Pub Date : 2025-10-14DOI: 10.1016/j.srs.2025.100314
Spencer G. Shields , Nicholas C. Coops , Alexis Achim , Richard C. Hamelin , Christopher Mulverhill
Satellite remote sensing has been a cornerstone of forest monitoring, enabling the observation of extensive areas at regular intervals. In 2014, Planet Labs introduced PlanetScope, a constellation of Earth observation CubeSats capable of delivering near-daily optical data at a 3 m resolution across the globe. The unique combination of high temporal and spatial resolution, along with comprehensive coverage, positions PlanetScope as a valuable tool for a wide range of forestry applications. This systematic literature review explores the diverse applications of PlanetScope in forestry research, detailing the ecosystems studied, the spatial and temporal characteristics of the datasets, analytical methods employed, and integration with other remote sensing technologies. We comment on potential strengths and weaknesses of the available datasets, compare models developed using PlanetScope with those derived from other remote sensing data sources, identify key areas for future research, and finally provide recommendations and considerations for prospective users of PlanetScope data.
{"title":"A review of PlanetScope CubeSats for forest monitoring","authors":"Spencer G. Shields , Nicholas C. Coops , Alexis Achim , Richard C. Hamelin , Christopher Mulverhill","doi":"10.1016/j.srs.2025.100314","DOIUrl":"10.1016/j.srs.2025.100314","url":null,"abstract":"<div><div>Satellite remote sensing has been a cornerstone of forest monitoring, enabling the observation of extensive areas at regular intervals. In 2014, Planet Labs introduced PlanetScope, a constellation of Earth observation CubeSats capable of delivering near-daily optical data at a 3 m resolution across the globe. The unique combination of high temporal and spatial resolution, along with comprehensive coverage, positions PlanetScope as a valuable tool for a wide range of forestry applications. This systematic literature review explores the diverse applications of PlanetScope in forestry research, detailing the ecosystems studied, the spatial and temporal characteristics of the datasets, analytical methods employed, and integration with other remote sensing technologies. We comment on potential strengths and weaknesses of the available datasets, compare models developed using PlanetScope with those derived from other remote sensing data sources, identify key areas for future research, and finally provide recommendations and considerations for prospective users of PlanetScope data.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100314"},"PeriodicalIF":5.2,"publicationDate":"2025-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320138","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-11DOI: 10.1016/j.srs.2025.100297
Narendra Nelli , Diana Francis , Charfeddine Cherif , Ricardo Fonseca , Hosni Ghedira
Fog significantly reduces visibility, impacting transportation and safety, especially in United Arab Emirates (UAE) during winter months. This study develops a machine learning (ML) approach for automated fog detection and masking from near real-time SEVIRI Satellite observations. We evaluate six ML models across four training strategies: (1) supervised training using SEVIRI nighttime microphysics Red-Green-Blue (RGB) pixels with Meteorological Aerodrome Reports (METAR) station labels; (2) as (1) but adding the three infrared channels; (3) k-means labels derived from Night Microphysics RGB; and (4) a fusion of station-labeled and k-means-labeled data. Among the models, the eXtreme Gradient Boosting (XGBoost) performs best overall. Using the same fog events analyzed by Weston and Temimi (2020), the fusion approach (Approach 4) with XGBoost more sharply delineates fog boundaries, accurately captures “fog holes”, and reduces false alarms and missed detections—including during marginal/light-mist episodes—relative to the thresholding method, with notable improvements over inland deserts and along the coast. At Abu Dhabi, station-trained models achieve a Probability of Detection of ∼0.73 with a False Alarm Ratio of ∼0.11; the fusion approach maintains strong detection skill with competitive false-alarm rates while improving spatial coherence. Regional case studies over Qatar and Saudi Arabia demonstrate that the trained model generalizes across the Arabian Peninsula. The workflow executes in seconds and relies only on three infrared channels, avoiding auxiliary reanalysis inputs and supporting near-real-time operations. These results show that combining complementary labels from stations and clustering substantially enhances satellite-based fog masking, providing a practical pathway for operational monitoring and a foundation for short-term nowcasting in arid environments.
