Lakes play a crucial role in supplying water resources, regulating regional climates, and supporting ecosystems. However, they are increasingly threatened by recurrent droughts. This study focuses on the Mu Us Sandy Land, the fourth largest desert in China, and presents an approach that combines machine-learning techniques with a newly constructed drought index—the detrended cumulative standardized precipitation-evapotranspiration index (DeCumSPEI)—to forecast monthly lake-area variations from 2001 to 2020. Remote sensing data, including lake-area measurements, were obtained using the Google Earth Engine platform. In addition, meteorological factors—such as precipitation, temperature, and actual evapotranspiration (ETa) as well as anthropogenic variables, such as crop evapotranspiration, the normalized difference vegetation index, and land-use and land-cover change—were collected. The study assessed the performance of six machine-learning models using fivefold cross validation: gradient boosting decision tree, extra trees, random forest, adaptive boosting (AB), bootstrap aggregating (Bagging), and eXtreme gradient boosting. These models were evaluated for their ability to predict lake areas under both short-term (monthly) and long-term (annual) drought conditions. In addition, the influence of ETa as an upper boundary condition was investigated. The results show that: First, all models, except for AB and Bagging, demonstrated strong predictive performance, achieving coefficients of determination (R2) as high as 0.833, and the lowest average root-mean-square error and standard deviation of 1.05 and 1.01 km2. Second, incorporating the 12-month scale DeCumSPEI significantly enhanced model accuracy, with performance improvements of up to 32.01%. Third, comparing models with and without ETa confirmed the critical role of ETa in improving prediction accuracy. These findings offer valuable insights for future lake area forecasting in drought-affected regions and underscore the potential of machine-learning models in hydrological and drought response research.

The spatiotemporal reorganization of sediment-supply mechanisms determines sediment availability in river-delta-wetland systems, yet a unified basin-scale diagnosis of Supply Limitation (SL), Normal Transport (NT), and Strong Mobilization (SM) is lacking for the Yangtze River Basin (YRB). Using 63,996 Landsat images and hydrological data (2006–2024) across eight sub-basins, we developed an SSC inversion framework coupling optical water-type clustering with machine learning, and identified the three regimes via station-scale SSC-Q relationships, hydrologically normalized anomalies, and extreme-flow analysis. Validation achieved R² = 0.56–0.73 and RMSE = 44–62 g·m⁻³. SSC declined significantly basin-wide, fastest in autumn (−1.09 g·m⁻³·yr⁻¹), indicating persistent water clearing. Sediment supply shifted from positive to negative anomalies, reflecting a transition from active mobilization to stronger constraints. Limitation intensified most in the Main Stem, Yalong River, and Poyang Lake; Han River and Dongting Lake remained stable, while Minjiang and Wuyang showed phase-dependent fluctuations. NT dominated, but SL and SM alternated frequently. High runoff did not always increase sediment export, nor low runoff always reduce it, revealing nonlinear, asymmetric responses. This underscores the diagnostic value of distinguishing discharge-controlled transport from supply constraints. The findings support wetland restoration and conservation in the Yangtze Delta.

Accelerating urban population concentration across the world has both positive and negative social-ecological consequences. Countries facing serious negative consequences like Japan need to explore enhanced deurbanization pathways, e.g., highlighting high quality of life in rural areas, to achieve a more desirable urban-rural population balance. Using a nation-wide survey of urban-to-rural migrants in Japan (N = 1315), public geospatial data and structural equation modelling, we identified five “rural capital” factors affecting quality of life of urban-to-rural migrants, i.e., touristic, rural land, human, urban-type and forest capital, and five “human need” factors, i.e., urban-like services, rural sociocultural benefits, rural subsistence, rural amenity and child-raising needs. A path analysis revealed the pathways in which these factors contribute to change in migrants’ subjective well-being (SWB): Rural touristic capital (e.g., unique natural/cultural capital) has strong positive indirect effects to increase SWB, through the provision of urban-like services and socio-cultural benefits. We also found a significant indirect contribution of forests to SWB through increased rural amenities. The contribution of household income to SWB was confirmed but marginal as compared with these factors. Furthermore, multiple group models revealed significantly different contributors to SWB among people in different demographic and socio-economic groups. These findings have important policy implications, particularly for local governments to developing and implementing rural revitalization strategies. Specifically, placing QoL as the ultimate goal, and understanding and responding to the difference in what constitute good quality of life among people would enable inclusive and effective actions.

