the precision of the three machine models; the decision coefficient (R2) rose by 7.25%, 5.71%, and 7.20%, respectively, to 0.76, 0.66, and 0.73. (2) The concentration of Chl-a in the lake region was projected from June to October 2019 to October 2021 using the RF optimal estimating model of meteorological fusion. The northeast, southwest, and south of the lake were where the comparatively high concentration values of Chl-a were located, whereas the lake’s center had a generally low concentration of the substance. Chromatically, Chl-a typically peaked in August after initially increasing and then declining. (3) The three rivers that feed into the river have varying levels of water pollution, with chemical oxygen demand (COD) and total nitrogen (TN) pollution being the most severe. This is what primarily caused the higher levels of Chl-a in the northeast, southwest, and south. This study is crucial for the preservation and restoration of Hulun Lake’s natural ecosystem and offers some technical support for the monitoring of the lake’s concentration of Chl-a. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic of study area.</p> Full article ">Figure 2
<p>Workflow chart of study.</p> Full article ">Figure 3
<p>Distribution of mean Chl-a concentrations at each sample point (<b>left</b>) and multi-year monthly mean concentrations of Chl-a for the 5 samples (<b>right</b>).</p> Full article ">Figure 4
<p>Weekly average values of temperature, precipitation, relative humidity, wind speed, and sunshine duration at four meteorological stations from 2016 to 2021.</p> Full article ">Figure 5
<p>Comparison of predicted Chl-a concentration with measured Chl-a concentration by the three machine models (without meteorological auxiliary data on the left side and with meteorological auxiliary data on the right side).</p> Full article ">Figure 6
<p>Estimation map of Chl-a concentration in Hulun Lake.</p> Full article ">Figure 7
<p>Monthly concentration change in main water quality parameters of rivers entering lake.</p> Full article ">Figure 8
<p>Hulun Lake Chl-a concentration monthly change box diagram from June to October. (The center line of the box represents the median value; the dots represent the average value; the bottom and top of the box represent the upper and lower quartiles, respectively; and the bottom and top lines represent the 10th and 90th percentiles, respectively.)</p> Full article ">
20 pages, 16712 KiB  
Article
Effects of Land Use/Cover Change on Terrestrial Carbon Stocks in the Yellow River Basin of China from 2000 to 2030
by Jiejun Zhang, Jie Yang, Pengfei Liu, Yi Liu, Yiwen Zheng, Xiaoyu Shen, Bingchen Li, Hongquan Song and Zongzheng Liang
Remote Sens. 2024, 16(10), 1810; https://doi.org/10.3390/rs16101810 - 20 May 2024
Viewed by 494
Abstract
Accurately assessing and predicting the impacts of land use changes on ecosystem carbon stocks in the Yellow River Basin (YRB) and exploring the optimization of land use structure to increase ecosystem carbon stocks are of great practical significance for China to achieve the [...] Read more.
Accurately assessing and predicting the impacts of land use changes on ecosystem carbon stocks in the Yellow River Basin (YRB) and exploring the optimization of land use structure to increase ecosystem carbon stocks are of great practical significance for China to achieve the goal of “double carbon”. In this study, we used multi-year remote sensing data, meteorological data and statistical data to measure the ecosystem carbon stock in the YRB from 2000 to 2020 based on the InVEST model, and then simulated and measured the ecosystem carbon stock under four different land use scenarios coupled with the FLUS model in 2030. The results show that, from 2000 to 2020, urban expansion in the YRB continued, but woodland and grassland grew more slowly. Carbon stock showed an increasing trend during the first 20 years, with an overall increase of 7.2 megatons, or 0.23%. Simulating the four land use scenarios in 2030, carbon stock will decrease the most under the cropland protection scenario, with a decrease of 17.7 megatons compared with 2020. However, carbon stock increases the most under the ecological protection scenario, with a maximum increase of 9.1 megatons. Furthermore, distinct trends in carbon storage were observed across different regions, with significant increases in the upstream under the natural development scenario, in the midstream under the ecological protection scenario and in the downstream under the cropland protection scenario. We suggest that the upstream should maintain the existing development mode, with ecological protection prioritized in the middle reaches and farmland protection prioritized in the lower reaches. This study provides a scientific basis for the carbon balance, land use structure adjustment and land management decision-making in the YRB. Full article
(This article belongs to the Special Issue Assessment of Ecosystem Services Based on Satellite Data)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Overview of the Yellow River Basin in China.</p> Full article ">Figure 2
<p>Spatial distribution of land use in the Yellow River Basin, 2000–2020.</p> Full article ">Figure 3
<p>LUCC in the Yellow River Basin from 2000 to 2020. (<b>a</b>) Area of land use type, (<b>b</b>) percentage of area in land use type.</p> Full article ">Figure 4
<p>Changes in carbon storage in the Yellow River Basin of various land use types and carbon pools from 2000 to 2020. (<b>a</b>,<b>b</b>) Carbon storage of land use type, (<b>c</b>,<b>d</b>) changes in carbon storage in land use type.</p> Full article ">Figure 5
<p>Changes in carbon storage of LUCC and carbon pools in the sub-basins from 2000 to 2020.</p> Full article ">Figure 6
<p>Spatial distribution of carbon storage in the Yellow River Basin from 2000 to 2020.</p> Full article ">Figure 7
<p>Spatial distribution of future LUCC in the Yellow River Basin under different scenarios.</p> Full article ">Figure 8
<p>Future LUCC in the Yellow River Basin under different scenarios. (<b>a</b>) Area of land use type, (<b>b</b>) percentage of area in land use type.</p> Full article ">Figure 9
<p>Carbon storage and changes in land use and carbon pools in the Yellow River Basin under different future scenarios. (<b>a</b>,<b>b</b>) Carbon storage of land use type, (<b>c</b>,<b>d</b>) changes of carbon storage in land use type.</p> Full article ">Figure 10
<p>Spatial distribution of carbon storage in the Yellow River Basin under different future scenarios.</p> Full article ">
18 pages, 16994 KiB  
Article
Inverse Synthetic Aperture Radar Imaging of Space Targets Using Wideband Pseudo-Noise Signals with Low Peak-to-Average Power Ratio
by Simon Anger, Matthias Jirousek, Stephan Dill and Markus Peichl
Remote Sens. 2024, 16(10), 1809; https://doi.org/10.3390/rs16101809 - 20 May 2024
Viewed by 480
Abstract
With the number of new satellites increasing dramatically, comprehensive space surveillance is becoming increasingly important. Therefore, high-resolution inverse synthetic aperture radar (ISAR) imaging of satellites can provide an in-situ assessment of the satellites. This paper demonstrates that pseudo-noise signals can also be used [...] Read more.
With the number of new satellites increasing dramatically, comprehensive space surveillance is becoming increasingly important. Therefore, high-resolution inverse synthetic aperture radar (ISAR) imaging of satellites can provide an in-situ assessment of the satellites. This paper demonstrates that pseudo-noise signals can also be used for satellite imaging, in addition to classical linear frequency-modulated chirp signals. Pseudo-noise transmission signals offer the advantage of very low cross-correlation values. This, for instance, enables the possibility of a system with multiple channels transmitting instantaneously. Furthermore, it can significantly reduce signal interference with other systems operating in the same frequency spectrum, which is of particular interest for high-bandwidth, high-power systems such as satellite imaging radars. A new routine has been introduced to generate a wideband pseudo-noise signal with a peak-to-average power ratio (PAPR) similar to that of a chirp signal. This is essential for applications where the transmit signal power budget is sharply limited by the high-power amplifier. The paper presents both theoretical descriptions and analysis of the generated pseudo-noise signal as well as the results of an imaging measurement of a real space target using the introduced pseudo-noise signals. Full article
(This article belongs to the Special Issue Radar for Space Observation: Systems, Methods and Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Iterative generation of an amplitude and bandwidth-limited pseudo-noise signal with a low peak-to-average power ratio (PAPR). The generated pseudo-noise signal is limited to the peak-to-peak amplitude <math display="inline"><semantics> <msub> <mi>u</mi> <mrow> <mi>p</mi> <mi>p</mi> </mrow> </msub> </semantics></math> in the time domain and to the transfer function <math display="inline"><semantics> <mrow> <msub> <mi>H</mi> <mrow> <mi>B</mi> <mi>P</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>f</mi> <mo>)</mo> </mrow> </mrow> </semantics></math> in the frequency domain.</p> Full article ">Figure 2
<p>Three different pulse signals with the same average power of 10 dBm. A Gaussian noise signal with the variance <math display="inline"><semantics> <mrow> <msup> <mi>σ</mi> <mn>2</mn> </msup> <mo>=</mo> <mn>0.5</mn> </mrow> </semantics></math> V and the corresponding probability density function (PDF) of the time domain amplitude values. A classical linear frequency modulated chirp signal and the corresponding PDF. A sharp amplitude limited pseudo-noise signal generated with the procedure described in <a href="#remotesensing-16-01809-f001" class="html-fig">Figure 1</a> and the corresponding PDF.</p> Full article ">Figure 3
<p>Short-time Fourier transform (STFT) of a chirp, Gaussian noise, and the generated pseudo-noise signal. The STFT indicates that the pseudo-noise signal exhibits very similar characteristics to a Gaussian noise signal.</p> Full article ">Figure 4
<p>Simulated point spread functions using a Gaussian noise as well as the pseudo-noise signal (PNS) with a signal bandwidth of 2.8 GHz. Both exhibit a nearly ideal characteristic down to an amplitude below <math display="inline"><semantics> <mrow> <mo>−</mo> <mn>35</mn> </mrow> </semantics></math> dB. Similar to the Gaussian noise signal, a pair of PNSs exhibits very low cross-correlation (CCF) values, indicating its quasi-orthogonality.</p> Full article ">Figure 5
<p>Simulated point spread functions using the PNS with different signal bandwidths.</p> Full article ">Figure 6
<p>Wideband ambiguity function of a linear frequency modulated (LFM) chirp signal (<b>left</b>) and a generated pseudo-noise (<b>right</b>) for typical radial velocities <math display="inline"><semantics> <msub> <mi>v</mi> <mi>r</mi> </msub> </semantics></math> as they occur in radar-based imaging of satellites.</p> Full article ">Figure 7
<p>Comparison of the Spektral Kurtosis (SK) of Gaussian noise and the generated pseudo-noise signal. The orange highlighted section shows the frequency range of the signal from <math display="inline"><semantics> <mrow> <mn>200</mn> </mrow> </semantics></math> MHz to <math display="inline"><semantics> <mrow> <mn>3</mn> </mrow> </semantics></math> GHz. The low SK values of the pseudo-noise signal show its stationary signal behavior and hence its low LPI properties.</p> Full article ">Figure 8
<p>A photo of the IoSiS setup is shown, featuring a 9 m transmitting antenna in a Cassegrain configuration and two 1.8 m receiving antennas. The radar electronics responsible for signal generation and acquisition are housed in a nearby container.</p> Full article ">Figure 9
<p>Simplified RF diagram of the IoSiS radar, showing the transmit channel and one of the two receive channels. Fully digital signal generation and acquisition enables the generation of arbitrary transmit signals. Filter banks in each channel enable operation in the upper (USB) and lower side band (LSB).</p> Full article ">Figure 10
<p>Measured point spread function (PSF) using the pseudo-noise signal routed through the calibration path of the radar system (<b>left</b>). Spectral weighted PSF of the pseudo-noise signal (<b>right</b>). In the non-weighted case, the half-power beam width (HPBW) corresponds to the applied pulse signal bandwidth of <math display="inline"><semantics> <mrow> <mn>2.8</mn> </mrow> </semantics></math> GHz and equals about <math display="inline"><semantics> <mrow> <mn>54</mn> </mrow> </semantics></math> mm. In both cases, the low cross-correlation (CCF) values show the quasi-orthogonality of a pair of pseudo-noise signals.</p> Full article ">Figure 11
<p>Phase values extracted in the maximum amplitude of the ACF point spread function (PSF) measured using the calibration path of the IoSiS radar (<b>left</b>). Corresponding phase probability function (PDF) of the measured phase values (<b>right</b>). With a phase standard deviation of <math display="inline"><semantics> <mrow> <mi>σ</mi> <mo>=</mo> <mn>0.3</mn> <mo>°</mo> </mrow> </semantics></math>, the phase is highly stable, which is a prerequisite for using the PNS in an ISAR system.</p> Full article ">Figure 12
<p>ISAR image of the ISS using a PNS transmit signal together with the IoSiS radar system. The achieved spatial resolution in range and azimuth direction was evaluated by analyzing two robust scatterers. The arrows indicate the positions of the scatterers under investigation.</p> Full article ">Figure 13
<p>ISAR image of the isolated robust scatterers of the ISS, together with the cuts in the cardinal direction, to analyze the PSF and to evaluate the PNS signal used. The HPBW in the focal plane (<math display="inline"><semantics> <mrow> <mo>Δ</mo> <mi>r</mi> </mrow> </semantics></math>) of both are in the expected range. In the azimuth direction, both exhibit a slightly higher HPBW (<math display="inline"><semantics> <mrow> <mo>Δ</mo> <msub> <mi>r</mi> <mrow> <mi>A</mi> <mi>z</mi> </mrow> </msub> </mrow> </semantics></math>) considering the integration angle.</p> Full article ">
25 pages, 11676 KiB  
Article
Deep Learning-Based Automatic River Flow Estimation Using RADARSAT Imagery
by Samar Ziadi, Karem Chokmani, Chayma Chaabani and Anas El Alem
Remote Sens. 2024, 16(10), 1808; https://doi.org/10.3390/rs16101808 - 20 May 2024
Viewed by 1476
Abstract
Estimating river flow is a key parameter for effective water resource management, flood risk prevention, and hydroelectric facilities planning. Yet, traditional gauging methods are not reliable under very high flows or extreme events. Hydrometric network stations are often sparse, and their spatial distribution [...] Read more.
Estimating river flow is a key parameter for effective water resource management, flood risk prevention, and hydroelectric facilities planning. Yet, traditional gauging methods are not reliable under very high flows or extreme events. Hydrometric network stations are often sparse, and their spatial distribution is not optimal. Therefore, many river sections cannot be monitored using traditional flow measurements and observations. In the last few decades, satellite sensors have been considered as complementary observation sources to traditional water level and flow measurements. This kind of approach has provided a way to maintain and expand the hydrometric observation network. Remote sensing data can be used to estimate flow from rating curves that relate instantaneous flow (Q) to channel cross-section geometry (effective width or depth of the water surface). Yet, remote sensing has limitations, notably its dependence on rating curves. Due to their empirical nature, rating curves are limited to specific river sections (reaches) and cannot be applied to other watercourses. Recently, deep-learning techniques have been successfully applied to hydrology. The primary goal of this study is to develop a deep-learning approach for estimating river flow in the Boreal Shield ecozone of Eastern Canada using RADARSAT-1 and -2 imagery and convolutional neural networks (CNN). Data from 39 hydrographic sites in this region were used in modeling. A new CNN architecture was developed to provide a straightforward estimation of the instantaneous river flow rate. Our results yielded a coefficient of determination (R2) and a Nash–Sutcliffe value of 0.91 and a root mean square error of 33 m3/s. Notably, the model performs exceptionally well for rivers wider than 40 m, reflecting its capability to adapt to varied hydrological contexts. These results underscore the potential of integrating advanced satellite imagery with deep learning to enhance hydrological monitoring across vast and remote areas. Full article
(This article belongs to the Topic Hydrology and Water Resources Management)
►▼ Show Figures

