Pub Date : 2025-02-16DOI: 10.1016/j.rse.2025.114651
Hengyi Chen , Chaoying Zhao , Roberto Tomás , Cristina Reyes-Carmona , Ya Kang
InSAR and optical pixel offset tracking (POT) are two efficient tools for monitoring landslide displacements, but limitations in resolving 3D displacements constrain the full exploration of kinematic behaviors, especially for complex landslides exhibiting diverse movement types. In this study, we propose a technical route that combines SAR and optical images to reveal the spatiotemporal evolution of the Lanuza landslide (Spain). In the temporal domain, ascending and descending Sentinel-1 SAR images were acquired to retrieve the line-of-sight (LOS) displacements. STL and cross wavelet transform were integrated to calculate the time lag between displacements and environmental factors. In the spatial domain, a two-stage method combining feature point matching and DeepFlow (FPM-DF) was proposed to retrieve the non-rigid horizontal displacements from optical images. A strain model and Bayesian inversion framework (SM-BIF) were integrated to invert 3D displacement fields. The mass conservation method was subsequently applied to estimate the landslide thickness. The results indicate that (1) the periodic terms of displacement are in phase with the freeze-thaw cycle of solifluction, which can intensify earthflow movement. (2) FPM-DF method is more efficient than the traditional POT method, especially for small-scale displacement fields, achieving reductions of standard deviations by 38 % and 51 % in the EW and NS directions, respectively. (3) the SM-BIF method reduces the maximum standard deviations of the 3D displacement field compared to the SM-VCE method, and the maximum thickness of the earthflow is approximately 22 m. This study can provide valuable insights into comprehensive monitoring of complex landslides with multi-platform remote sensing datasets.
{"title":"Integrating InSAR and non-rigid optical pixel offsets to explore the kinematic behaviors of the Lanuza complex landslide","authors":"Hengyi Chen , Chaoying Zhao , Roberto Tomás , Cristina Reyes-Carmona , Ya Kang","doi":"10.1016/j.rse.2025.114651","DOIUrl":"10.1016/j.rse.2025.114651","url":null,"abstract":"<div><div>InSAR and optical pixel offset tracking (POT) are two efficient tools for monitoring landslide displacements, but limitations in resolving 3D displacements constrain the full exploration of kinematic behaviors, especially for complex landslides exhibiting diverse movement types. In this study, we propose a technical route that combines SAR and optical images to reveal the spatiotemporal evolution of the Lanuza landslide (Spain). In the temporal domain, ascending and descending Sentinel-1 SAR images were acquired to retrieve the line-of-sight (LOS) displacements. STL and cross wavelet transform were integrated to calculate the time lag between displacements and environmental factors. In the spatial domain, a two-stage method combining feature point matching and DeepFlow (FPM-DF) was proposed to retrieve the non-rigid horizontal displacements from optical images. A strain model and Bayesian inversion framework (SM-BIF) were integrated to invert 3D displacement fields. The mass conservation method was subsequently applied to estimate the landslide thickness. The results indicate that (1) the periodic terms of displacement are in phase with the freeze-thaw cycle of solifluction, which can intensify earthflow movement. (2) FPM-DF method is more efficient than the traditional POT method, especially for small-scale displacement fields, achieving reductions of standard deviations by 38 % and 51 % in the EW and NS directions, respectively. (3) the SM-BIF method reduces the maximum standard deviations of the 3D displacement field compared to the SM-VCE method, and the maximum thickness of the earthflow is approximately 22 m. This study can provide valuable insights into comprehensive monitoring of complex landslides with multi-platform remote sensing datasets.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"320 ","pages":"Article 114651"},"PeriodicalIF":11.1,"publicationDate":"2025-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143417970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-14DOI: 10.1016/j.rse.2025.114649
Zhuoheng Chen, Stephen E. Grasby, Wanju Yuan, Di Lu, Christine Deblonde
Land surface temperature (LST) from satellite images contains meaningful signatures of geothermal heat flux (GHF) for geothermal exploration. However, the signal is mixed with solar radiation dominated features, making it difficult to identify GHF anomaly. Here we propose a novel method to tackle this problem that removes the time variant solar component based on principles of energy balance. Through an iteration process examining multiple LST maps from different seasons, the temporally invariant GHF component can be revealed. We tested this method by examining the Mount Meager Volcanic Complex area in British Columbia, Canada where data of known geothermal prospects are publicly accessible for validation. Seventy-two Landsat-8 cloud-free LST maps acquired in the last 10 years, were employed to extract the GHF component. Four high GHF anomalies are identified and two are consistent with areas of known hot spring swarms that occur above identified geothermal prospects. A third anomaly is spatially coincident with an active landslide site where warm seeps from the sliding surface and faults/fractures within the moving land mass are responsible for the anomalies. The anomalies in the fourth one are predominately anthropogenic, related to heat emission from hydropower facilities. The proposed method provides an efficient way to extract non-solar sourced LST anomalies, adding a cost-effective tool for geothermal exploration, and environmental/geohazard monitoring.