{"title":"Automated night-time fog detection and masking using machine learning from near real-time satellite observations","authors":"Narendra Nelli , Diana Francis , Charfeddine Cherif , Ricardo Fonseca , Hosni Ghedira","doi":"10.1016/j.srs.2025.100297","DOIUrl":"10.1016/j.srs.2025.100297","url":null,"abstract":"<div><div>Fog significantly reduces visibility, impacting transportation and safety, especially in United Arab Emirates (UAE) during winter months. This study develops a machine learning (ML) approach for automated fog detection and masking from near real-time SEVIRI Satellite observations. We evaluate six ML models across four training strategies: (1) supervised training using SEVIRI nighttime microphysics Red-Green-Blue (RGB) pixels with Meteorological Aerodrome Reports (METAR) station labels; (2) as (1) but adding the three infrared channels; (3) k-means labels derived from Night Microphysics RGB; and (4) a fusion of station-labeled and k-means-labeled data. Among the models, the eXtreme Gradient Boosting (XGBoost) performs best overall. Using the same fog events analyzed by Weston and Temimi (2020), the fusion approach (Approach 4) with XGBoost more sharply delineates fog boundaries, accurately captures “fog holes”, and reduces false alarms and missed detections—including during marginal/light-mist episodes—relative to the thresholding method, with notable improvements over inland deserts and along the coast. At Abu Dhabi, station-trained models achieve a Probability of Detection of ∼0.73 with a False Alarm Ratio of ∼0.11; the fusion approach maintains strong detection skill with competitive false-alarm rates while improving spatial coherence. Regional case studies over Qatar and Saudi Arabia demonstrate that the trained model generalizes across the Arabian Peninsula. The workflow executes in seconds and relies only on three infrared channels, avoiding auxiliary reanalysis inputs and supporting near-real-time operations. These results show that combining complementary labels from stations and clustering substantially enhances satellite-based fog masking, providing a practical pathway for operational monitoring and a foundation for short-term nowcasting in arid environments.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100297"},"PeriodicalIF":5.2,"publicationDate":"2025-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320240","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-11DOI: 10.1016/j.srs.2025.100310
Valtteri Soininen, Xiaowei Yu, Matti Hyyppä, Juha Hyyppä
Optimising bioeconomy-related ecosystem services requires more detailed forest information. One option is to go towards individual-tree-based precision forestry. Although many national airborne laser scanning (ALS) programmes can detect individual trees and predict their attributes, training the necessary models remains a challenge. Ideally, each site covered with ALS would have its own reference data, but this requires measuring millions of trees nationwide per every country-level ALS scan. Instead of always collecting site-specific training data, an alternative is to transfer individual tree models from other sites. This approach relies on good model transferability that ensures accurate and realistic estimates. This study tested the transferability of individual tree diameter at breast height (DBH) and stem volume models combining national laser scanning data and the random forest method in Finland. The model that was trained with coordinate information benefitted from training data that were collected within the range of 500 km. The root mean squared error (RMSE) and bias magnitude of the model that was trained without the coordinate information started to increase after 300 km, but the increase could be cancelled by using coordinates as predictor features. Furthermore, when the models were evaluated outside the area for which they were trained, the errors increased at a rate between 0.27–0.28 cm/100 km in RMSE in DBH prediction and 8.08–13.18 dm/100 km in stem volume prediction. The same values for bias magnitude were 0.39–0.42 cm/100 km in DBH prediction and 8.32–12.01 dm/100 km in stem volume prediction. The increase in training set size slightly slowed the rate. Quick convergence of RMSE was observed in a test in which small amounts of target site data were included in the training data. The same was also observed for bias magnitude, although the results were not as good as with RMSE.