Figure 1

Figure 1
<p>Overview of the study area.</p> Full article ">Figure 2
<p>Radar image coverage across the study area.</p> Full article ">Figure 3
<p>HAND computation workflow (inspired from [<a href="#B65-remotesensing-16-01808" class="html-bibr">65</a>]).</p> Full article ">Figure 4
<p>General flowchart of the river flow estimation approach.</p> Full article ">Figure 5
<p>Clipping data.</p> Full article ">Figure 6
<p>Data augmentation.</p> Full article ">Figure 7
<p>An overview of the convolutional neural network (CNN) architecture.</p> Full article ">Figure 8
<p>An illustration of the effect of adding HAND on the accuracy of river flow estimation. (<b>a</b>) Flow estimation using only SAR data and (<b>b</b>) flow estimation using SAR data plus HAND.</p> Full article ">Figure 9
<p>Illustration of some examples of overestimated or underestimated flow predictions.</p> Full article ">Figure 10
<p>Flow prediction using data that are obtained within a two-km buffer around the hydrometric station.</p> Full article ">Figure 11
<p>Flow estimation using data passing through the hydrometric station.</p> Full article ">Figure 12
<p>Occurrence of NSE in the 15 hydrometric stations.</p> Full article ">Figure 13
<p>Analysis of station-by-station flow estimation on different dates and under different flow conditions. River flow varies from 33 to 650 m<sup>3</sup>/s.</p> Full article ">Figure 14
<p>Analysis of station-by-station flow estimation on different dates and under different flow conditions. River flow varies from 12 to 22 m<sup>3</sup>/s.</p> Full article ">
20 pages, 8250 KiB  
Article
A Geometry-Compensated Sensitivity Study of Polarimetric Bistatic Scattering for Rough Surface Observation
by Yanting Wang and Thomas L. Ainsworth
Remote Sens. 2024, 16(10), 1807; https://doi.org/10.3390/rs16101807 - 20 May 2024
Viewed by 386
Abstract
The use of bistatic polarimetric SAR for rough surface observation has attracted increasing interest in recent years, with its acquisition of additional polarimetric information. In this paper, we investigate the sensitivity of polarimetric variables to soil moisture and surface roughness, with the intention [...] Read more.
The use of bistatic polarimetric SAR for rough surface observation has attracted increasing interest in recent years, with its acquisition of additional polarimetric information. In this paper, we investigate the sensitivity of polarimetric variables to soil moisture and surface roughness, with the intention of locating favorable bistatic geometries for soil moisture retrieval. However, in the bistatic setting, the expanded imaging geometry is convolved with the polarimetric scattering response along with the in-scene variations in the soil moisture and surface roughness. The probing polarization states continuously evolve with the bistatic geometry, incurring varying polarization orientation angles. In this investigation, we propose to first compensate the bistatic polarimetric observations for the geometry-induced polarization rotation. Simulations based on a two-scale rough surface scattering model are then used to evaluate the optimal imaging geometry for the best sensitivity to the soil moisture content. We show the different sensing geometries associated with a full list of common polarimetric variables, as we seek favorable bistatic geometries in non-specular directions. The influences of both surface roughness scales are evaluated, with the small-scale roughness parameter imposing the greatest limitation on our results. Full article
(This article belongs to the Special Issue Advances in Synthetic Aperture Radar Image Processing and Information Extraction)
►▼ Show Figures

