Remote Sensing of Environment最新文献

{"title":"Edge effects on forest dynamics in China from 2000 to 2020: Evidence from satellite remote sensing","authors":"Jinlong Chen,&nbsp;Zeng Cui,&nbsp;Zhiyao Tang","doi":"10.1016/j.rse.2025.115187","DOIUrl":"10.1016/j.rse.2025.115187","url":null,"abstract":"<div><div>In the 21st century, forest edge-to-interior gradients have exhibited remarkable spatiotemporal variation in China. In this study, we employed a high-resolution framework integrating NDVI, tree cover, canopy height, and forest age to evaluate edge effects on forest growth, structure, and maturity in China from 2000 to 2020. The results revealed that (1) the proportion of forest edges within 60 m declined from 38 % (653,000 km² out of 1,694,000 km²) in 2000 to 33 % (680,000 km² out of 2,033,000 km²) in 2020, accompanied by an increase in interior forests (&gt;300 m) from 22 % (381,000 km²) to 29 % (590,000 km²); (2) NDVI, tree cover, canopy height and forest age increased from the forest edge toward the interior, with the strongest effects occurring within 300 m of the edge and detectable up to 1.2 km into the interior. However, the edge forests demonstrated faster temporal increases in NDVI and tree cover; and (3) these effects varied with the land cover types adjacent to the edge. Specifically, NDVI (0.862) and tree cover (81.17 %) peaked at cropland edges and were lowest at grassland edges (0.793 and 41.24 %), while canopy height ranged from 9.77 m at grassland edges to 12.76 m at waterbody edges. These findings are valuable for advancing forest restoration, mitigating fragmentation, and enhancing natural carbon storage potential in China.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115187"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145703964","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}

{"title":"Spatial-X fusion for multi-source satellite imageries","authors":"Jiang He ,&nbsp;Liupeng Lin ,&nbsp;Zhuo Zheng ,&nbsp;Qiangqiang Yuan ,&nbsp;Jie Li ,&nbsp;Liangpei Zhang ,&nbsp;Xiao xiang Zhu","doi":"10.1016/j.rse.2025.115214","DOIUrl":"10.1016/j.rse.2025.115214","url":null,"abstract":"<div><div>Multi-source remote sensing data can highlight different types of information based on user needs, resulting in large volumes of data and significant challenges. Hardware and environmental constraints create mutual dependencies between information types, particularly between spatial data and other types, limiting the development of high-precision applications. Traditional methods are task-specific, leading to many algorithms without a unified solution, which greatly increases the computational and deployment costs of image fusion. In this paper, we summarize four remote sensing fusion tasks, including pan-sharpening, hyperspectral-multispectral fusion, spatio-temporal fusion, and polarimetric SAR fusion. By defining the spectral, temporal, and polarimetric information, as X, we propose the concept of generalized spatial-channel fusion, referred to as Spatial-X fusion. Then, we design an end-to-end network SpaXFus, a generalized spatial-channel fusion framework through a model-driven unfolding approach that exploits spatial-X intrinsic interactions to capture internal dependencies and self-interactions. Comprehensive experimental results demonstrate the superiority of SpaXFus, e.g., SpaXFus can achieve four remote sensing image fusion tasks with superior performance (across all fusion tasks, spectral distortion decreases by 25.48 %, while spatial details improve by 7.5 %) and shows huge improvements across multiple types of downstream applications, including vegetation index generation, fine-grained image classification, change detection, and SAR vegetation extraction.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115214"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146001412","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}