{"title":"Transferability of country-wide airborne laser scanning-based models for individual-tree attributes","authors":"Valtteri Soininen, Xiaowei Yu, Matti Hyyppä, Juha Hyyppä","doi":"10.1016/j.srs.2025.100310","DOIUrl":"10.1016/j.srs.2025.100310","url":null,"abstract":"<div><div>Optimising bioeconomy-related ecosystem services requires more detailed forest information. One option is to go towards individual-tree-based precision forestry. Although many national airborne laser scanning (ALS) programmes can detect individual trees and predict their attributes, training the necessary models remains a challenge. Ideally, each site covered with ALS would have its own reference data, but this requires measuring millions of trees nationwide per every country-level ALS scan. Instead of always collecting site-specific training data, an alternative is to transfer individual tree models from other sites. This approach relies on good model transferability that ensures accurate and realistic estimates. This study tested the transferability of individual tree diameter at breast height (DBH) and stem volume models combining national laser scanning data and the random forest method in Finland. The model that was trained with coordinate information benefitted from training data that were collected within the range of 500 km. The root mean squared error (RMSE) and bias magnitude of the model that was trained without the coordinate information started to increase after 300 km, but the increase could be cancelled by using coordinates as predictor features. Furthermore, when the models were evaluated outside the area for which they were trained, the errors increased at a rate between 0.27–0.28 cm/100 km in RMSE in DBH prediction and 8.08–13.18 dm<span><math><msup><mrow></mrow><mrow><mn>3</mn></mrow></msup></math></span>/100 km in stem volume prediction. The same values for bias magnitude were 0.39–0.42 cm/100 km in DBH prediction and 8.32–12.01 dm<span><math><msup><mrow></mrow><mrow><mn>3</mn></mrow></msup></math></span>/100 km in stem volume prediction. The increase in training set size slightly slowed the rate. Quick convergence of RMSE was observed in a test in which small amounts of target site data were included in the training data. The same was also observed for bias magnitude, although the results were not as good as with RMSE.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100310"},"PeriodicalIF":5.2,"publicationDate":"2025-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-10DOI: 10.1016/j.srs.2025.100311
Maud Henrion , Yanfei Li , Kaijun Wu , François Jonard , Sophie Opfergelt , Veerle Vanacker , Kristof Van Oost , Sébastien Lambot
Peatlands are important ecosystems, providing essential ecological services, such as carbon storage and biodiversity support. However, they are endangered by degradation due to land use and climate change. Their moisture status is a key factor, as it substantially impacts carbon storage and decomposition. Therefore, it is essential to accurately characterize, map, and monitor peatland moisture. This study assessed the potential of drone-borne Ground-penetrating radar (GPR), combined with full-wave inversion, to study peatland moisture. We applied this technique to a peatland in the Belgian Hautes Fagnes previously degraded by reforestation. We conducted GPR measurements over 4.5 ha for one and a half years, producing 19 different peatland root-zone moisture maps at a 5 m resolution. Our results demonstrate that this method can track moisture changes over the study site, with an overall temporal correlation of 0.71 with ground-based moisture sensors, but is less reliable in nearly saturated areas. The spatial correlation with ground-based probes is lower (0.23), due to the high micro-variability of moisture and the use of kriging interpolation to generate maps, resulting in a spatial mismatch as GPR measurements were not collected directly above the probes. We applied statistical clustering techniques on the moisture maps to delineate homogeneous moisture classes that align well with other specific site characteristics (peat depth, vegetation types, Normalized Difference Water Index and surface temperature). This technique shows potential for planning and monitoring peatland restoration efforts and provides a new and valuable approach for peatland moisture studies to complement existing satellite- and other drone-based methods.
{"title":"Drone-borne ground-penetrating radar reveals spatiotemporal moisture dynamics in peatland root zones","authors":"Maud Henrion , Yanfei Li , Kaijun Wu , François Jonard , Sophie Opfergelt , Veerle Vanacker , Kristof Van Oost , Sébastien Lambot","doi":"10.1016/j.srs.2025.100311","DOIUrl":"10.1016/j.srs.2025.100311","url":null,"abstract":"<div><div>Peatlands are important ecosystems, providing essential ecological services, such as carbon storage and biodiversity support. However, they are endangered by degradation due to land use and climate change. Their moisture status is a key factor, as it substantially impacts carbon storage and decomposition. Therefore, it is essential to accurately characterize, map, and monitor peatland moisture. This study assessed the potential of drone-borne Ground-penetrating radar (GPR), combined with full-wave inversion, to study peatland moisture. We applied this technique to a peatland in the Belgian Hautes Fagnes previously degraded by reforestation. We conducted GPR measurements over 4.5 ha for one and a half years, producing 19 different peatland root-zone moisture maps at a 5 m resolution. Our results demonstrate that this method can track moisture changes over the study site, with an overall temporal correlation of 0.71 with ground-based moisture sensors, but is less reliable in nearly saturated areas. The spatial correlation with ground-based probes is lower (0.23), due to the high micro-variability of moisture and the use of kriging interpolation to generate maps, resulting in a spatial mismatch as GPR measurements were not collected directly above the probes. We applied statistical clustering techniques on the moisture maps to delineate homogeneous moisture classes that align well with other specific site characteristics (peat depth, vegetation types, Normalized Difference Water Index and surface temperature). This technique shows potential for planning and monitoring peatland restoration efforts and provides a new and valuable approach for peatland moisture studies to complement existing satellite- and other drone-based methods.