Figure 1

Figure 1
<p>A bistatic imaging geometry for a horizontal surface with its normal vector pointing to <math display="inline"><semantics> <mrow> <mover accent="true"> <mrow> <mi>n</mi> </mrow> <mo stretchy="false">^</mo> </mover> </mrow> </semantics></math>. (<b>a</b>) The native reference system; (<b>b</b>) the redefined standard reference system by rotating the bistatic plane to X-Z. The bistatic plane is referred to the plane formed by <math display="inline"><semantics> <mrow> <msub> <mrow> <mover accent="true"> <mrow> <mi>k</mi> </mrow> <mo stretchy="false">^</mo> </mover> </mrow> <mrow> <mi>i</mi> </mrow> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mrow> <mover accent="true"> <mrow> <mi>k</mi> </mrow> <mo stretchy="false">^</mo> </mover> </mrow> <mrow> <mi>s</mi> </mrow> </msub> </mrow> </semantics></math>.</p> Full article ">Figure 2
<p>A top view of the scattering response in the full upper scattering hemisphere for a horizontal rough surface (<math display="inline"><semantics> <mrow> <msub> <mrow> <mi>ε</mi> </mrow> <mrow> <mi>r</mi> </mrow> </msub> <mo>=</mo> <mn>9.2</mn> <mo>−</mo> <mi>i</mi> <mn>0.5</mn> <mo>,</mo> </mrow> </semantics></math> the rms height equals <math display="inline"><semantics> <mrow> <mn>0.1</mn> <mi>λ</mi> </mrow> </semantics></math>, and the correlation length equals <math display="inline"><semantics> <mrow> <mn>0.5</mn> <mi>λ</mi> </mrow> </semantics></math>). The incidence direction is marked as ‘+’ and the specular direction is marked as ‘x’.</p> Full article ">Figure 3
<p>Same as <a href="#remotesensing-16-01807-f002" class="html-fig">Figure 2</a> except that the bistatic scattering response has been transformed into the standard reference frame that is defined based on the bistatic plane.</p> Full article ">Figure 4
<p>A top view of the changes in the normalized scattering intensities from low SMC (15%) to high SMC (25%) employing the same baseline roughness setting. The incident wave comes from direction (<math display="inline"><semantics> <mrow> <msup> <mrow> <mn>30</mn> </mrow> <mrow> <mo>∘</mo> </mrow> </msup> </mrow> </semantics></math>,<math display="inline"><semantics> <mrow> <msup> <mrow> <mn>180</mn> </mrow> <mrow> <mo>∘</mo> </mrow> </msup> </mrow> </semantics></math>), whereas the scattering direction varies in the full upper hemisphere.</p> Full article ">Figure 5
<p>A top view of the changes in the normalized scattering intensities from a long correlation length (<math display="inline"><semantics> <mrow> <mn>0.5</mn> <mi>λ</mi> </mrow> </semantics></math>) to a short correlation length (<math display="inline"><semantics> <mrow> <mn>0.1</mn> <mi>λ</mi> </mrow> </semantics></math>) using the same baseline moisture and orientation setting.</p> Full article ">Figure 6
<p>A top view of the quality factor <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>L</mi> </mrow> </msub> </mrow> </semantics></math> derived from the normalized scattering intensities shown in <a href="#remotesensing-16-01807-f004" class="html-fig">Figure 4</a> and <a href="#remotesensing-16-01807-f005" class="html-fig">Figure 5</a>.</p> Full article ">Figure 7
<p>A top view of the changes in the normalized scattering intensities from a high orientation concentration (<math display="inline"><semantics> <mrow> <mi>κ</mi> <mo>=</mo> <mn>50</mn> </mrow> </semantics></math>) to a low orientation concentration (<math display="inline"><semantics> <mrow> <mi>κ</mi> <mo>=</mo> <mn>30</mn> </mrow> </semantics></math>) under the same baseline moisture and surface roughness spectrum.</p> Full article ">Figure 8
<p>A top view of the quality factor <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>κ</mi> </mrow> </msub> </mrow> </semantics></math> derived from the normalized scattering intensities shown in <a href="#remotesensing-16-01807-f004" class="html-fig">Figure 4</a> and <a href="#remotesensing-16-01807-f007" class="html-fig">Figure 7</a>.</p> Full article ">Figure 9
<p>A top view of the changes in the scattering intensity ratios from low SMC (15%) to high SMC (25%) employing the same baseline roughness setting.</p> Full article ">Figure 10
<p>A top view of the changes in the scattering intensity ratios from a long correlation length (<math display="inline"><semantics> <mrow> <mn>0.5</mn> <mi>λ</mi> </mrow> </semantics></math>) to a short correlation length (<math display="inline"><semantics> <mrow> <mn>0.1</mn> <mi>λ</mi> </mrow> </semantics></math>) using the same baseline moisture and orientation setting.</p> Full article ">Figure 11
<p>A top view of the quality factor <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>L</mi> </mrow> </msub> </mrow> </semantics></math> derived from the scattering intensity ratios shown in <a href="#remotesensing-16-01807-f009" class="html-fig">Figure 9</a> and <a href="#remotesensing-16-01807-f010" class="html-fig">Figure 10</a>.</p> Full article ">Figure 12
<p>A top view of the changes in the scattering intensity ratios from a high orientation concentration (<math display="inline"><semantics> <mrow> <mi>κ</mi> <mo>=</mo> <mn>50</mn> </mrow> </semantics></math>) to a low orientation concentration (<math display="inline"><semantics> <mrow> <mi>κ</mi> <mo>=</mo> <mn>30</mn> </mrow> </semantics></math>) under the same baseline moisture and surface roughness spectrum.</p> Full article ">Figure 13
<p>A top view of the quality factor <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>κ</mi> </mrow> </msub> </mrow> </semantics></math> derived from the scattering intensity ratios shown in <a href="#remotesensing-16-01807-f009" class="html-fig">Figure 9</a> and <a href="#remotesensing-16-01807-f012" class="html-fig">Figure 12</a>.</p> Full article ">Figure 14
<p>A top view of the quality factor <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>L</mi> </mrow> </msub> </mrow> </semantics></math> derived for the eigenvalues.</p> Full article ">Figure 15
<p>A top view of the quality factor <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>κ</mi> </mrow> </msub> </mrow> </semantics></math> derived for the eigenvalues.</p> Full article ">Figure 16
<p>A top view of the changes in the characteristic POAs <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>τ</mi> </mrow> <mrow> <mi>i</mi> </mrow> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>τ</mi> </mrow> <mrow> <mi>s</mi> </mrow> </msub> </mrow> </semantics></math> from low SMC (15%) to high SMC (25%) under the same baseline roughness setting, shown in (<b>a</b>,<b>b</b>); the associated quality factors <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>L</mi> </mrow> </msub> </mrow> </semantics></math>, shown in (<b>c</b>,<b>d</b>); and the associated quality factors <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>κ</mi> </mrow> </msub> </mrow> </semantics></math>, shown in (<b>e</b>,<b>f</b>).</p> Full article ">Figure 17
<p>A top view of the changes in the characteristic scattering ratio from low SMC (15%) to high SMC (25%), shown in (<b>a</b>); the changes in the characteristic scattering ratio from a long correlation length (<math display="inline"><semantics> <mrow> <mn>0.5</mn> <mi>λ</mi> </mrow> </semantics></math>) to a short correlation length (<math display="inline"><semantics> <mrow> <mn>0.1</mn> <mi>λ</mi> </mrow> </semantics></math>), shown in (<b>b</b>), and the associated quality factor <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>L</mi> </mrow> </msub> </mrow> </semantics></math>, shown in (<b>c</b>); the changes in the scattering intensity ratios from a high orientation concentration (<math display="inline"><semantics> <mrow> <mi>κ</mi> <mo>=</mo> <mn>50</mn> </mrow> </semantics></math>) to a low orientation concentration (<math display="inline"><semantics> <mrow> <mi>κ</mi> <mo>=</mo> <mn>30</mn> </mrow> </semantics></math>), shown in (<b>d</b>), and the associated quality factor <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>I</mi> </mrow> <mrow> <mi>S</mi> <mi>M</mi> <mo>−</mo> <mi>κ</mi> </mrow> </msub> </mrow> </semantics></math>, shown in (<b>e</b>).</p> Full article ">
14 pages, 13233 KiB  
Communication
Radiometric Calibration of the Near-Infrared Bands of GF-5-02/DPC for Water Vapor Retrieval
by Yanqing Xie, Qingyu Zhu, Sifeng Zhu, Weizhen Hou, Liguo Zhang, Xuefeng Lei, Miaomiao Zhang, Yunduan Li, Zhenhai Liu, Yuan Wen and Zhengqiang Li
Remote Sens. 2024, 16(10), 1806; https://doi.org/10.3390/rs16101806 - 20 May 2024
Viewed by 420
Abstract
The GaoFen (GF)-5-02 satellite is one of the new generations of hyperspectral observation satellites launched by China in 2021. The directional polarimetric camera (DPC) is an optical sensor onboard the GF-5-02 satellite. The precipitable water vapor (PWV) is a key detection parameter of [...] Read more.
The GaoFen (GF)-5-02 satellite is one of the new generations of hyperspectral observation satellites launched by China in 2021. The directional polarimetric camera (DPC) is an optical sensor onboard the GF-5-02 satellite. The precipitable water vapor (PWV) is a key detection parameter of DPC. However, the existing PWV data developed using DPC data have significant errors due to the lack of the timely calibration of the two bands (865, 910 nm) of DPC used for PWV retrieval. In order to acquire DPC PWV data with smaller errors, a calibration method is developed for these two bands. The method consists of two parts: (1) calibrate the 865 nm band of the DPC using the cross-calibration method, (2) calibrate the 910 nm band of the DPC according to the calibrated 865 nm band of the DPC. This method effectively addresses the issue of the absence of a calibration method for the water vapor absorption band (910 nm) of the DPC. Regardless of whether AERONET PWV data or SuomiNET PWV data are used as the reference data, the accuracy of the DPC PWV data developed using calibrated DPC data is significantly superior to that of the DPC PWV data retrieved using data before recalibration. This means that the calibration method for the NIR bands of the DPC can effectively enhance the quality of DPC PWV data. Full article
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>The SRFs of the 865 nm bands of the DPC and POSP and the 910 nm band of the DPC.</p> Full article ">Figure 2
<p>Geographical distribution of AERONET and SuomiNET sites. Each solid blue dot represents an AERONET site, and each solid red dot represents a SuomiNET site.</p> Full article ">Figure 3
<p>The spectral reflectance data of typical ground objects. Each curve in subfigure (<b>a</b>) represents the spectral reflectance of bare soil at different wavelengths; each curve in subfigure (<b>b</b>) represents the spectral reflectance of vegetation at different wavelengths; and each curve in subfigure (<b>c</b>) represents the spectral reflectance of a rock at different wavelengths.</p> Full article ">Figure 3 Cont.
<p>The spectral reflectance data of typical ground objects. Each curve in subfigure (<b>a</b>) represents the spectral reflectance of bare soil at different wavelengths; each curve in subfigure (<b>b</b>) represents the spectral reflectance of vegetation at different wavelengths; and each curve in subfigure (<b>c</b>) represents the spectral reflectance of a rock at different wavelengths.</p> Full article ">Figure 4
<p>Matching results between the actual TOA reflectance data of the DPC at the NIR08 band and the TOA reflectance data of the POSP at the NIR08 band. (<b>a</b>) Actual TOA reflectance data of the POSP at the NIR08 band. (<b>b</b>) TOA reflectance data of the POSP at the NIR08 band after adding 1% random error. (<b>c</b>) TOA reflectance data of the POSP at the NIR08 band after adding 3% random error. (<b>d</b>) TOA reflectance data of the POSP at the NIR08 band after adding 5% random error.</p> Full article ">Figure 5
<p>The distribution of the values of <math display="inline"><semantics> <mrow> <msubsup> <mrow> <mi>R</mi> </mrow> <mrow> <mn>865</mn> <mi>n</mi> <mi>m</mi> </mrow> <mrow> <mo>′</mo> </mrow> </msubsup> <mo>/</mo> <msub> <mrow> <mi>R</mi> </mrow> <mrow> <mn>865</mn> <mi>n</mi> <mi>m</mi> </mrow> </msub> </mrow> </semantics></math>. The black line represents the normal distribution function corresponding to <math display="inline"><semantics> <mrow> <msubsup> <mrow> <mi>R</mi> </mrow> <mrow> <mn>865</mn> <mi>n</mi> <mi>m</mi> </mrow> <mrow> <mo>′</mo> </mrow> </msubsup> <mo>/</mo> <msub> <mrow> <mi>R</mi> </mrow> <mrow> <mn>865</mn> <mi>n</mi> <mi>m</mi> </mrow> </msub> </mrow> </semantics></math>.</p> Full article ">Figure 6
<p>Matching results between the simulated and actual TOA reflectance data of the 910 nm band of the DPC.</p> Full article ">Figure 7
<p>The distribution of the values of <math display="inline"><semantics> <mrow> <msubsup> <mrow> <mi>R</mi> </mrow> <mrow> <mn>910</mn> <mi>n</mi> <mi>m</mi> </mrow> <mrow> <mo>′</mo> </mrow> </msubsup> <mo>/</mo> <msub> <mrow> <mi>R</mi> </mrow> <mrow> <mn>910</mn> <mi>n</mi> <mi>m</mi> </mrow> </msub> </mrow> </semantics></math>. The black line represents the normal distribution function corresponding to <math display="inline"><semantics> <mrow> <msubsup> <mrow> <mi>R</mi> </mrow> <mrow> <mn>910</mn> <mi>n</mi> <mi>m</mi> </mrow> <mrow> <mo>′</mo> </mrow> </msubsup> <mo>/</mo> <msub> <mrow> <mi>R</mi> </mrow> <mrow> <mn>910</mn> <mi>n</mi> <mi>m</mi> </mrow> </msub> </mrow> </semantics></math>.</p> Full article ">Figure 8
<p>Validation results of PWV data retrieved using DPC data before and after radiometric calibration. (<b>a</b>) Validation results of data M1 based on AERONET data. (<b>b</b>) Validation results of data M2 based on AERONET data. (<b>c</b>) Validation results of data M1 based on SuomiNET data. (<b>d</b>) Validation results of data M2 based on SuomiNET data. (<b>e</b>) Validation results of data M1 based on both SuomiNET and AERONET data. (<b>f</b>) Validation results of data M2 based on both SuomiNET and AERONET data.</p> Full article ">Figure 8 Cont.
<p>Validation results of PWV data retrieved using DPC data before and after radiometric calibration. (<b>a</b>) Validation results of data M1 based on AERONET data. (<b>b</b>) Validation results of data M2 based on AERONET data. (<b>c</b>) Validation results of data M1 based on SuomiNET data. (<b>d</b>) Validation results of data M2 based on SuomiNET data. (<b>e</b>) Validation results of data M1 based on both SuomiNET and AERONET data. (<b>f</b>) Validation results of data M2 based on both SuomiNET and AERONET data.</p> Full article ">Figure 9
<p>The distribution of the absolute error of DPC PWV data. (<b>a</b>) The distribution of the absolute error of data M1. (<b>b</b>) The distribution of the absolute error of data M2. The dashed line is the linear regression line. The red points are the means for specific ranges of absolute error, and the red lines are the mean ± 2σ (standard deviation) of absolute error in a certain range.</p> Full article ">
38 pages, 9416 KiB  
Review
Remote Sensing of Forests in Bavaria: A Review
by Kjirsten Coleman, Jörg Müller and Claudia Kuenzer
Remote Sens. 2024, 16(10), 1805; https://doi.org/10.3390/rs16101805 - 20 May 2024
Viewed by 778
Abstract
In recent decades, climatic pressures have altered the forested landscape of Bavaria. Widespread loss of trees has unevenly impacted the entire state, of which 37% is covered by forests (5% more than the national average). In 2018 and 2019—due in large part to [...] Read more.
In recent decades, climatic pressures have altered the forested landscape of Bavaria. Widespread loss of trees has unevenly impacted the entire state, of which 37% is covered by forests (5% more than the national average). In 2018 and 2019—due in large part to drought and subsequent insect infestations—more tree-covered areas were lost in Bavaria than in any other German state. Moreover, the annual crown condition survey of Bavaria has revealed a decreasing trend in tree vitality since 1998. We conducted a systematic literature review regarding the remote sensing of forests in Bavaria. In total, 146 scientific articles were published between 2008 and 2023. While 88 studies took place in the Bavarian Forest National Park, only five publications covered the whole of Bavaria. Outside of the national park, the remaining 2.5 million hectares of forest in Bavaria are understudied. The most commonly studied topics were related to bark beetle infestations (24 papers); however, few papers focused on the drivers of infestations. The majority of studies utilized airborne data, while publications utilizing spaceborne data focused on multispectral; other data types were under-utilized- particularly thermal, lidar, and hyperspectral. We recommend future studies to both spatially broaden investigations to the state or national scale and to increase temporal data acquisitions together with contemporaneous in situ data. Especially in understudied topics regarding forest response to climate, catastrophic disturbances, regrowth and species composition, phenological timing, and in the sector of forest management. The utilization of remote sensing data in the forestry sector and the uptake of scientific results among stakeholders remains a challenge compared to other heavily forested European countries. An integral part of the Bavarian economy and the tourism sector, forests are also vital for climate regulation via atmospheric carbon reduction and land surface cooling. Therefore, forest monitoring remains centrally important to attaining more resilient and productive forests. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Bavaria is located in southeastern Germany, on the border with the Czech Republic and Austria (<b>a</b>). The elevation in Bavaria is highest in the south as low mountain ranges build into the Bavarian Alps (<b>b</b>). Slightly more than half of forests in Bavaria are privately owned (<b>c</b>). Bavaria has many protected nature parks and reserves, national parks, Natura2000 sites and Biosphere reserves (<b>d</b>) Reprinted/adapted with permission from Ref. [<a href="#B6-remotesensing-16-01805" class="html-bibr">6</a>] © Bayerisches Landesamt für Umwelt, <a href="http://www.lfu.bayern.de" target="_blank">www.lfu.bayern.de</a>. Forests cover 37% of Bavaria (10 m 2018 Copernicus Forest Type) (<b>e</b>) [<a href="#B7-remotesensing-16-01805" class="html-bibr">7</a>]. The climate of Bavaria is shifting from cold and dry winters and warm, wet summers, towards warmer, wetter winters and hotter, drier summers (<b>f</b>) (temperatures in red, precipitation in blue) (source: Climate Data Center, Deutscher Wetterdienst, monthly averages 1970–1990, 1991–2023) [<a href="#B2-remotesensing-16-01805" class="html-bibr">2</a>].</p> Full article ">Figure 2
<p>Common research areas for remote sensing of forests. Adapted graphic elements are courtesy of the University of Maryland (Center for Environmental Science, Integration and Application Network) Media Library, CC BY-SA [<a href="#B32-remotesensing-16-01805" class="html-bibr">32</a>].</p> Full article ">Figure 3
<p>Web of Science search methodology. We utilized so-called ‘wildcards’ (* and ?) to allow for spelling variations of search terms.</p> Full article ">Figure 4
<p>Number of publications per year.</p> Full article ">Figure 5
<p>Journals which published two or more articles. Those journals which have published only one article are combined into the category ‘Other’.</p> Full article ">Figure 6
<p>The distribution of first-authorship-affiliated institutes.</p> Full article ">Figure 7
<p>Study locations within Bavaria. Five studies covered the whole of Bavaria while nine studies covered Germany and two studies covered Europe with study sites in Bavaria (not pictured). Two study sites were not specified.</p> Full article ">Figure 8
<p>Area of interest (AOI) and the respective research topics investigated there.</p> Full article ">Figure 9
<p>The trend of AOI sizes throughout the review period.</p> Full article ">Figure 10
<p>The size of the AOI in relation to the data platform.</p> Full article ">Figure 11
<p>Data platform usage in terms of study area (<b>a</b>). The trend in data platform usage throughout the review period (<b>b</b>).</p> Full article ">Figure 12
<p>The distribution of spaceborne (<b>a</b>) sensors and data type and airborne (<b>b</b>) sensors and data type.</p> Full article ">Figure 13
<p>The type of data and platform used in each of the research topics.</p> Full article ">Figure 14
<p>Temporal resolution and data platform.</p> Full article ">Figure 15
<p>Data acquisition month and temporal resolution.</p> Full article ">Figure 16
<p>In situ and RS data acquisition dates by temporal resolution. Those publications which did not utilize any field data, or where the data source was an airborne dataset, have been excluded.</p> Full article ">Figure 17
<p>Forest types investigated over the study period.</p> Full article ">Figure 18
<p>Object scales investigated over the study period.</p> Full article ">Figure 19
<p>Spaceborne data resolution visualized with the object scale investigated.</p> Full article ">Figure 20
<p>Airborne data resolutions distinguished between lidar (<b>a</b>) and optical sensors (<b>b</b>). The resolution is presented with the object scales investigated.</p> Full article ">Figure 21
<p>Resolution trends for airborne lidar (<b>a</b>), optical (<b>b</b>), and spaceborne (<b>c</b>) data.</p> Full article ">Figure 22
<p>Research areas and publication trend.</p> Full article ">Figure 23
<p>Sub-topics (<b>right</b>) presented with the respective overarching research area (<b>left</b>).</p> Full article ">
18 pages, 4549 KiB  
Article
Investigating Dual-Source Satellite Image Data and ALS Data for Estimating Aboveground Biomass
by Wen Fan, Jiaojiao Tian, Thomas Knoke, Bisheng Yang, Fuxun Liang and Zhen Dong
Remote Sens. 2024, 16(10), 1804; https://doi.org/10.3390/rs16101804 - 19 May 2024
Viewed by 466
Abstract
Accurate estimation of above-ground biomass (AGB) in forested areas is essential for studying forest ecological functions, surface carbon cycling, and global carbon balance. Over the past decade, models that harness the distinct features of multi-source remote sensing observations for estimating AGB have gained [...] Read more.
Accurate estimation of above-ground biomass (AGB) in forested areas is essential for studying forest ecological functions, surface carbon cycling, and global carbon balance. Over the past decade, models that harness the distinct features of multi-source remote sensing observations for estimating AGB have gained significant popularity. It is worth exploring the differences in model performance by using simple and fused data. Additionally, quantitative estimation of the impact of high-cost laser point clouds on satellite imagery of varying costs remains largely unexplored. To address these challenges, model performance and cost must be considered comprehensively. We propose a comprehensive assessment based on three perspectives (i.e., performance, potential and limitations) for four typical AGB-estimation models. First, different variables are extracted from the multi-source and multi-resolution data. Subsequently, the performance of four regression methods is tested for AGB estimation with diverse indicator combinations. Experimental results prove that the combination of multi-source data provides a highly accurate AGB regression model. The proposed regression and variables rating approaches can flexibly integrate other data sources for modeling. Furthermore, the data cost is discussed against the AGB model performance. Our study demonstrates the potential of using low-cost satellite data to provide a rough AGB estimation for larger areas, which can allow different remote sensing data to meet different needs of forest management decisions. Full article
(This article belongs to the Special Issue Remote Sensing for Forest Morphological and Physiological Traits Monitoring)
►▼ Show Figures