{"title":"Identification of geothermal anomalies from Landsat derived land surface temperature, Mount Meager volcanic complex, British Columbia, Canada","authors":"Zhuoheng Chen, Stephen E. Grasby, Wanju Yuan, Di Lu, Christine Deblonde","doi":"10.1016/j.rse.2025.114649","DOIUrl":"10.1016/j.rse.2025.114649","url":null,"abstract":"<div><div>Land surface temperature (LST) from satellite images contains meaningful signatures of geothermal heat flux (GHF) for geothermal exploration. However, the signal is mixed with solar radiation dominated features, making it difficult to identify GHF anomaly. Here we propose a novel method to tackle this problem that removes the time variant solar component based on principles of energy balance. Through an iteration process examining multiple LST maps from different seasons, the temporally invariant GHF component can be revealed. We tested this method by examining the Mount Meager Volcanic Complex area in British Columbia, Canada where data of known geothermal prospects are publicly accessible for validation. Seventy-two Landsat-8 cloud-free LST maps acquired in the last 10 years, were employed to extract the GHF component. Four high GHF anomalies are identified and two are consistent with areas of known hot spring swarms that occur above identified geothermal prospects. A third anomaly is spatially coincident with an active landslide site where warm seeps from the sliding surface and faults/fractures within the moving land mass are responsible for the anomalies. The anomalies in the fourth one are predominately anthropogenic, related to heat emission from hydropower facilities. The proposed method provides an efficient way to extract non-solar sourced LST anomalies, adding a cost-effective tool for geothermal exploration, and environmental/geohazard monitoring.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"320 ","pages":"Article 114649"},"PeriodicalIF":11.1,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143417971","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-13DOI: 10.1016/j.rse.2025.114612
Thomas Roßberg, Michael Schmitt
The objective of this study is to investigate the relationship between the Normalized Difference Vegetation Index (NDVI) and Synthetic Aperture Radar (SAR) data at multiple frequencies, focusing on S- and C-band data with additional analysis for X- and L-band. This is the foundation for the translation of SAR data into NDVI values, thereby enabling the filling of gaps in NDVI data due to cloud cover. This study encompasses three distinct study areas in Argentina, Australia, and Vietnam, which exhibit considerable climatic and agricultural differences. NovaSAR-1 S-band and Sentinel-1 C-band data were acquired for all areas, with the addition of COSMO-SkyMed X-band and SAOCOM L-band SAR data for one region. Following the processing of the SAR data and the derivation of NDVI values from optical Sentinel-2 data, the relationship between them is analyzed for field-wise aggregated data.
The relationship between S- and C-band SAR data and NDVI values is observed to be strong for all fields. Consequently, cross-polarized (HV or VH) data demonstrated this relationship for all fields with a Pearson correlation coefficient , whereas for co-polarized data (HH or VV), this could only be shown for some fields and crops. In the case of rice paddy fields, however, a different relationship is observed. While both S- and C-band data demonstrate a good relationship, this is primarily evident in the case of co-polarized data, with cross-polarized data exhibiting a comparatively weaker relationship. A relationship was observed for X-band data, but no relationship could be attested for L-band data. Neither the cross-ratio nor the radar vegetation index (RVI) generally showed a stronger relationship with the NDVI compared to a single polarization.
The demonstrated relationship between NDVI values and SAR backscatter data allows for a translation to be feasible. Consequently, the planned launch of the NISAR satellite, comprising S- and L-band SAR sensors, will facilitate new opportunities for agricultural monitoring. However, the retrieval of NDVI values from SAR data is a complex topic, as numerous factors, including crop type, crop phenology, SAR geometry and frequency, and others, influence this relationship.