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100311"},"PeriodicalIF":5.2,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320127","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-09DOI: 10.1016/j.srs.2025.100292
Xiaohong Gao , Zhuoran Yan , Lun Bao , Xuan Li , Li Gao , Lingxue Yu
Wetlands influence local land surface temperature (LST) via biogeophysical processes, nevertheless, their temperature regulation under different moisture conditions remain unclear. This study quantified the spatial heterogeneity and drivers of wetland-induced temperature effects in the Amur River Basin using multi-year averaged LST data (2003–2022) within a space for time paired comparison framework. Our findings demonstrate that LST regulation of wetlands showed distinct diurnal asymmetry. During the growing season (May–September), natural wetlands induce substantial daytime cooling (−1.20 ± 0.77 K) and slight nighttime warming (0.05 ± 0.63 K). Artificial paddy wetlands show similar patterns but stronger nighttime warming (0.64 ± 0.41 K), reducing net cooling. Both wetland types absorb more solar radiation than adjacent drylands (natural: −0.67 % ± 0.88 %; artificial: −0.60 % ± 0.65 %) and dissipate energy primarily through enhanced evapotranspiration in early growing season (May–July) (0.13 ± 0.30 mm/d; 0.02 ± 0.22 mm/d). Nighttime heat release from water and soil partially offsets daytime cooling. Natural wetlands maintain superior cooling via stable non-radiative processes, with synchronized nighttime cooling in humid regions and compensatory nighttime warming in arid regions, ensuring consistent temperature reduction across hydrological gradients. Conversely, artificial paddy fields in semi-arid areas achieve strong cooling (−0.82 ± 0.34 K) through dual-phase evapotranspiration (0.128 ± 0.263 mm/d; 0.003 ± 0.236 mm/d). In humid regions, nighttime heat storage and release exceed daytime cooling, causing marginal warming. Thus, the cooling effect of artificial paddy fields is governed by moisture, evapotranspiration, and inundation. These results highlight that the artificial paddies cannot fully replace natural wetlands in climate regulation, underscoring the need to prioritize natural wetland conservation and restoration in land-use and climate strategies.
{"title":"Satellite observation reveals wetland-induced local cooling moderated by regional climate gradients","authors":"Xiaohong Gao , Zhuoran Yan , Lun Bao , Xuan Li , Li Gao , Lingxue Yu","doi":"10.1016/j.srs.2025.100292","DOIUrl":"10.1016/j.srs.2025.100292","url":null,"abstract":"<div><div>Wetlands influence local land surface temperature (LST) via biogeophysical processes, nevertheless, their temperature regulation under different moisture conditions remain unclear. This study quantified the spatial heterogeneity and drivers of wetland-induced temperature effects in the Amur River Basin using multi-year averaged LST data (2003–2022) within a space for time paired comparison framework. Our findings demonstrate that LST regulation of wetlands showed distinct diurnal asymmetry. During the growing season (May–September), natural wetlands induce substantial daytime cooling (−1.20 ± 0.77 K) and slight nighttime warming (0.05 ± 0.63 K). Artificial paddy wetlands show similar patterns but stronger nighttime warming (0.64 ± 0.41 K), reducing net cooling. Both wetland types absorb more solar radiation than adjacent drylands (natural: −0.67 % ± 0.88 %; artificial: −0.60 % ± 0.65 %) and dissipate energy primarily through enhanced evapotranspiration in early growing season (May–July) (0.13 ± 0.30 mm/d; 0.02 ± 0.22 mm/d). Nighttime heat release from water and soil partially offsets daytime cooling. Natural wetlands maintain superior cooling via stable non-radiative processes, with synchronized nighttime cooling in humid regions and compensatory nighttime warming in arid regions, ensuring consistent temperature reduction across hydrological gradients. Conversely, artificial paddy fields in semi-arid areas achieve strong cooling (−0.82 ± 0.34 K) through dual-phase evapotranspiration (0.128 ± 0.263 mm/d; 0.003 ± 0.236 mm/d). In humid regions, nighttime heat storage and release exceed daytime cooling, causing marginal warming. Thus, the cooling effect of artificial paddy fields is governed by moisture, evapotranspiration, and inundation. These results highlight that the artificial paddies cannot fully replace natural wetlands in climate regulation, underscoring the need to prioritize natural wetland conservation and restoration in land-use and climate strategies.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100292"},"PeriodicalIF":5.2,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145265185","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-09DOI: 10.1016/j.srs.2025.100308
Muyan Han , Ji Chai