Figure 1

Figure 1
<p>The workflow of the proposed method.</p> Full article ">Figure 2
<p>The study areas in Guangxi Province, China.</p> Full article ">Figure 3
<p>Distribution of sample sites and TLS collection locations: (<b>a</b>) Masson pine forest in Guigang municipality; (<b>b</b>) Masson pine forest in Laibin municipality; (<b>c</b>) Eucalyptus forest in Laibin municipality; and (<b>d</b>) Masson pine forest in Qinzhou municipality.</p> Full article ">Figure 4
<p>Scatterplots of the predicted and estimated AGB for the four fitting methods.</p> Full article ">Figure 5
<p>Scatterplots of the residuals and predicted AGB for the four fitting methods.</p> Full article ">Figure 6
<p>Scatterplot of <math display="inline"><semantics> <msup> <mi>R</mi> <mn>2</mn> </msup> </semantics></math> and cost. Cost results are for the multispectral products available from the Apollo Mapping archives (Standard Tasking) [<a href="#B28-remotesensing-16-01804" class="html-bibr">28</a>].</p> Full article ">Figure A1
<p>Satellite imagery variables selection. Importance ranking of the variables in (<b>a</b>) the GF2 model and (<b>c</b>) the LS8 model. Correlation analysis for (<b>b</b>) the GF2 model and (<b>d</b>) the LS8 model. Dark green and red indicate highly negative and positive correlations, respectively.</p> Full article ">Figure A2
<p>ALS variables selection. (<b>a</b>) Importance ranking of the variables in the ALS model. (<b>b</b>) Correlation analysis for the ALS model. Dark green and red indicate highly negative and positive correlations, respectively.</p> Full article ">Figure A3
<p>Fused data variables selection. (<b>c</b>) Importance ranking of variables in (<b>a</b>) the GF2-ALS model and (<b>c</b>) the LS8-ALS model. Correlation analysis for (<b>b</b>) the GF2-ALS model and (<b>d</b>) the LS8-ALS model. Dark green and red indicate highly negative positive correlations, respectively.</p> Full article ">
21 pages, 3243 KiB  
Article
Examining the Impacts of Pre-Fire Forest Conditions on Burn Severity Using Multiple Remote Sensing Platforms
by Kangsan Lee, Willem J. D. van Leeuwen, Jeffrey K. Gillan and Donald A. Falk
Remote Sens. 2024, 16(10), 1803; https://doi.org/10.3390/rs16101803 - 19 May 2024
Viewed by 603
Abstract
Pre-fire environmental conditions play a critical role in wildfire severity. This study investigated the impact of pre-fire forest conditions on burn severity as a result of the 2020 Bighorn Fire in the Santa Catalina Mountains in Arizona. Using a stepwise regression model and [...] Read more.
Pre-fire environmental conditions play a critical role in wildfire severity. This study investigated the impact of pre-fire forest conditions on burn severity as a result of the 2020 Bighorn Fire in the Santa Catalina Mountains in Arizona. Using a stepwise regression model and remotely sensed data from Landsat 8 and LiDAR, we analyzed the effects of structural and functional vegetation traits and environmental factors on burn severity. This analysis revealed that the difference normalized burn ratio (dNBR) was a more reliable indicator of burn severity compared to the relative dNBR (RdNBR). Stepwise regression identified pre-fire normalized difference vegetation index (NDVI), canopy cover, and tree density as significant variables across all land cover types that explained burn severity, suggesting that denser areas with higher vegetation greenness experienced more severe burns. Interestingly, residuals between the actual and estimated dNBR were lower in herbaceous zones compared to denser forested areas at similar elevations, suggesting potentially more predictable burn severity in open areas. Spatial analysis using Geary’s C statistics further revealed a strong negative autocorrelation: areas with high burn severity tended to be clustered, with lower severity areas interspersed. Overall, this study demonstrates the potential of readily available remote sensing data to predict potential burn severity values before a fire event, providing valuable information for forest managers to develop strategies for mitigating future wildfire damage. Full article
(This article belongs to the Special Issue Vegetation Mapping through Multiscale Remote Sensing)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Boundaries of significant fires since 2002 in the Santa Catalina Mountains north of Tucson, AZ, USA. The area of the 2020 Bighorn fire (red polygon) encompassed most areas that had already experienced burning within the last two decades. The study area (black outline) is located on the south side of the mountain.</p> Full article ">Figure 2
<p>Structural and functional traits of the forest and topographic variables were extracted from multiple data sources (marked with blue circles) including aerial LiDAR point cloud and multispectral Landsat 8 satellite spectral surface reflectance data.</p> Full article ">Figure 3
<p><b>dNBR and RdNBR for the Bighorn Fire for the Santa Catalina Mountains (2020).</b> The values of dNBR and RdNBR were multiplied by 1000 for visualization purposes. A total of 138,312 points were extracted from a 30-m grid derived from Landsat 8 images. Negative values indicate regrowth following the fire, while positive values indicate a higher degree of burn severity.</p> Full article ">Figure 4
<p><b>Correlation of all structural and functional variables.</b> Correlations among the topographic and pre-fire vegetation explanatory variables and the two burn severity indices (RdNBR and dNBR), where the dNBR represents a higher correlation coefficient overall.</p> Full article ">Figure 5
<p><b>Vegetation classification map, classified by Planet Scope and USGS LiDAR using classification and regression tree (CART), prior to the Bighorn Fire (overall accuracy 0.88).</b> This indicates a variety of vegetation types across the study area. Lower elevations are characterized by sparse vegetation, whereas as the elevation increases, a richer diversity of vegetation species can be observed.</p> Full article ">Figure 6
<p><b>Distributions of residuals of all classes.</b> It is pertinent to acknowledge that the three land cover types under consideration exhibited unequal population sizes. The live trees/shrubs class demonstrated the smallest residual errors in contrast to the remaining two vegetation types.</p> Full article ">Figure 7
<p><b>Residual values representing the difference between the predicted and observed burn severity index values.</b> Larger values were observed in higher elevation areas, where more severe burn severity was observed. Smaller differences were observed between the predicted and observed dNBR values in the lower elevation area.</p> Full article ">
25 pages, 18829 KiB  
Article
Enhanced Underwater Single Vector-Acoustic DOA Estimation via Linear Matched Stochastic Resonance Preprocessing
by Haitao Dong, Jian Suo, Zhigang Zhu, Haiyan Wang and Hongbing Ji
Remote Sens. 2024, 16(10), 1802; https://doi.org/10.3390/rs16101802 - 18 May 2024
Viewed by 559
Abstract
Underwater acoustic vector sensors (UAVSs) are increasingly utilized for remote passive sonar detection, but the accuracy of direction-of-arrival (DOA) estimation remains a challenging problem, particularly under low signal-to-noise ratio (SNR) conditions and complex background noise. In this paper, a comprehensive theoretical analysis is [...] Read more.
Underwater acoustic vector sensors (UAVSs) are increasingly utilized for remote passive sonar detection, but the accuracy of direction-of-arrival (DOA) estimation remains a challenging problem, particularly under low signal-to-noise ratio (SNR) conditions and complex background noise. In this paper, a comprehensive theoretical analysis is conducted on UAVS signal preprocessing subjected to gain-phase uncertainties for average acoustic intensity measurement (AAIM) and complex acoustic intensity measurement (CAIM)-based vector DOA estimation, aiming to explain the theoretical restrictions of intensity-based vector acoustic preprocessing approaches. On this basis, a generalized vector acoustic preprocessing optimization model is established in which the principle can be described as “maximizing the denoising performance under the constraints of an equivalent amplitude-gain response and phase-bias response”. A novel vector acoustic preprocessing method named linear matched stochastic resonance (LMSR) is proposed within the framework of matched stochastic resonance theory, which can naturally guarantee the linear gain-phase restrictions, as well achieving effective denoising performance. Numerical analyses demonstrate the superior vector DOA estimation performance of our proposed LMSR-AAIM and LMSR-CAIM methods in comparison to classical intensity-based AAIM and CAIM methods, especially under low-SNR conditions and non-Gaussian impulsive noise circumstances. Experimental verification conducted in the South China Sea further verifies its the effectiveness for practical application. This work can lay a solid foundation to break through the challenges of underwater remote vector acoustic DOA estimation under low-SNR conditions and complex ocean ambient noise and can provide important guidance for future research work. Full article
(This article belongs to the Special Issue Space-Air-Ground-Ocean Integrated Sensing and Information Transmission)
►▼ Show Figures