{"title":"Comparing the relationship between NDVI and SAR backscatter across different frequency bands in agricultural areas","authors":"Thomas Roßberg, Michael Schmitt","doi":"10.1016/j.rse.2025.114612","DOIUrl":"10.1016/j.rse.2025.114612","url":null,"abstract":"<div><div>The objective of this study is to investigate the relationship between the Normalized Difference Vegetation Index (NDVI) and Synthetic Aperture Radar (SAR) data at multiple frequencies, focusing on S- and C-band data with additional analysis for X- and L-band. This is the foundation for the translation of SAR data into NDVI values, thereby enabling the filling of gaps in NDVI data due to cloud cover. This study encompasses three distinct study areas in Argentina, Australia, and Vietnam, which exhibit considerable climatic and agricultural differences. NovaSAR-1 S-band and Sentinel-1 C-band data were acquired for all areas, with the addition of COSMO-SkyMed X-band and SAOCOM L-band SAR data for one region. Following the processing of the SAR data and the derivation of NDVI values from optical Sentinel-2 data, the relationship between them is analyzed for field-wise aggregated data.</div><div>The relationship between S- and C-band SAR data and NDVI values is observed to be strong for all fields. Consequently, cross-polarized (HV or VH) data demonstrated this relationship for all fields with a Pearson correlation coefficient <span><math><mrow><mi>ρ</mi><mo>></mo><mn>0</mn><mo>.</mo><mn>5</mn></mrow></math></span>, whereas for co-polarized data (HH or VV), this could only be shown for some fields and crops. In the case of rice paddy fields, however, a different relationship is observed. While both S- and C-band data demonstrate a good relationship, this is primarily evident in the case of co-polarized data, with cross-polarized data exhibiting a comparatively weaker relationship. A relationship was observed for X-band data, but no relationship could be attested for L-band data. Neither the cross-ratio nor the radar vegetation index (RVI) generally showed a stronger relationship with the NDVI compared to a single polarization.</div><div>The demonstrated relationship between NDVI values and SAR backscatter data allows for a translation to be feasible. Consequently, the planned launch of the NISAR satellite, comprising S- and L-band SAR sensors, will facilitate new opportunities for agricultural monitoring. However, the retrieval of NDVI values from SAR data is a complex topic, as numerous factors, including crop type, crop phenology, SAR geometry and frequency, and others, influence this relationship.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"319 ","pages":"Article 114612"},"PeriodicalIF":11.1,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143395941","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-13DOI: 10.1016/j.rse.2025.114648
Jiaqi Tian , Xiangzhong Luo , Weile Wang , Liyao Yu , Diane Tan Ting Ng , Kazuhito Ichii , Yao Zhang , Xiaoyang Zhang
Tropical forests in the Amazon are characterized by a dry-season green-up, indicating a light-dominated regime in the seasonal variation of ecosystem functions. Southeast Asia, which hosts some of the most carbon-dense and diverse ecosystems in the world, is also expected to green up in dry seasons, however, recent in-situ evidence suggests otherwise. Here, we utilized high-frequency observations from the Himawari-8 geostationary satellite to examine the seasonality of vegetation greenness across Southeast Asia and to investigate potential factors driving this seasonality. We found that evergreen forests in maritime Southeast Asia green up in dry seasons, which is linked to positive anomaly in incoming radiation, similar to the Amazon forests. However, deciduous forests, croplands, and evergreen forests in continental Southeast Asia tend to green up in wet seasons, which is associated with the changes in climate sensitivities of vegetation greenness between dry and wet seasons. Our study highlights seasonal differences in tropical vegetation across biomes and sheds light on the underlying mechanisms of climate-vegetation interactions in Southeast Asia.
{"title":"Seasonality of vegetation greenness in Southeast Asia unveiled by geostationary satellite observations","authors":"Jiaqi Tian , Xiangzhong Luo , Weile Wang , Liyao Yu , Diane Tan Ting Ng , Kazuhito Ichii , Yao Zhang , Xiaoyang Zhang","doi":"10.1016/j.rse.2025.114648","DOIUrl":"10.1016/j.rse.2025.114648","url":null,"abstract":"<div><div>Tropical forests in the Amazon are characterized by a dry-season green-up, indicating a light-dominated regime in the seasonal variation of ecosystem functions. Southeast Asia, which hosts some of the most carbon-dense and diverse ecosystems in the world, is also expected to green up in dry seasons, however, recent in-situ evidence suggests otherwise. Here, we utilized high-frequency observations from the Himawari-8 geostationary satellite to examine the seasonality of vegetation greenness across Southeast Asia and to investigate potential factors driving this seasonality. We found that evergreen forests in maritime Southeast Asia green up in dry seasons, which is linked to positive anomaly in incoming radiation, similar to the Amazon forests. However, deciduous forests, croplands, and evergreen forests in continental Southeast Asia tend to green up in wet seasons, which is associated with the changes in climate sensitivities of vegetation greenness between dry and wet seasons. Our study highlights seasonal differences in tropical vegetation across biomes and sheds light on the underlying mechanisms of climate-vegetation interactions in Southeast Asia.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"319 ","pages":"Article 114648"},"PeriodicalIF":11.1,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143403230","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-12DOI: 10.1016/j.rse.2025.114645
Qiming Zheng , Yiwen Zeng , Yuyu Zhou , Zhuosen Wang , Te Mu , Qihao Weng
While severe hurricanes continue to challenge the resilience of local communities, fine-scale knowledge of post-hurricane recovery remains scarce. Existing recovery tracking approaches mainly rely on aggregated metrics that would disguise the spatial heterogeneity in recovery patterns. Here, we present a spatiotemporally explicit investigation into the recovery of human activity after 10 recent severe hurricanes in the U.S., with daily nighttime light (NTL) time series images from NASA's Black Marble VIIRS NTL product suite. We utilized a Bayesian-based time series change detection model and temporal clustering algorithm to analyze the post-hurricane recovery of each built-up area pixel within 446 counties severely affected by the hurricanes. To investigate the potential inaccuracies stemming from assessments using aggregated statistics, we further compared the recovery pattern estimated at pixel scale with that estimated by aggregated NTL radiance at county and census tract scales. Last, we examined the inequality in post-hurricane recovery and how it related to socioeconomic factors and current hurricane assistance programs. Our analysis shows a 7-fold difference in the recovery duration of hurricane-affected built-up areas within a county, with one-third of the areas experiencing a prolonged recovery lasting over 200 days. We emphasize the necessity of fine-scale knowledge in recovery assessments as aggregated statistics tend to largely underestimate the severity of hurricane impact and spatial heterogeneity of recovery. More importantly, we identify a prevailing recovery inequality across minority and disadvantaged populations, as well as a continued disproportionate allocation of hurricane assistance served as a key driver of exacerbating recovery inequality. Our study offers nuanced insights into the spatial heterogeneity of post-hurricane recovery that can inform strategic and equitable recovery efforts, as well as more effective hurricane relief programs and protocols.