Accurate and timely corn mapping at regional scales supports effective agricultural management and policy development. However, traditional data-driven approaches rely heavily on extensive ground-truth samples and have limited applicability in data-scarce areas. And there is still a lack of simple and practical approaches that require few or no field observations for corn mapping. Therefore, this study presents a novel framework that integrates an improved Dual-peak Canopy Nitrogen Index (DCNI) with Green Chromatic Coordinate (GCC) analysis. The refined DCNI captures nitrogen-related spectral dynamics to distinguish corn from spectrally similar crops such as soybean and sorghum during peak growth stages. The GCC analysis identifies the optimal mapping window, mitigating the effects of cloud cover and temporal data gaps for regional corn mapping. We applied this improved index to three selected but representative corn growing regions: Nenjiang in China, Pocahontas in the United States, and Mayenne in France. Using the optimized DCNI thresholds, binary classification was performed to reduce both commission and omission errors. In the pixel-level validation for 2021, our DCNI index achieved better overall accuracy and F1 scores compared to the classical random forest model and its variants. Interannual tests from 2020 to 2022 also showed stable performance and strong agreement with official statistics, with all coefficients of determination R2 above 0.96. We also investigate uncertainties arising from data interpolation, limited field samples, and mixed cropping patterns, and propose the integration of higher-frequency satellite observations and multisource data fusion to improve early season monitoring and broaden large-scale applicability. The proposed framework requires minimal crop samples and computational resources, providing a simple, practical alternative for regional corn mapping with robust transferability.
{"title":"An efficient and transferable remote sensing spectral index for regional corn mapping","authors":"Muyan Han , Ji Chai","doi":"10.1016/j.srs.2025.100308","DOIUrl":"10.1016/j.srs.2025.100308","url":null,"abstract":"<div><div>Accurate and timely corn mapping at regional scales supports effective agricultural management and policy development. However, traditional data-driven approaches rely heavily on extensive ground-truth samples and have limited applicability in data-scarce areas. And there is still a lack of simple and practical approaches that require few or no field observations for corn mapping. Therefore, this study presents a novel framework that integrates an improved Dual-peak Canopy Nitrogen Index (DCNI) with Green Chromatic Coordinate (GCC) analysis. The refined DCNI captures nitrogen-related spectral dynamics to distinguish corn from spectrally similar crops such as soybean and sorghum during peak growth stages. The GCC analysis identifies the optimal mapping window, mitigating the effects of cloud cover and temporal data gaps for regional corn mapping. We applied this improved index to three selected but representative corn growing regions: Nenjiang in China, Pocahontas in the United States, and Mayenne in France. Using the optimized DCNI thresholds, binary classification was performed to reduce both commission and omission errors. In the pixel-level validation for 2021, our DCNI index achieved better overall accuracy and F1 scores compared to the classical random forest model and its variants. Interannual tests from 2020 to 2022 also showed stable performance and strong agreement with official statistics, with all coefficients of determination R<sup>2</sup> above 0.96. We also investigate uncertainties arising from data interpolation, limited field samples, and mixed cropping patterns, and propose the integration of higher-frequency satellite observations and multisource data fusion to improve early season monitoring and broaden large-scale applicability. The proposed framework requires minimal crop samples and computational resources, providing a simple, practical alternative for regional corn mapping with robust transferability.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100308"},"PeriodicalIF":5.2,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320136","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-09DOI: 10.1016/j.srs.2025.100307
J.C.B. da Silva , J.M. Magalhaes , A. Bosser , R. Huerre , Ariane Koch-Larrouy , Chloé Goret , Souley Diallo , Carina R. de Macedo , Alex Costa da Silva
Satellite remote sensing has revolutionized the study of Internal Solitary Waves (ISWs), revealing wave characteristics that are hardly obtainable through traditional in situ instrumentation. It enables the observation of their full two-dimensional horizontal structure, crest lengths, propagation direction, and phase speed, all on a global scale. However, some essential ISW parameters such as their amplitude and wavelength have been more difficult to assess from their surface manifestations. In this paper we employ an inversion method based on the quantitative relationship between sea surface current and ISW surface topography measured from SWOT KaRIn. The inversion method employs a fully nonlinear equation with continuous stratification to account for the strongly nonlinear nature of ISWs and uses the sea surface height anomaly from KaRIn measurements as a constraint to determine a unique solution. The method is tested on a case study in deep waters off the Amazon shelf in the Tropical Atlantic where in situ measurements quasi-coincident with a SWOT overpass allow evaluation of its accuracy. By directly contrasting the DJL retrievals with estimates from weakly nonlinear KdV theory, we show that KdV underestimates wave amplitudes and fits poorly surface expressions, whereas DJL yields accurate fits to both SWOT and mooring observations. We address a new possibility to calculate ISW parameters such as amplitude, wavelength, phase speed and wave induced velocity field based on fully nonlinear theory and discuss typical error margins that must be dealt with by researchers willing to use SWOT KaRIn in ISW studies.