Figure 1

Figure 1
<p>Performance comparison for vector DOA estimation (<b>a</b>) input and output amplitude response of CBSR with varied <span class="html-italic">a</span>; (<b>b</b>) input and output amplitude response of CBSR with varied <span class="html-italic">b</span>.</p> Full article ">Figure 2
<p>A comparison of MSR output SNR responses under different noise intensities (<span class="html-italic">D</span>).</p> Full article ">Figure 3
<p>The framework of LMSR-based vector acoustic DOA estimation method.</p> Full article ">Figure 4
<p>Performance comparison for vector DOA estimation. (<b>a</b>) Time-domain quantization of received noisy signal; (<b>b</b>) frequency-domain quantization of received noisy signal; (<b>c</b>) time-domain quantization of preprocessed signal with LMSR; (<b>d</b>) frequency-domain quantization of preprocessed signal with LMSR.</p> Full article ">Figure 5
<p>Performance comparison for vector acoustic DOA estimation of AAIM, CAIM, LMSR-AAIM, and LMSR-CAIM. (<b>a</b>) Mean error (ME); (<b>b</b>) root-mean-square error (RMSE).</p> Full article ">Figure 6
<p>Simulation of vector acoustic signal under impulsive noise (<math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>1.5</mn> </mrow> </semantics></math>, <math display="inline"><semantics> <mrow> <mi>θ</mi> <mo>=</mo> <msup> <mn>30</mn> <mo>∘</mo> </msup> </mrow> </semantics></math>).</p> Full article ">Figure 7
<p>Nonlinear filtering effect of LMSR under Lévy impulsive noise (<math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>1.5</mn> </mrow> </semantics></math>). (<b>a</b>) Received signal and its normalized FFT; (<b>b</b>) LMSR output signal and its normalized FFT.</p> Full article ">Figure 8
<p>Performance comparison for vector DOA estimation of AAIM, CAIM, LMSR-AAIM, and LMSR-CAIM under Lévy impulsive noise (<math display="inline"><semantics> <mrow> <mi>α</mi> <mo>=</mo> <mn>1.5</mn> </mrow> </semantics></math>). (<b>a</b>) Mean error (ME); (<b>b</b>) root-mean-square error (RMSE).</p> Full article ">Figure 9
<p>Sea experiment description. (<b>a</b>) Experimental ship’s navigation survey line; (<b>b</b>) latent sea buoy with a single UAVS.</p> Full article ">Figure 10
<p>Time-frequency analysis of received vector acoustic signal in 4.5 h. (<b>a</b>) P channel; (<b>b</b>) <math display="inline"><semantics> <msub> <mi mathvariant="normal">V</mi> <mi>x</mi> </msub> </semantics></math> channel; (<b>c</b>) <math display="inline"><semantics> <msub> <mi mathvariant="normal">V</mi> <mi>y</mi> </msub> </semantics></math> channel.</p> Full article ">Figure 11
<p>Estimation performance comparison for the moving ship. (<b>a</b>) AAIM; (<b>b</b>) CAIM; (<b>c</b>) LMSR-CAIM.</p> Full article ">
36 pages, 6112 KiB  
Article
Greenness and Actual Evapotranspiration in the Unrestored Riparian Corridor of the Colorado River Delta in Response to In-Channel Water Deliveries in 2021 and 2022
by Pamela L. Nagler, Ibrahima Sall, Martha Gomez-Sapiens, Armando Barreto-Muñoz, Christopher J. Jarchow, Karl Flessa and Kamel Didan
Remote Sens. 2024, 16(10), 1801; https://doi.org/10.3390/rs16101801 - 18 May 2024
Viewed by 767
Abstract
Natural resource managers may utilize remotely sensed data to monitor vegetation within their decision-making frameworks for improving habitats. Under binational agreements between the United States and Mexico, seven reaches were targeted for riparian habitat enhancement. Monitoring was carried out using Landsat 8 16-day [...] Read more.
Natural resource managers may utilize remotely sensed data to monitor vegetation within their decision-making frameworks for improving habitats. Under binational agreements between the United States and Mexico, seven reaches were targeted for riparian habitat enhancement. Monitoring was carried out using Landsat 8 16-day intervals of the two-band enhanced vegetation index 2 (EVI2) for greenness and actual evapotranspiration (ETa). In-channel water was delivered in 2021 and 2022 at four places in Reach 4. Three reaches (Reaches 4, 5 and 7) showed no discernable difference in EVI2 from reaches that did not receive in-channel water (Reaches 1, 2, 3 and 6). EVI2 in 2021 was higher than 2020 in all reaches except Reach 3, and EVI2 in 2022 was lower than 2021 in all reaches except Reach 7. ET(EVI2) was higher in 2020 than in 2021 and 2022 in all seven reaches; it was highest in Reach 4 (containing restoration sites) in all years. Excluding restoration sites, compared with 2020, unrestored reaches showed that EVI2 minimally increased in 2021 and 2022, while ET(EVI2) minimally decreased despite added water in 2021–2022. Difference maps comparing 2020 (no-flow year) to 2021 and 2022 (in-channel flows) reveal areas in Reaches 5 and 7 where the in-channel flows increased greenness and ET(EVI2). Full article
(This article belongs to the Section Biogeosciences Remote Sensing)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Colorado River and Delta depicting Reaches 1–7 as defined under Minute 319 and four water delivery sites used during the 2021 and 2022 in-channel water deliveries. The water delivery sites from north to south are Chausse, Km 18, Km 21, and Cori. The Yuma Valley AZMET [<a href="#B70-remotesensing-16-01801" class="html-bibr">70</a>] station is not shown; it is located north of the Northerly International Border (NIB) in Yuma, Arizona.</p> Full article ">Figure 2
<p>Peak growing season (1 May to 30 October) EVI2 (greenness) from Landsat 8 OLI imagery (30 m/98 ft resolution) for years 2014–2022 for the riparian corridor by river reach (Reach 1–7 includes restored areas in Reaches 2 and 4) and the weighted average by area of these seven reaches for all reaches (all) ((<b>a</b>) top bar plot) and the detrended EVI2 data ((<b>b</b>) bottom bar plot). Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 3
<p>Monthly variation in EVI2 (greenness) from Landsat 8 OLI (30 m/98 ft resolution) in Reach 4 (blue line) (excluding restorations sites) and in the unrestored reaches, (5, 6, and 7, green, red, and yellow lines, respectively), and the average of the unrestored Reaches 4–7 (dashed black line) for years 2014–2022. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 4
<p>Peak growing season (1 May to 30 October) ET(EVI2) (mm/day) from Landsat 8 OLI imagery (30 m/98 ft resolution) for years 2014–2022 for the riparian corridor for the seven Colorado River Delta reaches and the average of all reaches (all) with ET(EVI2) calculated with ETo calculated from AZMET [<a href="#B70-remotesensing-16-01801" class="html-bibr">70</a>] ((<b>a</b>) top bar plot) and detrended ET(EVI2) data ((<b>b</b>) bottom bar plot). Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 5
<p>Monthly variation in ET(EVI2) (mm/day) in unrestored reaches for years 2014–2022. The data correspond with Reach 4, which excludes restoration sites (blue line), Reaches 5, 6, and 7 (green, red, and yellow lines, respectively) and the average of Reaches 4–7 (dashed black line). Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 6
<p>(<b>a</b>) Peak growing season (1 May to 30 October) EVI2 change over one year (2021–2020) in Reaches 4–7 of the Colorado River Delta. Change maps show differences between 2021 (first in-channel flow year) and 2020 (non-flow year). Boxes show histograms based on the frequency distribution of pixels demonstrating EVI2 change in the reaches that received in-channel water deliveries (Reach 4, 5 and 7) and values less than zero indicate a decrease in EVI2. (<b>b</b>) Peak growing season (1 May to 30 October) EVI2 change over two years (2021–2019) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>c</b>) Peak growing season (1 May to 30 October) EVI2 change over two years (2022–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>d</b>) Peak growing season (1 May to 30 October) EVI2 change over one year (2022–2021) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 6 Cont.
<p>(<b>a</b>) Peak growing season (1 May to 30 October) EVI2 change over one year (2021–2020) in Reaches 4–7 of the Colorado River Delta. Change maps show differences between 2021 (first in-channel flow year) and 2020 (non-flow year). Boxes show histograms based on the frequency distribution of pixels demonstrating EVI2 change in the reaches that received in-channel water deliveries (Reach 4, 5 and 7) and values less than zero indicate a decrease in EVI2. (<b>b</b>) Peak growing season (1 May to 30 October) EVI2 change over two years (2021–2019) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>c</b>) Peak growing season (1 May to 30 October) EVI2 change over two years (2022–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>d</b>) Peak growing season (1 May to 30 October) EVI2 change over one year (2022–2021) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 6 Cont.
<p>(<b>a</b>) Peak growing season (1 May to 30 October) EVI2 change over one year (2021–2020) in Reaches 4–7 of the Colorado River Delta. Change maps show differences between 2021 (first in-channel flow year) and 2020 (non-flow year). Boxes show histograms based on the frequency distribution of pixels demonstrating EVI2 change in the reaches that received in-channel water deliveries (Reach 4, 5 and 7) and values less than zero indicate a decrease in EVI2. (<b>b</b>) Peak growing season (1 May to 30 October) EVI2 change over two years (2021–2019) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>c</b>) Peak growing season (1 May to 30 October) EVI2 change over two years (2022–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>d</b>) Peak growing season (1 May to 30 October) EVI2 change over one year (2022–2021) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 6 Cont.
<p>(<b>a</b>) Peak growing season (1 May to 30 October) EVI2 change over one year (2021–2020) in Reaches 4–7 of the Colorado River Delta. Change maps show differences between 2021 (first in-channel flow year) and 2020 (non-flow year). Boxes show histograms based on the frequency distribution of pixels demonstrating EVI2 change in the reaches that received in-channel water deliveries (Reach 4, 5 and 7) and values less than zero indicate a decrease in EVI2. (<b>b</b>) Peak growing season (1 May to 30 October) EVI2 change over two years (2021–2019) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>c</b>) Peak growing season (1 May to 30 October) EVI2 change over two years (2022–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>d</b>) Peak growing season (1 May to 30 October) EVI2 change over one year (2022–2021) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 7
<p>(<b>a</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over one year (2021–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>b</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over two years (2021–2019) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>c</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over two years (2022–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>d</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over one year (2022–2021) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 7 Cont.
<p>(<b>a</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over one year (2021–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>b</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over two years (2021–2019) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>c</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over two years (2022–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>d</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over one year (2022–2021) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 7 Cont.
<p>(<b>a</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over one year (2021–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>b</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over two years (2021–2019) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>c</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over two years (2022–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>d</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over one year (2022–2021) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 7 Cont.
<p>(<b>a</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over one year (2021–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>b</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over two years (2021–2019) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>c</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over two years (2022–2020) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. (<b>d</b>) Peak growing season (1 May to 30 October) ET(EVI2) change over one year (2022–2021) in Reaches 4–7 of the Colorado River Delta. Note: map legend and descriptions are the same as in the heading of <a href="#remotesensing-16-01801-f006" class="html-fig">Figure 6</a>a. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">Figure 8
<p>Nine years (2014–2022) of weighted average vegetation greenness (EVI2) and water use (Nagler ETa) for both restored sites in Reaches 2 and 4 and unrestored reaches (Reaches 1–7) in the Colorado River Delta. Data generated during this study are published and available [<a href="#B26-remotesensing-16-01801" class="html-bibr">26</a>].</p> Full article ">
22 pages, 6272 KiB  
Article
Modeling and Locating the Wind Erosion at the Dry Bottom of the Aral Sea Based on an InSAR Temporal Decorrelation Decomposition Model
by Yubin Song, Xuelian Xun, Hongwei Zheng, Xi Chen, Anming Bao, Ying Liu, Geping Luo, Jiaqiang Lei, Wenqiang Xu, Tie Liu, Olaf Hellwich and Qing Guan
Remote Sens. 2024, 16(10), 1800; https://doi.org/10.3390/rs16101800 - 18 May 2024
Viewed by 551
Abstract
The dust originating from the extinct lake of the Aral Sea poses a considerable threat to the surrounding communities and ecosystems. The accurate location of these wind erosion areas is an essential prerequisite for controlling sand and dust activity. However, few relevant indicators [...] Read more.
The dust originating from the extinct lake of the Aral Sea poses a considerable threat to the surrounding communities and ecosystems. The accurate location of these wind erosion areas is an essential prerequisite for controlling sand and dust activity. However, few relevant indicators reported in this current study can accurately describe and measure wind erosion intensity. A novel wind erosion intensity (WEI) of a pixel resolution unit was defined in this paper based on deformation due to the wind erosion in this pixel resolution unit. We also derived the relationship between WEI and soil InSAR temporal decorrelation (ITD). ITD is usually caused by the surface change over time, which is very suitable for describing wind erosion. However, within a pixel resolution unit, the ITD signal usually includes soil and vegetation contributions, and extant studies concerning this issue are considerably limited. Therefore, we proposed an ITD decomposition model (ITDDM) to decompose the ITD signal of a pixel resolution unit. The least-square method (LSM) based on singular value decomposition (SVD) is used to estimate the ITD of soil (SITD) within a pixel resolution unit. We verified the results qualitatively by the landscape photos, which can reflect the actual conditions of the soil. At last, the WEI of the Aral Sea from 23 June 2020, to 5 July 2020 was mapped. The results confirmed that (1) based on the ITDDM model, the SITD can be accurately estimated by the LSM; (2) the Aral Sea is experiencing severe wind erosion; and (3) the middle, northeast, and southeast bare areas of the South Aral Sea are where salt dust storms may occur. Full article
(This article belongs to the Special Issue Remote Sensing: 15th Anniversary)
►▼ Show Figures

Figure 1

Figure 1
<p>Study area and the corresponding land types. The 2017 land surface coverage map with a resolution of 10 m is from Tsinghua University (<a href="https://data-starcloud.pcl.ac.cn/zh/" target="_blank">https://data-starcloud.pcl.ac.cn/zh/</a> (accessed on 15 May 2024)).</p> Full article ">Figure 2
<p>Temporal variation trend of average wind speed from 23 June 2020 to 5 July 2020.</p> Full article ">Figure 3
<p>Spatial variation trend of average wind speed from 23 June 2020 to 5 July 2020.</p> Full article ">Figure 4
<p>Wind erosion at 1D soil surface. The blue dotted line represents topography before wind erosion, and the blue solid line represents the topography after wind erosion.</p> Full article ">Figure 5
<p>Wind erosion from the perspective of microwave remote sensing. The displacement of the scatterer in the pixel resolution unit caused by wind erosion can be described by the random variable <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>δ</mi> </mrow> <mrow> <mi>z</mi> </mrow> </msub> </mrow> </semantics></math>. The red dashed lines with arrows represent the wind erosion depth at any position of a pixel resolution unit.</p> Full article ">Figure 6
<p>Relationship between the ITD and WEI.</p> Full article ">Figure 7
<p>The VFC, total MBC, SMBC, VMBC, and SDV of the study area. (<b>a</b>–<b>e</b>) VFC, total MBC, SMBC, VMBC, and SDV, respectively. Due to the widespread presence of artificial forests, the estimated SMBC and VMBC in regions A, B, and C may be erroneous.</p> Full article ">Figure 8
<p>The potential wind erosion regions in the study area. The red line labeled as “AralSea1973” represents the boundary of the Aral Sea in 1973, which was manually delineated using optical remote sensing imagery from 1973 on Google Earth. Considering the presence of artificial forests, the estimated SMBC in regions A–F may be unreliable. Nevertheless, the presence of vegetation effectively mitigates the occurrence of wind erosion within these areas.</p> Full article ">Figure 9
<p>The SITD map of the Aral Sea. The legend in the bottom-right corner of <a href="#remotesensing-16-01800-f009" class="html-fig">Figure 9</a> denotes the areas where SITD calculations are unnecessary. These areas include water bodies, wetlands, regions with vegetation cover greater than 0.4, or areas where soil volumetric moisture content exceeds 0.1. This is because soil erosion is typically not expected to occur in these regions.</p> Full article ">Figure 10
<p>The results of SITD estimation and their corresponding landscape photos of 15 sampling sites.</p> Full article ">Figure 11
<p>WEI of the Aral Sea. When computing WEI, we still excluded regions encompassing water bodies, wetlands, VSM greater than or equal to 0.1, as well as areas with VFC greater than or equal to 0.4 (refer to the legend in the bottom-right corner of <a href="#remotesensing-16-01800-f011" class="html-fig">Figure 11</a>). Subsequently, we categorized WEI into eight classes to enhance the contrasting effect among regions exposed to varying intensities of wind erosion (refer to the legend in the top-left corner of <a href="#remotesensing-16-01800-f011" class="html-fig">Figure 11</a>).</p> Full article ">Figure 12
<p>The estimation process of the VMBC and SMBC for points A, B, and C. The buffer regions M, N, and Q are used to estimate the VMBC and SMBC of pixel points A, B, and C, respectively.</p> Full article ">Figure 13
<p>The aerosol optical thickness over the Aral Sea during the period from 23 June 2020 to 5 July 2020.</p> Full article ">Figure 14
<p>Soil temporal decorrelation with landscape photos of sampling sites in the bare land between the two branches of the South Aral Sea. (<b>a</b>) Soil temporal decorrelation, (<b>b</b>) Landscape photo of L16, (<b>c</b>) Landscape photo of L14, and (<b>d</b>) Landscape photo of L15. The value of soil time decorrelation is between 0 and 1. The smaller the value, the darker the corresponding pixel, and the larger the value, the brighter the corresponding pixel.</p> Full article ">
21 pages, 1775 KiB  
Article
Conditional Diffusion Model for Urban Morphology Prediction
by Tiandong Shi, Ling Zhao, Fanfan Liu, Ming Zhang, Mengyao Li, Chengli Peng and Haifeng Li
Remote Sens. 2024, 16(10), 1799; https://doi.org/10.3390/rs16101799 - 18 May 2024
Viewed by 497
Abstract
Predicting urban morphology based on local attributes is an important issue in urban science research. The deep generative models represented by generative adversarial network (GAN) models have achieved impressive results in this area. However, in such methods, the urban morphology is assumed to [...] Read more.
Predicting urban morphology based on local attributes is an important issue in urban science research. The deep generative models represented by generative adversarial network (GAN) models have achieved impressive results in this area. However, in such methods, the urban morphology is assumed to follow a specific probability distribution and be able to directly approximate the distribution via GAN models, which is not a realistic strategy. As demonstrated by the score-based model, a better strategy is to learn the gradient of the probability distribution and implicitly approximate the distribution. Therefore, in this paper, an urban morphology prediction method based on the conditional diffusion model is proposed. Implementing this approach results in the decomposition of the attribute-based urban morphology prediction task into two subproblems: estimating the gradient of the conditional distribution, and gradient-based sampling. During the training stage, the gradient of the conditional distribution is approximated by using a conditional diffusion model to predict the noise added to the original urban morphology. In the generation stage, the corresponding conditional distribution is parameterized based on the noise predicted by the conditional diffusion model, and the final prediction result is generated through iterative sampling. The experimental results showed that compared with GAN-based methods, our method demonstrated improvements of 5.5%, 5.9%, and 13.2% in the metrics of low-level pixel features, shallow structural features, and deep structural features, respectively. Full article
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Framework of the urban morphology prediction method based on the conditional diffusion model. During the training stage, the optimization objective of the conditional diffusion model is to minimize the distance between the predicted noise and the added noise. During the generation stage, the conditional diffusion model iteratively predicts noise and parameterizes the corresponding conditional distribution. The generation target is updated by sampling from this distribution until the end of the iterations, to obtain the final generated target.</p> Full article ">Figure 2
<p>Visualization of the dataset sample. The urban built-up area images and the water area images are binary images with pixel values of 0 and 1, respectively. Each pixel in the DEM and NTL images represents the average value of the corresponding geographic area, and the pixel value is scaled to 0-1.</p> Full article ">Figure 3
<p>Histograms of the dataset regarding the first feature. The left figure represents the results for the training dataset, and the right figure represents the results for the testing dataset.</p> Full article ">Figure 4
<p>Histograms of the dataset regarding the second feature. The left figure represents the results for the training dataset, and the right figure represents the results for the testing dataset.</p> Full article ">Figure 5
<p>Some generation targets. The color bar is consistent with <a href="#remotesensing-16-01799-f002" class="html-fig">Figure 2</a>. The left figure shows the generation targets obtained by constructing a guidance condition based on the first aggregation method. The right figure shows the generation targets obtained by constructing a guidance condition based on the second aggregation method.</p> Full article ">Figure 6
<p>Fractal dimension - statistics of the generation targets and corresponding labels.</p> Full article ">Figure 7
<p>Bad samples based on the first aggregation method.</p> Full article ">Figure 8
<p>Bad samples based on the second aggregation method.</p> Full article ">Figure 9
<p>Generation targets based on different values of the parameter <span class="html-italic">w</span> under the first aggregation method. The first column shows the corresponding labels.</p> Full article ">Figure 10
<p>Generation targets based on different values of the parameter <span class="html-italic">w</span> under the second aggregation method. The first column shows the corresponding labels.</p> Full article ">Figure 11
<p>Confusion matrix for the prediction results of the two types of fusion method.</p> Full article ">
22 pages, 1843 KiB  
Article
Long-Time Coherent Integration for the Spatial-Based Bistatic Radar Based on Dual-Scale Decomposition and Conditioned CPF
by Suqi Li, Yihan Wang, Yanfeng Liang and Bailu Wang
Remote Sens. 2024, 16(10), 1798; https://doi.org/10.3390/rs16101798 - 18 May 2024
Viewed by 491
Abstract
This paper addresses the problem of weak maneuvering target detection in the space-based bistatic radar system through long-time coherent integration (LTCI). The space-based bistatic radar is vulnerable to the high-order range migration (RM) and Doppler frequency migration (DFM), since the target, the receiver [...] Read more.
This paper addresses the problem of weak maneuvering target detection in the space-based bistatic radar system through long-time coherent integration (LTCI). The space-based bistatic radar is vulnerable to the high-order range migration (RM) and Doppler frequency migration (DFM), since the target, the receiver and the transmitter all can play fast movement independently. To correct high- order RM and DFM, this usually involves joint high-dimensional parameter searching, incurring a large computational burden. In our previous work, a dual-scale (DS) decomposition of motion parameters was proposed, in which the optimal GRFT is conditionally decoupled into two cascade procedures called the modified generalized inverse Fourier transform (GIFT) and generalized Fourier transform (GFT), resulting in the DS-GRFT detector. However, even if the DS-GRFT detector preserves the superior performance and dramatically decreases the complexity, high-dimensional searching is still required. In this paper, by analyzing the structure of the DS-GRFT detector, we further designed a conditioned cubic phase function (CCPF) tailored to the range–slow-time signal after GIFT, breaking the joint high-dimensional searching into independent one-dimensional searching. Then, by connecting the proposed CCPF with the GIFT, we achieved a new LTCI detector called the DS-GIFT-CCPF detector, which obtained a significant computational cost reduction with acceptable performance loss, as demonstrated in numerical experiments. Full article
(This article belongs to the Special Issue Advances in Synthetic Aperture Radar (SAR) Signal and Image Processing)
►▼ Show Figures