{"title":"GPP-net: a robust high-resolution GPP estimation network for Sentinel-2 using only surface reflectance and photosynthetically active radiation","authors":"Shaoyu Wang ,&nbsp;Youngryel Ryu ,&nbsp;Benjamin Dechant ,&nbsp;Helin Zhang ,&nbsp;Huaize Feng ,&nbsp;Jeongho Lee ,&nbsp;Changhyun Choi","doi":"10.1016/j.rse.2025.115198","DOIUrl":"10.1016/j.rse.2025.115198","url":null,"abstract":"<div><div>High-resolution gross primary productivity (GPP) estimation is crucial for ecological and agricultural applications that require fine spatial details to capture GPP heterogeneity. Satellite-based GPP estimation usually relies on land cover and meteorological data. However, the misclassification of land cover data and coarse resolution of meteorological data greatly increase the uncertainty. Here, we propose a robust high-resolution GPP estimation deep learning (DL) network, named GPP-net, using only satellite surface reflectance (SR) from Sentinel-2 and photosynthetically active radiation (PAR). Specifically, GPP-net is based on a fully 1-D convolutional encoder-decoder network combined with a spectral band importance estimation module. To enhance the generalization of GPP-net, we ran the soil-canopy energy balance radiative transfer (SCOPE) model, and then combined these SCOPE-simulated reflectance data with GPP and PAR data extracted from FLUXNET2015 to pre-train GPP-net. Compared to benchmark models including near-infrared reflectance of vegetation multiplied by incoming sunlight (NIRvP), partial least squares (PLS) and random forest (RF), GPP-net improved half-hourly and daily GPP retrieval across seven plant functional types (PFTs) including four forest types, cropland, grassland and wetland. Owing to its robust nonlinear feature learning capabilities, GPP-net also facilitated robust GPP estimation across both C3 and C4 vegetation. We found that GPP-net could reliably estimate GPP under drought and heatwave conditions, with minimal improvement from including vapor pressure deficit (VPD) as a predictor. Furthermore, GPP-net demonstrated great robustness to soil effects in GPP mapping, and had strong ability in capturing inter-annual variability of GPP. The pretraining paradigm enabled us to fully leverage historical data, and the DL framework ensured that the model generalization continually improves as new data is integrated. Our model dispenses with land cover data and minimizes the requirements of coarse-resolution meteorological data for high-resolution GPP estimation, which could support future efforts in global high-resolution GPP mapping.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115198"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145771648","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}

{"title":"Retrieval of global surface phytoplankton community structure using a minimal set of predictors","authors":"Xueyin Li ,&nbsp;Shuguo Chen ,&nbsp;Junwei Wang ,&nbsp;Qing Zhu","doi":"10.1016/j.rse.2025.115174","DOIUrl":"10.1016/j.rse.2025.115174","url":null,"abstract":"<div><div>Phytoplankton underpin marine food webs and drive essential biogeochemical processes. Their pigment composition serves as a vital indicator of community structure and physiological state. In this study, we developed a multilayer perceptron (MLP)-based model to estimate the concentrations of 26 pigments and infer phytoplankton community structure in global surface waters. By integrating prior knowledge and quantifying through SHapley Additive exPlanations (SHAP), we selected a physically meaningful subset of 10 input features to construct pigment inversion models, including environmental parameters such as sea surface temperature (SST), salinity (SSS), and sea surface height (SSH), as well as remote sensing reflectance (<em>R</em>_<em>rs</em>) at seven spectral bands. For 24 pigments, the model achieved high predictive accuracy and strong generalization performance, with an average coefficient of determination (<em>R</em>^<em>2</em>) of 0.76 and median absolute percentage difference (MAPD) of 12.01 % under random cross-validation, and slightly lower accuracy under temporal cross-validation (<em>R</em>^<em>2</em> = 0.67; MAPD = 14.35 %). Feature analysis revealed that <em>R</em>_<em>rs</em>, particularly spectral regions encompassing absorption peaks and adjacent gradients, dominated pigment predictions. Moreover, the model captured nonlinear thermal responses of phytoplankton to SST, consistent with known ecophysiological patterns, and reflected synergistic interactions between SST and SSS affecting pigment variability. The inferred phytoplankton community structures, estimated <em>via</em> Diagnostic Pigment Analysis (DPA), showed strong agreement with previous studies, validating the model's ecological reliability. Additionally, Hydrolight simulations based on <em>in situ a</em>_<em>ph</em> spectra demonstrated that the model performs reliably under water conditions with suspended particulate matter (SPM) ≤ 1 g m⁻³ and colored dissolved organic matter (CDOM) absorption at 440 nm is ≤0.25 m⁻¹, maintaining MAPD below 25 %. Our research demonstrates that neural networks operate in a physically informed manner rather than as purely data-driven models, enabling them to represent complex interactions between phytoplankton and ecological environment. This underscores the necessity of designing ecologically informed and physically interpretable input features in remote sensing applications, offering valuable guidance for future biogeochemical modeling efforts.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115174"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145732128","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}