{"title":"Internal solitary wave parameters from SWOT KaRIn sea surface topography: a case study in the Tropical Atlantic","authors":"J.C.B. da Silva , J.M. Magalhaes , A. Bosser , R. Huerre , Ariane Koch-Larrouy , Chloé Goret , Souley Diallo , Carina R. de Macedo , Alex Costa da Silva","doi":"10.1016/j.srs.2025.100307","DOIUrl":"10.1016/j.srs.2025.100307","url":null,"abstract":"<div><div>Satellite remote sensing has revolutionized the study of Internal Solitary Waves (ISWs), revealing wave characteristics that are hardly obtainable through traditional in situ instrumentation. It enables the observation of their full two-dimensional horizontal structure, crest lengths, propagation direction, and phase speed, all on a global scale. However, some essential ISW parameters such as their amplitude and wavelength have been more difficult to assess from their surface manifestations. In this paper we employ an inversion method based on the quantitative relationship between sea surface current and ISW surface topography measured from SWOT KaRIn. The inversion method employs a fully nonlinear equation with continuous stratification to account for the strongly nonlinear nature of ISWs and uses the sea surface height anomaly from KaRIn measurements as a constraint to determine a unique solution. The method is tested on a case study in deep waters off the Amazon shelf in the Tropical Atlantic where in situ measurements quasi-coincident with a SWOT overpass allow evaluation of its accuracy. By directly contrasting the DJL retrievals with estimates from weakly nonlinear KdV theory, we show that KdV underestimates wave amplitudes and fits poorly surface expressions, whereas DJL yields accurate fits to both SWOT and mooring observations. We address a new possibility to calculate ISW parameters such as amplitude, wavelength, phase speed and wave induced velocity field based on fully nonlinear theory and discuss typical error margins that must be dealt with by researchers willing to use SWOT KaRIn in ISW studies.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100307"},"PeriodicalIF":5.2,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-07DOI: 10.1016/j.srs.2025.100304
Yasser Maghsoudi , Andrew J. Hooper , Tim J. Wright , Milan Lazecky , Muriel Pinheiro
Interferometric synthetic aperture radar (InSAR) phase bias, most commonly observed in short-term, multilooked interferograms, can significantly impact the accuracy of ground displacement measurements, particularly in regions with vegetation or temporal decorrelation. While phase bias can largely be corrected using phase linking methods, these approaches are computationally intensive and ineffective for pixels with coherence limited to very short time intervals. This paper aims to consolidate our empirical correction algorithm for mitigating phase bias in InSAR data, demonstrating its application across three diverse study areas: the Azores in Portugal, Campi Flegrei in Italy, and Tien Shan in China. Our method estimates bias terms using only short-term wrapped interferograms and applies these terms to correct any desired interferograms. The algorithm also addresses gaps and missing interferograms within time-series data by incorporating temporal smoothing constraints, which minimize differences between estimated bias terms over time. Additionally, the study examines the temporal and spatial behavior of the calibration parameters and explores the choice of long-term interferograms for their estimation. Validation against a phase linking approach shows that our phase bias correction algorithm effectively reduces phase bias, achieving close alignment with the benchmark results. This work contributes a robust framework for correcting short-term interferograms, leading to improved InSAR velocity estimates.