Figure 1

Figure 1
<p>Geometric relationship between the spaced-based bistatic radar system and the target.</p> Full article ">Figure 2
<p>Range spectrum with slow time at different stages: (<b>a</b>) before RM correction; (<b>b</b>) after coarse acceleration compensation; (<b>c</b>) after RM compensation with respect to coarse acceleration and coarse velocity; (<b>d</b>) after RM compensation with respect to coarse jerk, coarse acceleration and coarse velocity.</p> Full article ">Figure 3
<p>Range–Doppler spectrum after RM correction.</p> Full article ">Figure 4
<p>Maximum outputs of each range bin after RM and DFM corrections.</p> Full article ">Figure 5
<p>Estimation of target motion parameters: (<b>a</b>) coarse velocity, coarse acceleration and coarse jerk spectrum at range bin 15,361; (<b>b</b>) fine acceleration and fine jerk spectrum with estimated coarse motion parameters of (<b>a</b>); (<b>c</b>) range–Doppler spectrum after target motion compensation with estimated motion parameters of (<b>a</b>,<b>b</b>); (<b>d</b>) coarse velocity, coarse acceleration and coarse jerk spectrum at range bin 15,232; (<b>e</b>) fine acceleration and fine jerk spectrum with estimated coarse motion parameters of (<b>c</b>); (<b>f</b>) range–Doppler spectrum after target motion compensation with estimated motion parameters of (<b>d</b>,<b>e</b>).</p> Full article ">Figure 6
<p>Doppler spectrum for targets 1, 2 and 4 at different stages: (<b>a</b>) after DFM compensation with respect to coarse motion parameters for target 1 at range bin 15,361; (<b>b</b>) after DFM compensation with respect to coarse motion parameters and fine acceleration of target 1 at range bin 15,361; (<b>c</b>) after DFM compensation with respect to coarse motion parameters, fine acceleration and fine jerk of target 1 at range bin 15,361; (<b>d</b>) after DFM compensation with respect to coarse motion parameters of target 2 at range bin 15,361; (<b>e</b>) coarse motion parameters and fine acceleration of target 2 at range bin 15,361; (<b>f</b>) after DFM compensation with respect to coarse motion parameters, fine acceleration and fine jerk of target 2 at range bin 15,361; (<b>g</b>) after DFM compensation with respect to coarse motion parameters of target 4 at range bin 15,232; (<b>h</b>) after DFM compensation with respect to coarse motion parameters and fine acceleration of target 4 at range bin 15,232; (<b>i</b>) after DFM compensation with respect to coarse motion parameters, fine acceleration and fine jerk of target 4 at range bin 15,232.</p> Full article ">Figure 7
<p>Detection probability versus pulse-compressed SNR for the proposed DS-GIFT-CCPF with different sizes of search space of <math display="inline"><semantics> <mo>Ω</mo> </semantics></math>.</p> Full article ">Figure 8
<p>Performance comparison of different detectors: (<b>a</b>) detection probability versus pulse-compressed SNR; (<b>b</b>) computational complexity.</p> Full article ">
19 pages, 5287 KiB  
Article
GGMNet: Pavement-Crack Detection Based on Global Context Awareness and Multi-Scale Fusion
by Yong Wang, Zhenglong He, Xiangqiang Zeng, Juncheng Zeng, Zongxi Cen, Luyang Qiu, Xiaowei Xu and Qunxiong Zhuo
Remote Sens. 2024, 16(10), 1797; https://doi.org/10.3390/rs16101797 - 18 May 2024
Viewed by 459
Abstract
Accurate and comprehensive detection of pavement cracks is important for maintaining road quality and ensuring traffic safety. However, the complexity of road surfaces and the diversity of cracks make it difficult for existing methods to accomplish this challenging task. This paper proposes a [...] Read more.
Accurate and comprehensive detection of pavement cracks is important for maintaining road quality and ensuring traffic safety. However, the complexity of road surfaces and the diversity of cracks make it difficult for existing methods to accomplish this challenging task. This paper proposes a novel network named the global graph multiscale network (GGMNet) for automated pixel-level detection of pavement cracks. The GGMNet network has several innovations compared with the mainstream road crack detection network: (1) a global contextual Res-block (GC-Resblock) is proposed to guide the network to emphasize the identities of cracks while suppressing background noises; (2) a graph pyramid pooling module (GPPM) is designed to aggregate the multi-scale features and capture the long-range dependencies of cracks; (3) a multi-scale features fusion module (MFF) is established to efficiently represent and deeply fuse multi-scale features. We carried out extensive experiments on three pavement crack datasets. These were DeepCrack dataset, with complex background noises; the CrackTree260 dataset, with various crack structures; and the Aerial Track Detection dataset, with a drone’s perspective. The experimental results demonstrate that GGMNet has excellent performance, high accuracy, and strong robustness. In conclusion, this paper provides support for accurate and timely road maintenance and has important reference values and enlightening implications for further linear feature extraction research. Full article
(This article belongs to the Special Issue Road Extraction and Distress Assessment by Spaceborne, Airborne and Terrestrial Platforms (Second Edition))
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Network framework of the proposed GGMNet.</p> Full article ">Figure 2
<p>Architecture of the proposed GC-Resblock.</p> Full article ">Figure 3
<p>Framework of the proposed GPPM.</p> Full article ">Figure 4
<p>Schematic diagram of spatial reasoning.</p> Full article ">Figure 5
<p>Framework of the proposed CWF.</p> Full article ">Figure 6
<p>Visualization of the outcomes produced by diverse methods for DeepCrack.</p> Full article ">Figure 7
<p>Visualization of the outcomes produced by diverse methods for CrackTree260.</p> Full article ">Figure 8
<p>Visualization of the outcomes produced by diverse methods for Aerial Track Detection.</p> Full article ">Figure 9
<p>Visualization outcomes for different levels of GC-Resblock. (<b>a</b>,<b>b</b>) Before and after addition of the first layer of the encoder. (<b>c</b>,<b>d</b>) Before and after the addition of the second layer of the encoder. (<b>e</b>,<b>f</b>) Before and after the addition of the third layer of the encoder.</p> Full article ">Figure 10
<p>F1 score and IOU value of the different feature fusion methods for DeepCrack.</p> Full article ">
17 pages, 19375 KiB  
Article
Deep Blind Fault Activity—A Fault Model of Strong Mw 5.5 Earthquake Seismogenic Structures in North China
by Guanshen Liu, Renqi Lu, Dengfa He, Lihua Fang, Yang Zhang, Peng Su and Wei Tao
Remote Sens. 2024, 16(10), 1796; https://doi.org/10.3390/rs16101796 - 18 May 2024
Viewed by 444
Abstract
North China is one of the high-risk areas for destructive and strong earthquakes in mainland China and has experienced numerous strong historical earthquakes. An earthquake of magnitude MW 5.5 struck Pingyuan County, Dezhou city, in Shandong Province, China, on 6 August 2023. [...] Read more.
North China is one of the high-risk areas for destructive and strong earthquakes in mainland China and has experienced numerous strong historical earthquakes. An earthquake of magnitude MW 5.5 struck Pingyuan County, Dezhou city, in Shandong Province, China, on 6 August 2023. This earthquake was the strongest in the eastern North China Craton since the 1976 Tangshan earthquake. Since the earthquake did not produce surface ruptures, the seismogenic structure for fault responsible for the Pingyuan MW 5.5 earthquake is still unclear. To reveal the subsurface geological structure near the earthquake epicenter, this study used high-resolution two-dimensional (2D) seismic reflection profiles and constructed a three-dimensional (3D) geometric model of the Tuqiao Fault by interpreting the faults in the seismic reflection profiles. This study further combined focal mechanism solutions, aftershock clusters, and other seismological data to discuss the seismogenic fault of the Pingyuan MW 5.5 earthquake. The results show that the Tuqiao Fault is not the seismogenic fault of the MW 5.5 earthquake. The actual seismogenic structure may be related to the NE-oriented high-angle strike-slip blind fault developed in the basement. We further propose three possible fault models for the strong seismogenic structure in North China to discuss the potential seismotectonics in this region. Full article
(This article belongs to the Special Issue Earthquake Disaster Monitoring Using Remote Sensing Image Processing and Geophysical Techniques)
►▼ Show Figures

Figure 1

Figure 1
<p>Geological structure map of the eastern NCC. The red dots represent the earthquakes that occurred in the eastern NCC from 1900 to 2023, and the focal mechanism solutions are all lower hemisphere projections [<a href="#B7-remotesensing-16-01796" class="html-bibr">7</a>,<a href="#B14-remotesensing-16-01796" class="html-bibr">14</a>]. The black solid lines represent faults that reach the surface, while the black dashed lines represent buried faults. The yellow pentagram represents the epicenter of the <span class="html-italic">M<sub>W</sub></span> 5.5 earthquake. The grayscale represents the elevation. The pink dots superimposed on the pentagram represent the aftershocks of the <span class="html-italic">M<sub>W</sub></span> 5.5 earthquake. The red box represents the location of the study area; the blue box represents the area where <a href="#remotesensing-16-01796-f002" class="html-fig">Figure 2</a>a is located. The dark-blue solid lines L1 and L2 represent the positions of the deep seismic reflection profiles passing through the epicenters of the Xingtai and Tangshan earthquakes, respectively.</p> Full article ">Figure 2
<p>Geological map and seismic intensity map near the epicenter. (<b>a</b>) is the tectonic unit map of the Linqing Depression. (<b>b</b>) is the seismic intensity map. The yellow solid line represents the position of the 2D seismic reflection profile, the blue solid line represents the position of the velocity profile, the light-blue solid line represents the river, and the black square represents the city. The focal mechanism solution and aftershock relocation is cited from Zhang et al., (2024) [<a href="#B1-remotesensing-16-01796" class="html-bibr">1</a>]. The purple and pink ovals represent the distribution ranges of areas of different seismic intensity (<a href="https://www.cea.gov.cn/" target="_blank">https://www.cea.gov.cn/</a>; accessed on 8 August 2023).</p> Full article ">Figure 3
<p>Lithology in the Linqing Depression. The dashed lines between the strata represent parallel unconformities, while the wavy lines represent angular unconformities.</p> Full article ">Figure 4
<p>Seismic reflection profile AA’ and its interpretation. (<b>a</b>) is seismic reflection profile. (<b>b</b>) is interpretation scheme. The yellow pentagram represents the centroid position of the <span class="html-italic">M<sub>W</sub></span> 5.5 earthquake. The solid red line represents the fault, and the red arrow represents the direction of movement of the fault. The black box represents the position of <a href="#remotesensing-16-01796-f005" class="html-fig">Figure 5</a>. The F1 fault is the Kongzhuang Fault; F2 represents the Tuxi Fault, and F3 represents the Tuqiao Fault.</p> Full article ">Figure 5
<p>Partial detail of the breakpoint on the Tuqiao Fault (located in the black box of <a href="#remotesensing-16-01796-f004" class="html-fig">Figure 4</a>). (<b>a</b>) is the seismic reflection profile. (<b>b</b>) is the seismic interpretation. The red solid line represents the fault, the red dashed line represents the inferred fault, and the red arrow represents the location of the breakpoint and fold. The blue dashed line represents the bottom boundary of different strata. T1 is the bottom boundary of the Neogene Minghuazhen Formation, T2 is the bottom boundary of the Neogene Guantao Formation, and T3 is the bottom boundary of the Paleogene Kongdian Formation.</p> Full article ">Figure 6
<p>The 3D geometric model of the Tuqiao Fault. (<b>a</b>) shows the depth map of the Tuqiao Fault; (<b>b</b>) shows the distribution of dip angles of Tuqiao Fault. The red solid line represents the fault trace at the top of the fault, the two short red lines represent the position of the hanging wall of the fault, the gray dashed line represents the contour of the fault, and the light-blue transparent area represents nodal plane I in the focal mechanism solution.</p> Full article ">Figure 7
<p>Surface deformation signal from our InSAR observation. The red dashed line represents the concealed seismogenic fault. The two red solid lines indicate the position of the hanging wall of the seismogenic fault. The black thin dashed lines represent pre-existing concealed faults.</p> Full article ">Figure 8
<p>(<b>a</b>) is velocity structure profile BB’. (<b>b</b>) is location map of faults and earthquakes. The white dashed line represents the basement fault inferred from the cluster of aftershocks. The green circle represents the location of an aftershock. The long red solid line represents the fault trace at the top of the fault, the two short red solid lines represent the tendency of the fault, the gray dashed line represents the contour of the fault, the different colors of the sources represent different magnitudes, the different colors of the velocity data represent different velocity magnitudes, and a transparent blue area represents low-velocity body.</p> Full article ">Figure 9
<p>Seismogenic models of the Xingtai and Tangshan earthquakes in North China. (<b>a</b>) shows the seismogenic model of the Xingtai earthquake (modified by Wang et al., (1997) [<a href="#B7-remotesensing-16-01796" class="html-bibr">7</a>]); (<b>b</b>) shows the seismogenic model of the Tangshan earthquake (modified by Liu et al., (2011) [<a href="#B9-remotesensing-16-01796" class="html-bibr">9</a>]). The red solid line represents the fault, and the red dashed line represents the inferred basement fault. Tc is the boundary between the upper and lower crust. The locations of the deep seismic reflection profiles are shown in <a href="#remotesensing-16-01796-f001" class="html-fig">Figure 1</a>.</p> Full article ">Figure 10
<p>Tectonic evolution and current strong earthquake structural models of the eastern graben basin in North China. (<b>a</b>) shows the development pattern of faults during the early Cenozoic basin graben period; (<b>b</b>–<b>d</b>) show possible seismogenic patterns since the late Cenozoic. The red solid line represents the fault during the active period, the black solid line represents the fault during the inactive period, and the red dashed line represents the basement detachment. The red arrow indicates the direction of movement of the active faults, while the black arrow indicates the direction of movement of the inactive faults. The yellow pentagram represents the hypocenter of strong earthquake.</p> Full article ">
25 pages, 10663 KiB  
Article
DDFNet-A: Attention-Based Dual-Branch Feature Decomposition Fusion Network for Infrared and Visible Image Fusion
by Qiancheng Wei, Ying Liu, Xiaoping Jiang, Ben Zhang, Qiya Su and Muyao Yu
Remote Sens. 2024, 16(10), 1795; https://doi.org/10.3390/rs16101795 - 18 May 2024
Viewed by 451
Abstract
The fusion of infrared and visible images aims to leverage the strengths of both modalities, thereby generating fused images with enhanced visible perception and discrimination capabilities. However, current image fusion methods frequently treat common features between modalities (modality-commonality) and unique features from each [...] Read more.
The fusion of infrared and visible images aims to leverage the strengths of both modalities, thereby generating fused images with enhanced visible perception and discrimination capabilities. However, current image fusion methods frequently treat common features between modalities (modality-commonality) and unique features from each modality (modality-distinctiveness) equally during processing, neglecting their distinct characteristics. Therefore, we propose a DDFNet-A for infrared and visible image fusion. DDFNet-A addresses this limitation by decomposing infrared and visible input images into low-frequency features depicting modality-commonality and high-frequency features representing modality-distinctiveness. The extracted low and high features were then fused using distinct methods. In particular, we propose a hybrid attention block (HAB) to improve high-frequency feature extraction ability and a base feature fusion (BFF) module to enhance low-frequency feature fusion ability. Experiments were conducted on public infrared and visible image fusion datasets MSRS, TNO, and VIFB to validate the performance of the proposed network. DDFNet-A achieved competitive results on three datasets, with EN, MI, VIFF, QAB/F, FMI, and Qs metrics reaching the best performance on the TNO dataset, achieving 7.1217, 2.1620, 0.7739, 0.5426, 0.8129, and 0.9079, respectively. These values are 2.06%, 11.95%, 21.04%, 21.52%, 1.04%, and 0.09% higher than those of the second-best methods, respectively. The experimental results confirm that our DDFNet-A achieves better fusion performance than state-of-the-art (SOTA) methods. Full article
(This article belongs to the Special Issue Remote Sensing: 15th Anniversary)
►▼ Show Figures