{"title":"Nighttime lights reveal substantial spatial heterogeneity and inequality in post-hurricane recovery","authors":"Qiming Zheng , Yiwen Zeng , Yuyu Zhou , Zhuosen Wang , Te Mu , Qihao Weng","doi":"10.1016/j.rse.2025.114645","DOIUrl":"10.1016/j.rse.2025.114645","url":null,"abstract":"<div><div>While severe hurricanes continue to challenge the resilience of local communities, fine-scale knowledge of post-hurricane recovery remains scarce. Existing recovery tracking approaches mainly rely on aggregated metrics that would disguise the spatial heterogeneity in recovery patterns. Here, we present a spatiotemporally explicit investigation into the recovery of human activity after 10 recent severe hurricanes in the U.S., with daily nighttime light (NTL) time series images from NASA's Black Marble VIIRS NTL product suite. We utilized a Bayesian-based time series change detection model and temporal clustering algorithm to analyze the post-hurricane recovery of each built-up area pixel within 446 counties severely affected by the hurricanes. To investigate the potential inaccuracies stemming from assessments using aggregated statistics, we further compared the recovery pattern estimated at pixel scale with that estimated by aggregated NTL radiance at county and census tract scales. Last, we examined the inequality in post-hurricane recovery and how it related to socioeconomic factors and current hurricane assistance programs. Our analysis shows a 7-fold difference in the recovery duration of hurricane-affected built-up areas within a county, with one-third of the areas experiencing a prolonged recovery lasting over 200 days. We emphasize the necessity of fine-scale knowledge in recovery assessments as aggregated statistics tend to largely underestimate the severity of hurricane impact and spatial heterogeneity of recovery. More importantly, we identify a prevailing recovery inequality across minority and disadvantaged populations, as well as a continued disproportionate allocation of hurricane assistance served as a key driver of exacerbating recovery inequality. Our study offers nuanced insights into the spatial heterogeneity of post-hurricane recovery that can inform strategic and equitable recovery efforts, as well as more effective hurricane relief programs and protocols.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"319 ","pages":"Article 114645"},"PeriodicalIF":11.1,"publicationDate":"2025-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143388254","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-12DOI: 10.1016/j.rse.2024.114585
Yiqun Xie , Anh N. Nhu , Xiao-Peng Song , Xiaowei Jia , Sergii Skakun , Haijun Li , Zhihao Wang
Spatial variability has been one of the major challenges for large-area crop monitoring and classification with remote sensing. Recent works on deep learning have introduced spatial transformation methods to automatically partition a heterogeneous region into multiple homogeneous sub-regions during the training process. However, the framework is only designed for deep learning and is not available for other models, e.g., decision tree and random forest, which are frequently the models of choice in many crop mapping products. This paper develops a geo-aware random forest (Geo-RF) model to enable new capabilities to automatically recognize spatial variability during training, partition the space, and learn local models. Specifically, Geo-RF can capture spatial partitions with flexible shapes via an efficient bi-partitioning optimization algorithm. Geo-RF also automatically determines the number of partitions needed in a hierarchical manner via statistical tests and builds local RF models along the partitioning process to explicitly address spatial variability and improve classification quality. We used both synthetic and real-world data to evaluate the effectiveness of Geo-RF. First, through the controlled synthetic experiment, Geo-RF demonstrated the ability to capture the artificially-inserted true partition where a different relationship between the inputs and outputs is used. Second, we showed the improvements from Geo-RF using crop classification for five major crops over the contiguous US. The results demonstrated that Geo-RF is able to significantly improve classification performance in sub-regions that are otherwise compromised in a single RF model. For example, the partition around downstream Mississippi for soybean classification led to major improvements for about 0.10-0.25 in F1 scores in the area, and the score increased from 0.57 to 0.82 at certain locations. Similarly, for rice classification, the partition in Arkansas led to F1 scores increasing from 0.59 to 0.88 in local areas. In addition, we evaluated the models under different parameter settings, and the results showed that Geo-RF led to improvements over RF in the vast majority of scenarios (e.g., varying model complexity and training sizes). Computationally, Geo-RF took about one to three times more training time while its execution time during testing was similar to that of RF. Overall, Geo-RF showed the ability to automatically address spatial variability via partitioning optimization, which is an important skill for improving crop classification over heterogeneous geographic areas at large scale. Future research can explore the use of Geo-RF for other geographic regions and applications, interpretable methods to understand the data-driven partitioning, and new designs to further enhance the computational efficiency.