{"title":"Advances in mitigating InSAR non-closure phase bias: A refined processing approach","authors":"Yasser Maghsoudi , Andrew J. Hooper , Tim J. Wright , Milan Lazecky , Muriel Pinheiro","doi":"10.1016/j.srs.2025.100304","DOIUrl":"10.1016/j.srs.2025.100304","url":null,"abstract":"<div><div>Interferometric synthetic aperture radar (InSAR) phase bias, most commonly observed in short-term, multilooked interferograms, can significantly impact the accuracy of ground displacement measurements, particularly in regions with vegetation or temporal decorrelation. While phase bias can largely be corrected using phase linking methods, these approaches are computationally intensive and ineffective for pixels with coherence limited to very short time intervals. This paper aims to consolidate our empirical correction algorithm for mitigating phase bias in InSAR data, demonstrating its application across three diverse study areas: the Azores in Portugal, Campi Flegrei in Italy, and Tien Shan in China. Our method estimates bias terms using only short-term wrapped interferograms and applies these terms to correct any desired interferograms. The algorithm also addresses gaps and missing interferograms within time-series data by incorporating temporal smoothing constraints, which minimize differences between estimated bias terms over time. Additionally, the study examines the temporal and spatial behavior of the calibration parameters and explores the choice of long-term interferograms for their estimation. Validation against a phase linking approach shows that our phase bias correction algorithm effectively reduces phase bias, achieving close alignment with the benchmark results. This work contributes a robust framework for correcting short-term interferograms, leading to improved InSAR velocity estimates.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100304"},"PeriodicalIF":5.2,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320137","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-05DOI: 10.1016/j.srs.2025.100303
Yi Zhao , Haibin Shi , Xianyue Li , Shuya Yang , Qingfeng Miao , Jianwen Yan , Cong Hou
The Hetao Irrigation District in Inner Mongolia represents a significant agricultural irrigation area in China, where the precise extraction of planting structure is crucial for the advancement of precision agriculture and smart irrigation practices in the region. To investigate the trends in crop cultivation area changes at the branch canal scale within the Hetao Irrigation District, a 3-year experiment was conducted in the Zuo Er Branch Canal, which is located downstream of the district. This study employed an innovative multiple extraction method combined with a machine learning model to accurately extract the planting structure. The results showed that from April to September over a 3-year period, the NDVI spectral curves for cultivated land and ditch canal exhibited a trend of increasing first and then decreasing, while wasteland and road displayed relatively stable curves with minimal variation. When extracting land use types, the overall accuracy of the gradient boosting tree model was improved by 1.74 %, and 11.43 % compared with that of the random forest and decision tree, and the accuracy of the detail validation was higher. For planting structure extraction, the gradient boosting tree model has an average 3-year overall accuracy improvement of 3.24 % and 8.35 % over the random forest and decision tree models. The area designated for sunflower planting has increased annually, showing a 27.47 % rise in 2024 compared to 2022. In contrast, the area allocated for maize planting has decreased each year, with a significant 66.18 % reduction in 2024 relative to 2022. This study offers crucial theoretical insights and practical implications for the dynamic analysis of planting structure and the modern management of agriculture within the Hetao Irrigation District.
{"title":"Fine extraction of planting structure at branch canal scale in the Hetao Irrigation District based on multiple extraction method","authors":"Yi Zhao , Haibin Shi , Xianyue Li , Shuya Yang , Qingfeng Miao , Jianwen Yan , Cong Hou","doi":"10.1016/j.srs.2025.100303","DOIUrl":"10.1016/j.srs.2025.100303","url":null,"abstract":"<div><div>The Hetao Irrigation District in Inner Mongolia represents a significant agricultural irrigation area in China, where the precise extraction of planting structure is crucial for the advancement of precision agriculture and smart irrigation practices in the region. To investigate the trends in crop cultivation area changes at the branch canal scale within the Hetao Irrigation District, a 3-year experiment was conducted in the Zuo Er Branch Canal, which is located downstream of the district. This study employed an innovative multiple extraction method combined with a machine learning model to accurately extract the planting structure. The results showed that from April to September over a 3-year period, the NDVI spectral curves for cultivated land and ditch canal exhibited a trend of increasing first and then decreasing, while wasteland and road displayed relatively stable curves with minimal variation. When extracting land use types, the overall accuracy of the gradient boosting tree model was improved by 1.74 %, and 11.43 % compared with that of the random forest and decision tree, and the accuracy of the detail validation was higher. For planting structure extraction, the gradient boosting tree model has an average 3-year overall accuracy improvement of 3.24 % and 8.35 % over the random forest and decision tree models. The area designated for sunflower planting has increased annually, showing a 27.47 % rise in 2024 compared to 2022. In contrast, the area allocated for maize planting has decreased each year, with a significant 66.18 % reduction in 2024 relative to 2022. This study offers crucial theoretical insights and practical implications for the dynamic analysis of planting structure and the modern management of agriculture within the Hetao Irrigation District.