Figure 1

Figure 1
<p>The architecture of the DDFNet-A.</p> Full article ">Figure 2
<p>The architecture of the DFE and HAB.</p> Full article ">Figure 3
<p>The architecture of the BFF.</p> Full article ">Figure 4
<p>Qualitative comparison of selected images from the MSRS dataset: (<b>a</b>) 00196D; (<b>b</b>) 00131D; and (<b>c</b>) 00770N. Some targets and details are annotated with red and green boxes to highlight noteworthy information.</p> Full article ">Figure 5
<p>Object comparisons of 20 pairs of images selected from the MSRS dataset.</p> Full article ">Figure 6
<p>Qualitative comparison of selected images from the TNO dataset: (<b>a</b>) Kaptein 1123; (<b>b</b>) Street; and (<b>c</b>) Nato camp. Some targets and details are annotated with red and green boxes to highlight noteworthy information.</p> Full article ">Figure 7
<p>Object comparisons of 25 pairs of images selected from the TNO dataset.</p> Full article ">Figure 8
<p>Qualitative comparison of selected images from the VIFB dataset: (<b>a</b>) fight; (<b>b</b>) people shallow; and (<b>c</b>) running. Some targets and details are annotated with red and green boxes to highlight noteworthy information.</p> Full article ">Figure 9
<p>Object comparisons of 18 pairs of images selected from the VIFB dataset.</p> Full article ">Figure 10
<p>Qualitative comparison of selected images from the Lytro and Grayscale dataset: (<b>a</b>) far-focus images; (<b>b</b>) near-focus images; and (<b>c</b>) fused images.</p> Full article ">
23 pages, 8508 KiB  
Article
An Evaluation of Optimization Algorithms for the Optimal Selection of GNSS Satellite Subsets
by Abdulaziz Alluhaybi, Panos Psimoulis and Rasa Remenyte-Prescott
Remote Sens. 2024, 16(10), 1794; https://doi.org/10.3390/rs16101794 - 18 May 2024
Viewed by 564
Abstract
Continuous advancements in GNSS systems have led, apart from the broadly used GPS, to the development of other satellite systems (Galileo, BeiDou, GLONASS), which have significantly increased the number of available satellites for GNSS positioning applications. However, despite GNSS satellites’ redundancy, a potential [...] Read more.
Continuous advancements in GNSS systems have led, apart from the broadly used GPS, to the development of other satellite systems (Galileo, BeiDou, GLONASS), which have significantly increased the number of available satellites for GNSS positioning applications. However, despite GNSS satellites’ redundancy, a potential poor GNSS satellite signal (i.e., low signal-to-noise ratio) can negatively affect the GNSS’s performance and positioning accuracy. On the other hand, selecting high-quality GNSS satellite signals by retaining a sufficient number of GNSS satellites can enhance the GNSS’s positioning performance. Various methods, including optimization algorithms, which are also commonly adopted in artificial intelligence (AI) methods, have been applied for satellite selection. In this study, five optimization algorithms were investigated and assessed in terms of their ability to determine the optimal GNSS satellite constellation, such as Artificial Bee Colony optimization (ABC), Ant Colony Optimization (ACO), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Simulated Annealing (SA). The assessment of the optimization algorithms was based on two criteria, such as the robustness of the solution for the optimal satellite constellation and the time required to find the solution. The selection of the GNSS satellites was based on the weighted geometric dilution of precision (WGDOP) parameter, where the geometric dilution of precision (GDOP) is modified by applying weights based on the quality of the satellites’ signal. The optimization algorithms were tested on the basis of 24 h of tracking data gathered from a permanent GNSS station, for GPS-only and multi-GNSS data (GPS, GLONASS, and Galileo). According to the comparison results, the ABC, ACO, and PSO algorithms were equivalent in terms of selection accuracy and speed. However, ABC was determined to be the most suitable algorithm due it requiring the fewest number of parameters to be set. To further investigate ABC’s performance, the method was applied for the selection of an optimal GNSS satellite subset according to the number of total available tracked GNSS satellites (up to 31 satellites), leading to more than 300 million possible combinations of 15 GNSS satellites. ABC was able to select the optimal satellite subsets with 100% accuracy. Full article
(This article belongs to the Topic Artificial Intelligence in Navigation)
►▼ Show Figures

Figure 1

Figure 1
<p>Representation of the ABC searching process and the roles of employed scout and onlooker bees.</p> Full article ">Figure 2
<p>Schematic representation of the solution building process by ants in ACO.</p> Full article ">Figure 3
<p>Representation of the GA processing steps.</p> Full article ">Figure 4
<p>Representation of PSO travelling technique for the solution optimization.</p> Full article ">Figure 5
<p>Schematic representation of SA algorithm procedure.</p> Full article ">Figure 6
<p>Time period of satellites’ mean movement by one degree, considering satellite azimuth and elevation angles.</p> Full article ">Figure 7
<p>(<b>Left</b>) View of the roof of NGI building, with the location of control point NGB2 and (<b>right</b>) the GNSS antenna installed on the top of the pillar of NGB2.</p> Full article ">Figure 8
<p>Number of available GNSS satellites at NGB2 GNSS station for a 24 h period on 20 September 2021.</p> Full article ">Figure 9
<p>Sample of the file of the GPS data information, which includes (i) date–time, (ii) satellite PRN, (iii) azimuth, (iv) elevation angle, and (v) CNR (in dB-Hz).</p> Full article ">Figure 10
<p>The possible combinations of satellite constellations for (<b>left</b>) GPS-only in the cases of 8 and 13 available GPS satellites and (<b>right</b>) multi-GNSS satellite constellation in the case of 18 and 31 available GNSS satellites.</p> Full article ">Figure 11
<p>(<b>Left</b>) The quality of match (accuracy) of the selection of the optimal GPS satellite subset by the optimization algorithms with respect to the actual optimal GPS satellite subset derived by the TM. (<b>Right</b>) The time required for the TM and the optimization algorithms to perform optimal GPS satellite subset selection.</p> Full article ">Figure 12
<p>The comparison of the performance of the optimization algorithms with respect to TM, expressed as the difference between the CNR-WGDOP of the optimal satellite constellation of each optimization algorithm and the corresponding CNR-WDGOP of TM. The results of the four cases of GPS satellite constellations (4, 5, 6, and 7 satellites) are presented. On the left axis, the CNR-WGDOP value of the optimal satellite constellation based on the TM is presented, and on the right axis is the difference between each of the optimization algorithms and the TM.</p> Full article ">Figure 13
<p>The sky plots of epoch 55 (<b>left</b>) and epoch 184 (<b>right</b>) presenting the selection of the optimal GPS satellite subset by the ACO and the TM, and showing the satellites commonly selected (blue) by the two methods, but also those that were differently selected by ACO (yellow) and TM (orange).</p> Full article ">Figure 14
<p>Same as <a href="#remotesensing-16-01794-f013" class="html-fig">Figure 13</a>, this figure presents the sky plots for epoch 81 (<b>left</b>) and epoch 215 (<b>right</b>), as well as differences between the selection of the optimal GPS satellite subset for PSO and TM.</p> Full article ">Figure 15
<p>(<b>Left</b>) The accuracy of the selection of the optimal GNSS satellite subset of ABC with respect to TM for the various cases of satellite constellations and parameter settings and (<b>right</b>) the required time of the TM and ABC algorithm to compute the selection of optimal GNSS satellite subset.</p> Full article ">Figure 16
<p>A comparison of the performance of the ABC algorithm for the three sets of parameter settings with respect to TM, expressed as the difference between the CNR-WGDOP of the optimal satellite constellation of each ABC parameter setting and the corresponding CNR-WDGOP of TM. On the left axis, the CNR-WGDOP value of the optimal satellite constellation based on the TM is presented, and on the right axis is the difference between each of the ABC parameter settings and the TM.</p> Full article ">Figure 17
<p>Sky plots of epochs 20 (<b>left</b>) and 88 (<b>right</b>), presenting the difference in the selection of optimal GNSS satellite subset between ABC setting 1 and the actual TM, by showing the common satellites (blue) and the differences between ABC’s (yellow) and TM’s (orange) satellite selection.</p> Full article ">
24 pages, 12245 KiB  
Article
How Representative Are European AERONET-OC Sites of European Marine Waters?
by Ilaria Cazzaniga and Frédéric Mélin
Remote Sens. 2024, 16(10), 1793; https://doi.org/10.3390/rs16101793 - 18 May 2024
Viewed by 392
Abstract
Data from the Ocean Color component of the Aerosol Robotic Network (AERONET-OC) have been extensively used to assess Ocean Color radiometric products from various satellite sensors. This study, focusing on Ocean Color radiometric operational products from the Sentinel-3 Ocean and Land Colour Instrument [...] Read more.
Data from the Ocean Color component of the Aerosol Robotic Network (AERONET-OC) have been extensively used to assess Ocean Color radiometric products from various satellite sensors. This study, focusing on Ocean Color radiometric operational products from the Sentinel-3 Ocean and Land Colour Instrument (OLCI), aims at investigating where in the European seas the results of match-up analyses at the European marine AERONET-OC sites could be applicable. Data clustering is applied to OLCI remote sensing reflectance RRS(λ) from the various sites to define different sets of optical classes, which are later used to identify class-based uncertainties. A set of fifteen classes grants medium-to-high classification levels to most European seas, with exceptions in the South-East Mediterranean Sea, the Atlantic Ocean, or the Gulf of Bothnia. In these areas, RRS(λ) spectra are very often identified as novel with respect to the generated set of classes, suggesting their under-representation in AERONET-OC data. Uncertainties are finally mapped onto European seas according to class membership. The largest uncertainty values are obtained in the blue spectral region for almost all classes. In clear waters, larger values are obtained in the blue bands. Conversely, larger values are shown in the green and red bands in coastal and turbid waters. Full article
(This article belongs to the Special Issue Editorial Board Members’ Collection Series: Recent Progress in Ocean Colour Remote Sensing)
►▼ Show Figures