{"title":"Accounting for spatial variability with geo-aware random forest: A case study for US major crop mapping","authors":"Yiqun Xie , Anh N. Nhu , Xiao-Peng Song , Xiaowei Jia , Sergii Skakun , Haijun Li , Zhihao Wang","doi":"10.1016/j.rse.2024.114585","DOIUrl":"10.1016/j.rse.2024.114585","url":null,"abstract":"<div><div>Spatial variability has been one of the major challenges for large-area crop monitoring and classification with remote sensing. Recent works on deep learning have introduced spatial transformation methods to automatically partition a heterogeneous region into multiple homogeneous sub-regions during the training process. However, the framework is only designed for deep learning and is not available for other models, e.g., decision tree and random forest, which are frequently the models of choice in many crop mapping products. This paper develops a geo-aware random forest (Geo-RF) model to enable new capabilities to automatically recognize spatial variability during training, partition the space, and learn local models. Specifically, Geo-RF can capture spatial partitions with flexible shapes via an efficient bi-partitioning optimization algorithm. Geo-RF also automatically determines the number of partitions needed in a hierarchical manner via statistical tests and builds local RF models along the partitioning process to explicitly address spatial variability and improve classification quality. We used both synthetic and real-world data to evaluate the effectiveness of Geo-RF. First, through the controlled synthetic experiment, Geo-RF demonstrated the ability to capture the artificially-inserted true partition where a different relationship between the inputs and outputs is used. Second, we showed the improvements from Geo-RF using crop classification for five major crops over the contiguous US. The results demonstrated that Geo-RF is able to significantly improve classification performance in sub-regions that are otherwise compromised in a single RF model. For example, the partition around downstream Mississippi for soybean classification led to major improvements for about 0.10-0.25 in F1 scores in the area, and the score increased from 0.57 to 0.82 at certain locations. Similarly, for rice classification, the partition in Arkansas led to F1 scores increasing from 0.59 to 0.88 in local areas. In addition, we evaluated the models under different parameter settings, and the results showed that Geo-RF led to improvements over RF in the vast majority of scenarios (e.g., varying model complexity and training sizes). Computationally, Geo-RF took about one to three times more training time while its execution time during testing was similar to that of RF. Overall, Geo-RF showed the ability to automatically address spatial variability via partitioning optimization, which is an important skill for improving crop classification over heterogeneous geographic areas at large scale. Future research can explore the use of Geo-RF for other geographic regions and applications, interpretable methods to understand the data-driven partitioning, and new designs to further enhance the computational efficiency.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"319 ","pages":"Article 114585"},"PeriodicalIF":11.1,"publicationDate":"2025-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143393807","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-11DOI: 10.1016/j.rse.2025.114619
Chenqian Tang , Chong Shi , Husi Letu , Shuai Yin , Teruyuki Nakajima , Miho Sekiguchi , Jian Xu , Mengjie Zhao , Run Ma , Wenwu Wang
The aerosol optical thickness (AOT) and fine-mode fraction (FMF) are crucial to understanding the radiative and environmental effects of aerosols. However, accurately retrieving these properties simultaneously from monodirectional multispectral satellite data remains challenging. Inversion algorithms based on lookup tables typically leverage information from only two or three channels, resulting in limited retrieval parameters. Although optimal estimation methods can enhance the utilization of multispectral information, they are mostly constrained by fixed aerosol types and have higher computational overhead due to the multiple iterations. To achieve real-time, high-precision, and simultaneous retrieval of the AOT and FMF for geostationary satellites with high-frequency observation, we propose a novel hybrid algorithm, AIRTrans, for the Himawari-8/AHI by integrating radiative transfer (RT) and transfer learning (TL) approaches. Specifically, RT is used to construct a simulation dataset that covers multiple aerosol types and surface conditions corresponding to the simulated multispectral observation, which pre-trains an artificial neural network model. The TL strategy is then employed to fine-tune this model using in situ data, enhancing its representativeness in real scenarios. AIRTrans performs direct retrieval using satellite observations and surface reflectance constructed via the second minimum reflectance method but considering background AOT. Results indicate that the AIRTrans-retrieved AOT and FMF are generally more consistent with ground-based observations from AERONET than official AHI products, through three years of independent validation across the full-disk region. Specifically, AIRTrans achieves retrieval with RMSEs of 0.132 and 0.146 for AOT and FMF, respectively, compared to 0.216 and 0.284 for AHI products. AIRTrans shows a remarkable improvement on FMF, particularly in addressing the significant underestimation of the AHI products at over 90 % of individual sites. The performance of AIRTrans during two severe aerosol pollution events (intense dust storms and haze) further demonstrates its robust ability to capture spatiotemporal variations of AOT and FMF simultaneously.