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100303"},"PeriodicalIF":5.2,"publicationDate":"2025-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145265186","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-03DOI: 10.1016/j.srs.2025.100302
Annika M. Zuleger , Martina M. Viti , Luise Quoss , Filipe S. Dias , Luís Borda-de-Água , Miguel N. Bugalho , Henrique M. Pereira
Herbivore-accessible biomass (HAB), defined as aboveground biomass under 2 m, including leaves and soft branches, is a key metric for understanding ecosystem function, but remains poorly quantified. We estimated HAB across diverse habitats in the Peneda-Gerês National Park using high-resolution NDVI, LiDAR, topography and field data. Generalized Additive Mixed Models (GAMMs) revealed habitat-specific effects of NDVI and vegetation height, as well as terrain, and structural metrics across plant types. Models were evaluated using hold-out cross-validation on a 20 % subset of the field data. The total HAB model performed well (Deviance Explained = 0.77, RMSE20 = 172.38 g/m2), while the shrub model performed slightly worse (Deviance Explained = 0.71, RMSE20 = 410.21 g/m2), and the herbaceous model exhibited a moderate fit and accuracy (Deviance Explained = 0.69, RMSE20 = 34.25 g/m2). Average total HAB was 1.31 ± 0.83 tons/ha, dominated by shrubs (1.02 tons/ha) compared to herbaceous HAB (0.14 tons/ha). HAB density varied by habitat, highest in shrublands (up to 1.83 ton/ha) and lowest in oak forests (0.85 tons/ha), while agricultural areas supported the most herbaceous HAB (0.68 tons/ha). These values are substantially lower than shrub biomass estimates reported in other studies (e.g., up to 30 tons/ha), reflecting our focus on live biomass <2 m. Prediction uncertainty was low (CV: 22–34 %), improving on other studies reporting up to 190 %, and highlighting the strength of combining spectral and structural data for fine-scale forage estimation. This study provides the first spatially explicit HAB estimates for the area, supporting herbivore ecology and management.
{"title":"Mapping herbivore-accessible biomass across a heterogeneous mountain landscape using multisensor high-resolution UAV data","authors":"Annika M. Zuleger , Martina M. Viti , Luise Quoss , Filipe S. Dias , Luís Borda-de-Água , Miguel N. Bugalho , Henrique M. Pereira","doi":"10.1016/j.srs.2025.100302","DOIUrl":"10.1016/j.srs.2025.100302","url":null,"abstract":"<div><div>Herbivore-accessible biomass (HAB), defined as aboveground biomass under 2 m, including leaves and soft branches, is a key metric for understanding ecosystem function, but remains poorly quantified. We estimated HAB across diverse habitats in the Peneda-Gerês National Park using high-resolution NDVI, LiDAR, topography and field data. Generalized Additive Mixed Models (GAMMs) revealed habitat-specific effects of NDVI and vegetation height, as well as terrain, and structural metrics across plant types. Models were evaluated using hold-out cross-validation on a 20 % subset of the field data. The total HAB model performed well (Deviance Explained = 0.77, RMSE<sub>20</sub> = 172.38 g/m<sup>2</sup>), while the shrub model performed slightly worse (Deviance Explained = 0.71, RMSE<sub>20</sub> = 410.21 g/m<sup>2</sup>), and the herbaceous model exhibited a moderate fit and accuracy (Deviance Explained = 0.69, RMSE<sub>20</sub> = 34.25 g/m<sup>2</sup>). Average total HAB was 1.31 ± 0.83 tons/ha, dominated by shrubs (1.02 tons/ha) compared to herbaceous HAB (0.14 tons/ha). HAB density varied by habitat, highest in shrublands (up to 1.83 ton/ha) and lowest in oak forests (0.85 tons/ha), while agricultural areas supported the most herbaceous HAB (0.68 tons/ha). These values are substantially lower than shrub biomass estimates reported in other studies (e.g., up to 30 tons/ha), reflecting our focus on live biomass <2 m. Prediction uncertainty was low (CV: 22–34 %), improving on other studies reporting up to 190 %, and highlighting the strength of combining spectral and structural data for fine-scale forage estimation. This study provides the first spatially explicit HAB estimates for the area, supporting herbivore ecology and management.</div></div>","PeriodicalId":101147,"journal":{"name":"Science of Remote Sensing","volume":"12 ","pages":"Article 100302"},"PeriodicalIF":5.2,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320135","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","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