{"title":"Performance assessment of canopy gap fraction retrieval using multiple airborne LiDAR instrument configurations","authors":"Yi Li ,&nbsp;Hao Tang ,&nbsp;Shanshan Wei ,&nbsp;Donghui Xie ,&nbsp;Xihan Mu ,&nbsp;Guangjian Yan","doi":"10.1016/j.rse.2025.115166","DOIUrl":"10.1016/j.rse.2025.115166","url":null,"abstract":"<div><div>Canopy gap fraction (GF) is one of the most important vegetation structure parameters widely used in forest ecology and meteorology. Small-footprint airborne LiDAR has gained great popularity for canopy GF retrieval in the past decade; yet most studies were carried out at individual sites using different commercial-off-the-shelf instruments, often ignoring possible inconsistency due to the use of different instruments or scanning configurations. This study aims to provide an assessment of the performance of canopy GF retrieval from small-footprint airborne LiDAR under different collection scenarios. We utilized small-footprint airborne LiDAR data and field digital hemispherical photo (DHP) measurements from the National Ecological Observatory Network (NEON), encompassing 30 sites across 16 eco-regions in the United States from 2016 to 2022. A total of eight baseline cases together with different instrument configurations were included in this analysis. LiDAR-based canopy GF was first retrieved using established methods and then compared to a total of 4596 DHPs from 383 plots. Results show that all instruments could reach an accuracy better than 10 % RMSE under a proper configuration, however, the difference of using waveforms and point clouds alone could reach an RMSE up to 18 %. Scan angle showed the greatest impact among all sensor related parameters and could lead to a mean bias up to 5 %. Interestingly, waveforms did not always outperform point clouds, likely due either to the varying pulse shape of transmitted waveform or to low digitizer performance. In sum, our results caution the use of airborne LiDAR as the only means for validating large-scale satellite vegetation structure products or monitoring subtle long-term canopy changes. Assessing both instrument and acquisition specifications can help minimize potential bias in canopy GF retrieval over different ecosystems.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115166"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145731846","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}

{"title":"Integrated scan simultaneous trajectory enhancement and mapping (IS2-TEAM) for fine resolution forest inventory using backpack LiDAR","authors":"Chunxi Zhao ,&nbsp;Songlin Fei ,&nbsp;Ayman Habib","doi":"10.1016/j.rse.2025.115212","DOIUrl":"10.1016/j.rse.2025.115212","url":null,"abstract":"<div><div>Advancements in close-range remote sensing offer promises in automated forest inventory, facilitating sustainable forest management. Backpack LiDAR is a mobile mapping system that can be deployed for different applications, including forest inventory. However, under-canopy trajectories derived from backpacks equipped with GNSS units are often unreliable due to signal outages, resulting in degraded mapping results. To overcome this challenge, an Integrated Scan Simultaneous Trajectory Enhancement and Mapping (IS^{<strong>2</strong>}-TEAM) framework is proposed for fine-resolution forest inventory. For multi-beam spinning LiDAR, a single scan refers to points from a full revolution of the laser beam assembly, while an integrated scan combines multiple successive single scans. The framework introduces feature extraction strategies for defining reliable semantic features (tree trunk and ground) from integrated scans to enhance trajectory and mapping results. If available, the framework can incorporate a Digital Terrain Model (DTM) extracted from existing geospatial data to enhance georeferencing accuracy of backpack LiDAR. Finally, extracted features together with enhanced point cloud undergo a machine-learning post-processing strategy to evaluate the IS^{<strong>2</strong>}-TEAM's performance in tree detection and provide tree classification results based on maturity. The proposed approach has been thoroughly evaluated across multiple study sites with different forest types and terrain conditions. Through the experimental results, it has been shown that the IS^{<strong>2</strong>}-TEAM extracts reliable ground and tree trunk features and generates point clouds with alignment quality range of 2–4 cm. Furthermore, DTM-assisted IS²-TEAM significantly improves the georeferencing accuracy of the derived point cloud in forests with varying terrain conditions, achieving an average vertical accuracy improvement of 1 m across all test datasets. Finally, the proposed self-evaluation strategy successfully identifies all mature trees and achieves over 90 % accuracy in tree classification.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115212"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145881025","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}