{"title":"Development of a hybrid algorithm for the simultaneous retrieval of aerosol optical thickness and fine-mode fraction from multispectral satellite observation combining radiative transfer and transfer learning approaches","authors":"Chenqian Tang , Chong Shi , Husi Letu , Shuai Yin , Teruyuki Nakajima , Miho Sekiguchi , Jian Xu , Mengjie Zhao , Run Ma , Wenwu Wang","doi":"10.1016/j.rse.2025.114619","DOIUrl":"10.1016/j.rse.2025.114619","url":null,"abstract":"<div><div>The aerosol optical thickness (AOT) and fine-mode fraction (FMF) are crucial to understanding the radiative and environmental effects of aerosols. However, accurately retrieving these properties simultaneously from monodirectional multispectral satellite data remains challenging. Inversion algorithms based on lookup tables typically leverage information from only two or three channels, resulting in limited retrieval parameters. Although optimal estimation methods can enhance the utilization of multispectral information, they are mostly constrained by fixed aerosol types and have higher computational overhead due to the multiple iterations. To achieve real-time, high-precision, and simultaneous retrieval of the AOT and FMF for geostationary satellites with high-frequency observation, we propose a novel hybrid algorithm, AIRTrans, for the Himawari-8/AHI by integrating radiative transfer (RT) and transfer learning (TL) approaches. Specifically, RT is used to construct a simulation dataset that covers multiple aerosol types and surface conditions corresponding to the simulated multispectral observation, which pre-trains an artificial neural network model. The TL strategy is then employed to fine-tune this model using in situ data, enhancing its representativeness in real scenarios. AIRTrans performs direct retrieval using satellite observations and surface reflectance constructed via the second minimum reflectance method but considering background AOT. Results indicate that the AIRTrans-retrieved AOT and FMF are generally more consistent with ground-based observations from AERONET than official AHI products, through three years of independent validation across the full-disk region. Specifically, AIRTrans achieves retrieval with RMSEs of 0.132 and 0.146 for AOT and FMF, respectively, compared to 0.216 and 0.284 for AHI products. AIRTrans shows a remarkable improvement on FMF, particularly in addressing the significant underestimation of the AHI products at over 90 % of individual sites. The performance of AIRTrans during two severe aerosol pollution events (intense dust storms and haze) further demonstrates its robust ability to capture spatiotemporal variations of AOT and FMF simultaneously.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"319 ","pages":"Article 114619"},"PeriodicalIF":11.1,"publicationDate":"2025-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143385117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-11DOI: 10.1016/j.rse.2025.114644
Esmaeel Adrah, Jesse Pan Wong, He Yin
Mapping tree crops is essential for resource management and supporting local livelihoods and ecosystem services. However, tree crops are often overlooked or misclassified in regional and global cropland maps. Employing multi-sensor imagery presents new opportunities for mapping tree crops by providing additional observations and distinct characteristics. Nevertheless, challenges regarding the scarcity of ground references and the lack of robust approaches to integrating multi-sensor imagery pose obstacles to the production of reliable tree crop maps. Herein, we evaluate the integration of the Global Ecosystem Dynamic Investigation (GEDI) LiDAR with Sentinel-2 and Sentinel-1 to facilitate tree crops mapping in the eastern Mediterranean region (including Syria, part of Turkey, and Jordan) and southern France. First, we systematically filtered the GEDI relative heights (RH) metrics and above-ground biomass density (AGBD) using ancillary data (e.g., cloud, topography, land cover) and applied spatial constraints to combine the high-quality GEDI shots with Sentinel-2 normalized difference vegetation index (NDVI) and Sentinel-1 VV and VH backscatter. Second, we used Time-Weighted Dynamic Time Warping (TW-DTW) and random forest (RF) models to test the classification performance using different combinations of input features at the GEDI footprint level. Finally, we used GEDI footprint level classification as training samples to train RF classifiers to generate wall-to-wall tree crops maps using a combined Sentinel-2 and Sentinel-1 imagery composite. We found that, at the GEDI footprint level, using GEDI variables only, we achieved an F1 score of 73–78 % for tree crops, approximately 4–10 % higher compared to that using Sentinel-2 and Sentinel-1 imagery for classification. However, by combining GEDI with Sentinel-2 and Sentinel-1 imagery, we achieved the highest accuracy (F1 score: 73–86 %) at the GEDI footprint level classification. The mapping accuracy of our wall-to-wall map varied across different agroclimatic zones with higher accuracy in dryer regions reaching up to 91 % and lowest at 69 %. Our finding demonstrates the value of using structural information from the GEDI data to map tree crops across different agroclimatic zones. Our study emphasizes the importance of tree crops in regional maps and offers insights to support the efforts to integrate data from multiple remote sensing platforms.