Figure 1

Figure 1
<p>Location of the European AERONET-OC marine sites.</p> Full article ">Figure 2
<p>Spectral classes represented by mean and standard deviation (through error bars) values of the fourteen clusters identified by ISODATA from in situ <math display="inline"><semantics> <mrow> <mi>n</mi> <msub> <mrow> <mi>R</mi> </mrow> <mrow> <mi>R</mi> <mi>S</mi> </mrow> </msub> <mfenced separators="|"> <mrow> <mi mathvariant="sans-serif">λ</mi> </mrow> </mfenced> </mrow> </semantics></math> spectra at the coastal European AERONET-OC sites considered as a bulk (‘in situ’ set).</p> Full article ">Figure 3
<p>Spectral classes represented by mean and standard deviation values of the clusters identified by ISODATA from OLCI <math display="inline"><semantics> <mrow> <mi>n</mi> <msub> <mrow> <mi>R</mi> </mrow> <mrow> <mi>R</mi> <mi>S</mi> </mrow> </msub> <mfenced separators="|"> <mrow> <mi mathvariant="sans-serif">λ</mi> </mrow> </mfenced> </mrow> </semantics></math> spectra at each coastal European AERONET-OC site (‘OLCI-<span class="html-italic">XX</span>’ sets).</p> Full article ">Figure 4
<p>Mean and standard deviation values of the 15 clusters identified by ISODATA from OLCI <math display="inline"><semantics> <mrow> <mi>n</mi> <msub> <mrow> <mi>R</mi> </mrow> <mrow> <mi>R</mi> <mi>S</mi> </mrow> </msub> <mfenced separators="|"> <mrow> <mi mathvariant="sans-serif">λ</mi> </mrow> </mfenced> </mrow> </semantics></math> spectra at all the coastal European AERONET-OC sites (‘OLCI-all’ set).</p> Full article ">Figure 5
<p>AERONET-OC sites relative contribution to the 15 clusters. <span class="html-italic">N</span> is the number of spectra in each cluster and <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>p</mi> </mrow> <mrow> <mi>t</mi> </mrow> </msub> </mrow> </semantics></math> the average total membership for the cluster, while the number on the bottom-left corner indicates the cluster.</p> Full article ">Figure 6
<p>Cluster distribution among AERONET-OC sites. <span class="html-italic">N</span> is the number of spectra at each site.</p> Full article ">Figure 7
<p>(<b>Top</b>) Most frequently assigned site in European seas in 2022 (S3A and S3B). (<b>Bottom</b>) Mean total membership values in 2022. Black pixels represent no data pixels.</p> Full article ">Figure 8
<p>(<b>Top</b>) Most frequently assigned class (from the 15-cluster set) in European seas in 2022 (S3A and S3B). Black and white pixels represent no data and unclassified pixels, respectively. (<b>Bottom</b>) Mean total membership values in 2022.</p> Full article ">Figure 9
<p><math display="inline"><semantics> <mrow> <msub> <mrow> <mi>R</mi> <mi>M</mi> <mi>S</mi> <mi>D</mi> </mrow> <mrow> <mi>u</mi> </mrow> </msub> <mo>(</mo> <mi mathvariant="sans-serif">λ</mi> <mo>)</mo> </mrow> </semantics></math> for AERONET-OC and OLCI match-ups for (<b>top</b>) S3A and (<b>bottom</b>) S3B. Crosses indicate the values for 620 nm center-wavelengths calculated without ZEE and TCP match-up values (where <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>R</mi> </mrow> <mrow> <mi>R</mi> <mi>S</mi> </mrow> </msub> <mo>(</mo> <mn>620</mn> <mo>)</mo> </mrow> </semantics></math> is not available).</p> Full article ">Figure 10
<p>Average <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>R</mi> <mi>M</mi> <mi>S</mi> <mi>D</mi> </mrow> <mrow> <mi>u</mi> </mrow> </msub> <mo>(</mo> <mi mathvariant="sans-serif">λ</mi> <mo>)</mo> </mrow> </semantics></math> (in sr<sup>−1</sup>) at (<b>a</b>) 412.5 nm, (<b>b</b>) 560 nm, and (<b>c</b>) 665 nm and (<b>d</b>) novelty frequency map (in %) for S3A data in 2022.</p> Full article ">Figure 11
<p>S3A mean total membership <math display="inline"><semantics> <mrow> <msub> <mrow> <mi>p</mi> </mrow> <mrow> <mi>t</mi> </mrow> </msub> </mrow> </semantics></math> for the various 2022 seasons: (<b>a</b>) winter, (<b>b</b>) spring, (<b>c</b>) summer, and (<b>d</b>) autumn.</p> Full article ">
15 pages, 4772 KiB  
Technical Note
Eutrophication and HAB Occurrence Control in Lakes of Different Origins: A Multi-Source Remote Sensing Detection Strategy
by Giovanni Laneve, Alejandro Téllez, Ashish Kallikkattil Kuruvila, Milena Bruno and Valentina Messineo
Remote Sens. 2024, 16(10), 1792; https://doi.org/10.3390/rs16101792 - 18 May 2024
Viewed by 485
Abstract
Remote sensing techniques have become pivotal in monitoring algal blooms and population dynamics in freshwater bodies, particularly to assess the ecological risks associated with eutrophication. This study focuses on remote sensing methods for the analysis of 4 Italian lakes with diverse geological origins, [...] Read more.
Remote sensing techniques have become pivotal in monitoring algal blooms and population dynamics in freshwater bodies, particularly to assess the ecological risks associated with eutrophication. This study focuses on remote sensing methods for the analysis of 4 Italian lakes with diverse geological origins, leveraging water quality samples and data from the Sentinel-2 and Landsat 5.7–8 platforms. Chl-a, a well-correlated indicator of phytoplankton biomass abundance and eutrophication, was estimated using ordinary least squares linear regression to calibrate surface reflectance with chl-a concentrations. Temporal gaps between sample and image acquisition were considered, and atmospheric correction dedicated to water surfaces was implemented using ACOLITE and those specific to each satellite platform. The developed models achieved determination coefficients higher than 0.69 with mean square errors close to 3 mg/m3 for water bodies with low turbidity. Furthermore, the time series described by the models portray the seasonal variations in the lakes water bodies. Full article
(This article belongs to the Special Issue Satellite-Based Climate Change and Sustainability Studies)
►▼ Show Figures

Figure 1

Figure 1
<p>Study area with in situ samples according to Sentinel 2.</p> Full article ">Figure 2
<p>Flowchart of Sentinel-2 and Landsat image processing.</p> Full article ">Figure 3
<p>Seasonal (Nov.-Mar.) chl-<span class="html-italic">a</span> average in the five superficial Vico stations with their regression lines, trend lines and values.</p> Full article ">Figure 4
<p>BMAA vs. microcystins concentration (μg/L) in lakes Vico and Albano.</p> Full article ">Figure 5
<p>Scatter plot, estimated chl-<span class="html-italic">a</span> values vs. in situ values.</p> Full article ">Figure 6
<p>Time series of chl<span class="html-italic">-a</span> for Trasimeno, Bolsena, Albano, and Vico lakes.</p> Full article ">Figure 7
<p>Landsat and Sentinel 2 historical series.</p> Full article ">Figure 8
<p>Optical water property estimate with Sentinel-2 images.</p> Full article ">Figure 9
<p>Temporal trends of chlorophyll-<span class="html-italic">a</span> concentrations: a comparative analysis of in situ and Landsat images.</p> Full article ">
25 pages, 9092 KiB  
Article
Constrained Iterative Adaptive Algorithm for the Detection and Localization of RFI Sources Based on the SMAP L-Band Microwave Radiometer
by Xinxin Wang, Xiang Wang, Lin Wang, Jianchao Fan and Enbo Wei
Remote Sens. 2024, 16(10), 1791; https://doi.org/10.3390/rs16101791 - 18 May 2024
Viewed by 372
Abstract
The Soil Moisture Active Passive (SMAP) satellite carries an L-band microwave radiometer. This sensor can be used to observe global soil moisture (SM) and sea surface salinity (SSS) within the protected L-band spectrum (1400–1427 MHz). Owing to the complex effects of radio frequency [...] Read more.
The Soil Moisture Active Passive (SMAP) satellite carries an L-band microwave radiometer. This sensor can be used to observe global soil moisture (SM) and sea surface salinity (SSS) within the protected L-band spectrum (1400–1427 MHz). Owing to the complex effects of radio frequency interference (RFI), the SM and SSS data are missing or have low accuracy. In this paper, a constrained iterative adaptive algorithm for the detection, identification, and localization of RFI sources is designed, named MICA-BEID. The algorithm synthesizes antenna temperatures for the third and fourth Stokes parameters before RFI filtering, creating a new polarization parameter called WSPDA, designed to approximate the level of RFI interference on the L-band microwave radiometer. The algorithm then utilizes the WSPDA intensity and distribution density of RFI detection samples to enhance the identification and classification of RFI sources across various intensity levels. By utilizing statistical methods such as the probability density function (PDF) and the cumulative distribution function (CDF), the algorithm dynamically adjusts adaptive parameters, including the RFI detection threshold and the maximum effective radius of RFI sources. Through the application of multiple iterative clustering methods, the algorithm can adaptively detect and identify RFI sources at various satellite orbits and intensity levels. Through extensive comparative analysis with other localization results and known RFI sources, the MICA-BEID algorithm can achieve optimal localization accuracy of approximately 1.2 km. The localization of RFI sources provides important guidance for identifying and turning off illegal RFI sources. Moreover, the localization and long-time-series characteristic analysis of RFI sources that cannot be turned off is of significant value for simulating the spatial distribution characteristics of localized RFI source intensity in local areas. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Flowchart for the constrained iterative adaptive algorithm.</p> Full article ">Figure 2
<p>Statistical histogram of calculation results of SMAP L1B half-orbit data parameter <span class="html-italic">W<sub>SPDA</sub></span> for (<b>a</b>) the ascending orbit and (<b>b</b>) the descending orbit.</p> Full article ">Figure 3
<p>Spatial distribution of SMAP RFI detection sample data. (<b>a</b>,<b>b</b>) RFI detection samples inside (green) and on the edges (red) of the ascending and descending orbits, respectively; (<b>c</b>,<b>d</b>) SMAP RFI SRDF samples (blue), SPDA samples (red) and overlapped samples (green) in ascending and descending orbits, respectively, with the samples on the edges being removed.</p> Full article ">Figure 3 Cont.
<p>Spatial distribution of SMAP RFI detection sample data. (<b>a</b>,<b>b</b>) RFI detection samples inside (green) and on the edges (red) of the ascending and descending orbits, respectively; (<b>c</b>,<b>d</b>) SMAP RFI SRDF samples (blue), SPDA samples (red) and overlapped samples (green) in ascending and descending orbits, respectively, with the samples on the edges being removed.</p> Full article ">Figure 4
<p>Local spatial distribution of SMAP RFI detection sample data at the water–land boundary. The background is the spatial interpolation results of <span class="html-italic">W<sub>SPDA</sub></span>. The detection results of the descending orbit at a detection threshold of 5.5 K are in red, and those at a detection threshold of 9.3 K are in green.</p> Full article ">Figure 5
<p>Spatial distribution of RFI detection samples after interpolation for (<b>a</b>) land only and (<b>b</b>) at the water–land boundary.</p> Full article ">Figure 6
<p><span class="html-italic">R<sub>max</sub></span> statistical analysis examples. (<b>a</b>,<b>b</b>) Statistical results of RFI detection samples for fore and aft looks, respectively; (<b>c</b>,<b>d</b>) spatial distribution of <span class="html-italic">W<sub>SPDA</sub></span> intensity of clusters for fore and aft looks, respectively. The background is the spatial interpolation results of <span class="html-italic">W<sub>SPDA</sub></span> for all fore and aft detection samples.</p> Full article ">Figure 7
<p>The changing characteristics of clustering identification results with iteration count. (<b>a</b>) Number of detected samples; (<b>b</b>) number of clustered samples; (<b>c</b>) number of RFI sources identified; (<b>d</b>) <span class="html-italic">W<sub>max</sub></span> of RFI sources.</p> Full article ">Figure 8
<p>Spatial distribution of RFI detection samples for the SMAP satellite’s fore and aft looks footprints.</p> Full article ">Figure 9
<p>Filtering results of clusters at the water–land boundary. (<b>a</b>) All clustering results for all samples, where various colors denote different clusters; (<b>b</b>) results of cluster filtering, with red indicating the removed clusters and green indicating the retained clusters.</p> Full article ">Figure 10
<p>A typical example of the effects of Faraday rotation during a period of high geomagnetic disturbance on 27 February 2023.</p> Full article ">Figure 11
<p>Statistical analysis results of the impact of Faraday rotation. (<b>a</b>) Data from half-orbit of the SMAP satellite. (<b>b</b>) Data within the blue box range, as depicted in <a href="#remotesensing-16-01791-f010" class="html-fig">Figure 10</a>, of a typical example. The red line in both (<b>a</b>) and (<b>b</b>) represents the result of linear fitting.</p> Full article ">Figure 12
<p>(<b>a</b>) Two- and (<b>b</b>) three-dimensional intensity distributions of <span class="html-italic">W<sub>SPDA</sub></span> of clusters for descending orbit. Red and black circles represent fore and aft antenna scanning orbits, respectively.</p> Full article ">Figure 13
<p>Statistical diagrams of spatial distribution of the normalized parameter with distance for descending orbits.</p> Full article ">Figure 14
<p>RFI identification and determination criteria. The red stars represent outlier data that have been removed.</p> Full article ">Figure 15
<p>Spatial distribution characteristic of long-time-series RFI location data points and the RFI location centroid.</p> Full article ">Figure 16
<p>The characteristics of the relationship between the <span class="html-italic">W<sub>SPDA</sub></span> parameter and the antenna temperatures for (<b>a</b>) horizontal (<span class="html-italic">Ta_h</span>) and (<b>b</b>) vertical (<span class="html-italic">Ta_v</span>) polarizations.</p> Full article ">Figure 17
<p>Comparison results between MICA-BEID algorithm and SMAP RFI surveys. (<b>a</b>) Comparison of RFI location results; (<b>b</b>) comparison of RFI levels.</p> Full article ">Figure 18
<p>Comparative analysis of MICA-BEID algorithm RFI localization results with known RFI sources on a long-term scale. (<b>a</b>) Hebei. (<b>b</b>) Tianjin. (<b>c</b>) Shandong. (<b>d</b>) Hunan.</p> Full article ">Figure 19
<p>Comparative analysis of RFI localization results with MICA-BEID algorithm with <span class="html-italic">W<sub>SPDA</sub></span> of the SMOS satellite.</p> Full article ">Figure 20
<p>Comparative analysis of RFI localization results with MICA-BEID algorithm with the average TB-RMSE.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-171
Previous Issue
Next Issue
Back to TopTop