{"title":"Comparing the performance of different hyperspectral satellite imaging spectroscopy in mapping methane point-source emissions","authors":"Fei Li ,&nbsp;Keer Lin ,&nbsp;Yingqi Yan ,&nbsp;Shengxi Bai ,&nbsp;Qidan Huang ,&nbsp;Chenxi Feng ,&nbsp;Shiwei Sun ,&nbsp;Shaohua Zhao ,&nbsp;Wei Zhou ,&nbsp;Chunyan Zhou ,&nbsp;Jun Lin ,&nbsp;Xinwei Zhang ,&nbsp;Yongguang Zhang","doi":"10.1016/j.rse.2025.115224","DOIUrl":"10.1016/j.rse.2025.115224","url":null,"abstract":"<div><div>Accurate detection and quantification of methane point sources are critical for climate change mitigation. Recent advances in spaceborne hyperspectral imaging spectrometers offer new opportunities for global monitoring, yet systematic evaluation across missions remains limited. Here, we assess the spectral and radiometric performance of six operational hyperspectral satellites for methane plume detection. Results show that EMIT, EnMAP, and the GF5 series provide superior capabilities. EMIT achieved the best retrieval precision (<em>σ</em> = 18 ppb), followed by GF5–02-AHSI (<em>σ</em> = 31 ppb), EnMAP (<em>σ</em> = 33 ppb), and GF5-01 A-AHSI (<em>σ</em> = 47 ppb), outperforming PRISMA (<em>σ</em> = 84 ppb) and ZY1-02E-AHSI (<em>σ</em> = 71 ppb). These differences arise from finer spectral sampling distances (SSD: 7.4–8.4 nm), higher signal-to-noise ratios (SNR: 170–250), and more stable central wavelength shifts (0.3–1.3 nm) within the 2300 nm methane absorption region, which collectively enhance plume discrimination against surface backgrounds. In contrast, PRISMA and ZY1-02E-AHSI exhibit lower sensitivity due to larger wavelength shifts (up to 2.8 nm) and lower SNRs (140–150). Case studies illustrate successful detection of methane plumes from diverse sources, including oil and gas (O&amp;G) infrastructure, coal mines, and a landfill site, highlighting both straightforward and challenging retrieval scenarios. This cross-sensor comparison underscores the importance of spectral fidelity and radiometric performance for methane monitoring. The findings provide a quantitative basis for prioritizing existing assets and guiding the design of future missions, emphasizing that high SNR and stable spectral calibration are key for advancing global detection of point-source methane emissions.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115224"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145894482","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}

{"title":"Probabilistic mapping of high-intensity forest fire potential via time series machine learning and remote sensing-informed fire spread simulations","authors":"Rui Chen ,&nbsp;Yiru Zhang ,&nbsp;Yanxi Li ,&nbsp;Marta Yebra ,&nbsp;Chunquan Fan ,&nbsp;Hongguo Zhang ,&nbsp;Binbin He","doi":"10.1016/j.rse.2026.115233","DOIUrl":"10.1016/j.rse.2026.115233","url":null,"abstract":"<div><div>High-intensity forest fires have significant destructive impacts on ecosystems and society, and are an increasing concern worldwide. Accurate probabilistic risk assessment of these fires can effectively enhance the ability to guide wildfire management, particularly for large and extreme fires. However, forecasting large-scale fire behavior characteristics remains challenging, limiting the effectiveness of spatial estimations of high-intensity forest fire potential (HIFFP). This study aims to integrate fire spread simulations and machine learning (ML) algorithms to enhance HIFFP estimations through multi-step time-series forecasting on fire rate of spread and fireline intensity at regional scales. We first established a high-intensity forest fire dataset based on remote sensing-informed fire spread simulations from the Weather Research and Forecasting coupled fire-spread model (WRF-SFIRE), incorporating explanatory variables on fuel, weather, climate, and topography. Then, the knowledge-guided framework (multi-step time series-based ML, MTS-ML) was designed to estimate HIFFP within different hours after fires occur, integrating with Bayesian Network (BN), Random Forest (RF), and copula models. Results indicate that MTS-ML improved HIFFP modeling compared with ML-based methods, achieving AUC (the area under the receiver operating characteristic curve) &gt; 0.95 (with ∼0.04 increments), F1 score &gt; 0.85 (with ∼0.08 increments), and MAE &lt; 0.15. Topographic index, foliage fuel load, and wind speed are identified as primary contributors to HIFFP. Probabilistic mapping of HIFFP represents wildfire danger, which is closely linked to burn severity and fire-induced carbon emissions. This study presents a novel framework for enhancing regional risk assessment of high-intensity forest fires, providing valuable guidance in wildfire control and management.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115233"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938801","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}