{"title":"Integrating GEDI, Sentinel-2, and Sentinel-1 imagery for tree crops mapping","authors":"Esmaeel Adrah, Jesse Pan Wong, He Yin","doi":"10.1016/j.rse.2025.114644","DOIUrl":"10.1016/j.rse.2025.114644","url":null,"abstract":"<div><div>Mapping tree crops is essential for resource management and supporting local livelihoods and ecosystem services. However, tree crops are often overlooked or misclassified in regional and global cropland maps. Employing multi-sensor imagery presents new opportunities for mapping tree crops by providing additional observations and distinct characteristics. Nevertheless, challenges regarding the scarcity of ground references and the lack of robust approaches to integrating multi-sensor imagery pose obstacles to the production of reliable tree crop maps. Herein, we evaluate the integration of the Global Ecosystem Dynamic Investigation (GEDI) LiDAR with Sentinel-2 and Sentinel-1 to facilitate tree crops mapping in the eastern Mediterranean region (including Syria, part of Turkey, and Jordan) and southern France. First, we systematically filtered the GEDI relative heights (RH) metrics and above-ground biomass density (AGBD) using ancillary data (e.g., cloud, topography, land cover) and applied spatial constraints to combine the high-quality GEDI shots with Sentinel-2 normalized difference vegetation index (NDVI) and Sentinel-1 VV and VH backscatter. Second, we used Time-Weighted Dynamic Time Warping (TW-DTW) and random forest (RF) models to test the classification performance using different combinations of input features at the GEDI footprint level. Finally, we used GEDI footprint level classification as training samples to train RF classifiers to generate wall-to-wall tree crops maps using a combined Sentinel-2 and Sentinel-1 imagery composite. We found that, at the GEDI footprint level, using GEDI variables only, we achieved an F1 score of 73–78 % for tree crops, approximately 4–10 % higher compared to that using Sentinel-2 and Sentinel-1 imagery for classification. However, by combining GEDI with Sentinel-2 and Sentinel-1 imagery, we achieved the highest accuracy (F1 score: 73–86 %) at the GEDI footprint level classification. The mapping accuracy of our wall-to-wall map varied across different agroclimatic zones with higher accuracy in dryer regions reaching up to 91 % and lowest at 69 %. Our finding demonstrates the value of using structural information from the GEDI data to map tree crops across different agroclimatic zones. Our study emphasizes the importance of tree crops in regional maps and offers insights to support the efforts to integrate data from multiple remote sensing platforms.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"319 ","pages":"Article 114644"},"PeriodicalIF":11.1,"publicationDate":"2025-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143379285","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-09DOI: 10.1016/j.rse.2025.114640
Qingyan Meng , Shize Chen , Linlin Zhang , Xiaolin Zhu , Yeping Zhang , Peter M. Atkinson
Land surface temperature (LST) data are crucial for global climate change research. While remote sensing data serve as a key source for LST, single-source sensor data often lack spatiotemporal continuity due to long satellite revisit intervals and cloud cover. Spatiotemporal fusion, which combines the strengths of multiple sources, can increase the available information. However, most current spatiotemporal fusion methods are designed for local-scale applications. This research proposes the Global Spatiotemporal Fusion Model (GLOSTFM) to generate global LST products. GLOSTFM, built on image pyramid principles, addresses the computational and complexity challenges of global-scale spatiotemporal fusion. Moreover, the model utilizes data from the novel Fengyun-3D satellite, which has a daily revisit capability and provides LST products separately derived from its thermal infrared (MERSI, 1 km) and microwave (MWRI, 25 km) sensors. By leveraging the cloud-penetrating capabilities of the microwave data to compensate for missing information, GLOSTFM increases the available information and reduces observational uncertainties. The results showcase high processing efficiency and enhanced spatiotemporal continuity, with an average RMSE of 2.874 K and an excellent R2 of 0.980. The utility of the GLOSTFM model for monitoring urban heat island effects in Beijing was explored to illustrate one application among a broad range of potential applications of the proposed GLOSTFM that require global data on LST across the Earth's surface.
{"title":"GLOSTFM: A global spatiotemporal fusion model integrating multi-source satellite observations to enhance land surface temperature resolution","authors":"Qingyan Meng , Shize Chen , Linlin Zhang , Xiaolin Zhu , Yeping Zhang , Peter M. Atkinson","doi":"10.1016/j.rse.2025.114640","DOIUrl":"10.1016/j.rse.2025.114640","url":null,"abstract":"<div><div>Land surface temperature (LST) data are crucial for global climate change research. While remote sensing data serve as a key source for LST, single-source sensor data often lack spatiotemporal continuity due to long satellite revisit intervals and cloud cover. Spatiotemporal fusion, which combines the strengths of multiple sources, can increase the available information. However, most current spatiotemporal fusion methods are designed for local-scale applications. This research proposes the Global Spatiotemporal Fusion Model (GLOSTFM) to generate global LST products. GLOSTFM, built on image pyramid principles, addresses the computational and complexity challenges of global-scale spatiotemporal fusion. Moreover, the model utilizes data from the novel Fengyun-3D satellite, which has a daily revisit capability and provides LST products separately derived from its thermal infrared (MERSI, 1 km) and microwave (MWRI, 25 km) sensors. By leveraging the cloud-penetrating capabilities of the microwave data to compensate for missing information, GLOSTFM increases the available information and reduces observational uncertainties. The results showcase high processing efficiency and enhanced spatiotemporal continuity, with an average RMSE of 2.874 K and an excellent <em>R</em><sup>2</sup> of 0.980. The utility of the GLOSTFM model for monitoring urban heat island effects in Beijing was explored to illustrate one application among a broad range of potential applications of the proposed GLOSTFM that require global data on LST across the Earth's surface.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"319 ","pages":"Article 114640"},"PeriodicalIF":11.1,"publicationDate":"2025-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143375582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-08DOI: 10.1016/j.rse.2025.114637
Chan Li , Penghai Wu , Si-Bo Duan , Yixuan Jia , Shuai Sun , Chunxiang Shi , Zhixiang Yin , Huifang Li , Huanfeng Shen
Due to the great advancements in land surface models (LSMs), integrating data from thermal infrared (TIR) and LSMs is a promising way for obtaining gapless all-weather land surface temperature (LST). However, the differences of spatial resolution and discrepancy of data acquisition ways between TIR LST and model-simulated LST usually brought great challenges to traditional methods in terms of accuracy and texture details. This study proposes a low-resolution filling then super-resolution reconstruction (LFSR) framework for generating gapless all-weather LST using Moderate Resolution Imaging Spectroradiometer (MODIS) LST and China Meteorological Administration Land Data Assimilation System (CLDAS) LST. For the LFSR, a multi-source multi-temporal low-resolution filling (MSMTLF) network is first designed to alleviate the discrepancy of data acquisition ways between the MODIS LST and CLDAS LST, and generate gapless low-resolution degraded LSTs. A multi-scale multi-temporal super-resolution reconstruction (MSMTSR) network is then used to reconstruct the gapless low-resolution degraded LSTs into gapless high-resolution MODIS-like LSTs with rich-texture, which is mainly used to deal with resolution differences between the two LSTs. The experiments suggested that the LFSR achieved satisfactory results, and the maximal RMSE is less 2.5 K in the simulated experiments. When validated against the in-situ LST data under clear and cloudy skies, the small difference of the overall average bias (−0.91 K for clear skies VS -0.88 K for cloudy skies) and overall average RMSE (4.15 K for clear skies VS 5.68 K for cloudy skies) were obtained. Compared with results from the different input data, the other strategies and the other methods, the generated gapless all-weather MODIS-like LSTs from the LFSR were closer to the actual labels or have better consistency and spatial details. These results indicated the LFSR achieves impressive performance for fusing MODIS and CLDAS data. The LFSR actually provides a new framework for fusing TIR LST and simulation-based LST with considerable data inconsistency, and has the potential for generating gapless all-weather TIR LST records.
{"title":"LFSR: Low-resolution Filling then Super-resolution Reconstruction framework for gapless all-weather MODIS-like land surface temperature generation","authors":"Chan Li , Penghai Wu , Si-Bo Duan , Yixuan Jia , Shuai Sun , Chunxiang Shi , Zhixiang Yin , Huifang Li , Huanfeng Shen","doi":"10.1016/j.rse.2025.114637","DOIUrl":"10.1016/j.rse.2025.114637","url":null,"abstract":"<div><div>Due to the great advancements in land surface models (LSMs), integrating data from thermal infrared (TIR) and LSMs is a promising way for obtaining gapless all-weather land surface temperature (LST). However, the differences of spatial resolution and discrepancy of data acquisition ways between TIR LST and model-simulated LST usually brought great challenges to traditional methods in terms of accuracy and texture details. This study proposes a low-resolution filling then super-resolution reconstruction (LFSR) framework for generating gapless all-weather LST using Moderate Resolution Imaging Spectroradiometer (MODIS) LST and China Meteorological Administration Land Data Assimilation System (CLDAS) LST. For the LFSR, a multi-source multi-temporal low-resolution filling (MSMTLF) network is first designed to alleviate the discrepancy of data acquisition ways between the MODIS LST and CLDAS LST, and generate gapless low-resolution degraded LSTs. A multi-scale multi-temporal super-resolution reconstruction (MSMTSR) network is then used to reconstruct the gapless low-resolution degraded LSTs into gapless high-resolution MODIS-like LSTs with rich-texture, which is mainly used to deal with resolution differences between the two LSTs. The experiments suggested that the LFSR achieved satisfactory results, and the maximal RMSE is less 2.5 K in the simulated experiments. When validated against the in-situ LST data under clear and cloudy skies, the small difference of the overall average bias (−0.91 K for clear skies VS -0.88 K for cloudy skies) and overall average RMSE (4.15 K for clear skies VS 5.68 K for cloudy skies) were obtained. Compared with results from the different input data, the other strategies and the other methods, the generated gapless all-weather MODIS-like LSTs from the LFSR were closer to the actual labels or have better consistency and spatial details. These results indicated the LFSR achieves impressive performance for fusing MODIS and CLDAS data. The LFSR actually provides a new framework for fusing TIR LST and simulation-based LST with considerable data inconsistency, and has the potential for generating gapless all-weather TIR LST records.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"319 ","pages":"Article 114637"},"PeriodicalIF":11.1,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143371518","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号