{"title":"Unveiling the long-term cascading effects of the 2018 Baige landslide and subsequent outburst flood with satellite radar observations","authors":"Bo Chen ,&nbsp;Zhenhong Li ,&nbsp;Chuang Song ,&nbsp;Roberto Tomás ,&nbsp;Chen Yu ,&nbsp;Wu Zhu ,&nbsp;Jianbing Peng","doi":"10.1016/j.rse.2026.115231","DOIUrl":"10.1016/j.rse.2026.115231","url":null,"abstract":"<div><div>Landslide-dammed lakes (LDL) and landslide lake outburst flood (LLOF) can significantly alter the kinematic behavior of upstream and downstream landslides, posing severe threats to human life and infrastructure. However, the long-term impacts of LDL and LLOF on surrounding landslide stability remain poorly understood. In this study, we systematically examine the cascading effects triggered by the 2018 Baige LDL and LLOF on adjacent landslides, based on time series interferometric synthetic aperture radar (InSAR) analysis of 1437 satellite radar images. Unlike previous studies that focused on individual landslides or localized areas, we developed an automated method to detect the onset of landslide acceleration, leading to the establishment of an inventory of 65 accelerated landslides (ALs) and a quantitative evaluation of their controlling factors. Our results show that approximately 30 % of the flood-affected active landslides changed their deformation mechanisms, which can be categorized into five distinct types. Among the landslides accelerated by the Baige event, 43 % exhibited persistent acceleration, whereas 57 % showed signs of self-recovery. For the latter, deformation velocity typically decayed by 90 % within an average of 9.3 years after the outburst, returning to near pre-event levels. Furthermore, compared to 378 flood-involved but non-ALs, ALs preferentially occur on gentler slopes and in areas with lower vegetation cover. More notably, those ALs generally experienced greater flood depth, higher flow velocity, and stronger flood power. This study is the first to assess the long-lasting cascading effects of LDL and LLOF on creep landslides. These findings advance our understanding of LDL and LLOF-induced landslide mechanisms and offer valuable insights for the long-term risk assessment and geohazard mitigation of landslide-prone regions affected by similar cascading processes.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115231"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938859","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}

{"title":"Estimating the upper depth of subsurface water on the Greenland Ice Sheet using multi-frequency passive microwave remote sensing, radiative transfer modeling, and machine learning","authors":"Baptiste Vandecrux ,&nbsp;Ghislain Picard ,&nbsp;Pierre Zeiger ,&nbsp;Marion Leduc-Leballeur ,&nbsp;Andreas Colliander ,&nbsp;Alamgir Hossan ,&nbsp;Andreas Ahlstrøm","doi":"10.1016/j.rse.2025.115197","DOIUrl":"10.1016/j.rse.2025.115197","url":null,"abstract":"<div><div>As the Arctic warms, surface melt extends into the Greenland Ice Sheet's accumulation zone, where much of the water infiltrates into the snowpack. This makes monitoring the subsurface water depth and spatial extent important for accurate ice sheet runoff estimations. Subsurface water can be detected using remotely sensed microwave brightness temperatures (T_B). We use vertically polarized T_B at 1.4 GHz from Soil Moisture and Ocean Salinity satellite (SMOS) and at 6.9, 10.7, and 18.7 GHz from the Advanced Microwave Scanning Radiometers (AMSR-E/2) to estimate the upper depth of liquid water (UDLW) on the ice sheet accumulation area. We build a catalogue of simulated UDLW and T_B: realistic UDLW are modeled by the Geological Survey of Denmark and Greenland (GEUS) snow model, forced by the Copernicus Arctic Regional Reanalysis (CARRA), and the corresponding T_B are calculated by the Snow Microwave Radiative Transfer (SMRT) model at 19 sites. We train on this catalogue an ensemble of cross-validated Random Forest (RF) models to predict UDLW and its uncertainty from T_B at four frequencies. On hold-out modeled data and for water within 5 m of the surface, the RF ensemble achieves a median RMSE of 0.68 m and mean error of −0.09 m. Our retrieval, when applied to observed T_B, matches within 2 m UDLW inferred from subsurface temperature profiles down to 4–6 m depth. Performances decrease beyond 5 m depth and for low liquid water amounts. Our retrieval produces daily UDLW maps over the ice sheet's accumulation area during 2010–2023 which reveal the seasonal evolution of UDLW, deliver the first quantitative estimates of subsurface liquid water depth on the ice sheet and offer new insights into meltwater infiltration and storage processes.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"334 ","pages":"Article 115197"},"PeriodicalIF":11.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938861","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}