Over the last decade, L-band synthetic aperture radar (SAR) satellite data has become more widely available globally, providing new opportunities for biodiversity and ecosystem services (BES) monitoring. To better understand these opportunities, we conducted a systematic scoping review of articles that utilized L-band synthetic aperture radar (SAR) satellite data for BES monitoring. We found that the data have mainly been analyzed using image classification and regression methods, with classification methods attempting to understand how the extent, spatial distribution, and/or changes in different types of land use/land cover affect BES, and regression methods attempting to generate spatially explicit maps of important BES-related indicators like species richness or vegetation above-ground biomass. Random forest classification and regression algorithms, in particular, were used frequently and found to be promising in many recent studies. Deep learning algorithms, while also promising, have seen relatively little usage thus far. PALSAR-1/-2 annual mosaic data was by far the most frequently used dataset. Although free, this data is limited by its low temporal resolution. To help overcome this and other limitations of the existing L-band SAR datasets, 64% of studies combined them with other types of remote sensing data (most commonly, optical multispectral data). Study sites were mainly subnational in scale and located in countries with high species richness. Future research opportunities include investigating the benefits of new free, high temporal resolution L-band SAR datasets (e.g., PALSAR-2 ScanSAR data) and the potential of combining L-band SAR with new sources of SAR data (e.g., P-band SAR data from the “Biomass” satellite) and further exploring the potential of deep learning techniques.

National-scale monitoring of mangrove forests, including their spatial structure (e.g., canopy height and above-ground biomass), is necessary for the implementation of global agreements related to climate change and biodiversity conservation. Field measurements of mangrove structural parameters, however, are costly to collect. Here, we investigated the use of freely available data from multiple satellite sensors to monitor mangrove height and above-ground biomass at the national scale in Mauritius. L-band synthetic aperture radar data, optical multispectral data, and an existing global tree canopy height map were used as the main inputs to a regression model for estimating mangrove canopy height, while satellite LIDAR height measurements (Global Ecosystem Dynamics Inves-tigation (GEDI) data) were used to calibrate and validate the models. Based on 5-fold cross-validation, random forest regression achieved an R2 value of 0.46 and root-mean-square error (RMSE) of 4.45m. The main factor limiting higher model accuracy was likely the sparsity of satellite LIDAR data points in man-grove areas of Mauritius (n = 65), and this may be an issue in other SIDS also due to the limited geographic coverage of the LIDAR data. Still, our approach outperformed the existing global tree canopy height map (RMSE = 5.33) and a linear regression modeling approach (R2 = 0.34, RMSE = 4.85m).

Over the past few decades, Scenario analysis emerged as a useful tool for environmental decision-making amidst multiple uncertainties. Using the influential drivers of change, Scenarios portray the range of plausible alternative futures useful for quantifying the synergies and trade-offs of vital ecosystem services across multiple development trajectories. In this research, we demonstrate two case examples of the application of Scenarios in quantifying current and future mangrove ecosystem services. The case studies are selected from two representative sites: Tamsui River Estuary in Taiwan and Bhitarkanika mangroves in Odisha, India. Using the combination of Land Change Modeller (LCM) and InVEST ecosystem services simulation Tool, the research demonstrates the application and use of spatially explicit Scenarios for mangroves’ current and future conservation. As such, the case studies identify an ameliorative way of future planning, particularly with respect to the eco-sensitive development of coastal regions and small islands.

Integration of solar photovoltaics on croplands ("agrivoltaics") has been promoted as an environmentally-friendly approach for solar energy deployment. Past studies, however, have not considered that these croplands could alternatively be used as sites for expanding agroforestry, a practice which provides important ecosystem services to the neighbouring environment. We assessed the potential of agrivoltaics on herbaceous croplands in ASEAN, considering potential trade-offs with agroforestry. We assumed that croplands located in environmentally sensitive areas (ESAs) – including protected areas, key biodiversity areas, forests, wetlands/inland water bodies, their buffer zones, and areas with steep slopes – were better suited for agroforestry than agrivoltaics. We found that even if agrivoltaics are prohibited on all croplands located in ESAs, using just 10 % of the remaining land for agrivoltaics can still allow it to provide most of ASEAN's electricity generation needs. Thus, large-scale expansion of agrivoltaics need not conflict with regional efforts to enhance biodiversity/ecosystem services through agroforestry.

Globally, urbanization constitutes one of the major underlying drivers of global ecological degradation. Hence, deurbanization, i.e., demographic shift from urban to distant rural areas in a way that increases quality of life (QoL), can be one of the key pathways to address this global challenge. In this study, we investigated the contribution of nature and other types of rural capital to QoL and to people’s decision to migrate from urban to rural areas by studying residents in Hokuto City, a popular urban-to-rural migration destination in Japan. An integrated analysis of the 414 responses to a questionnaire survey and open and commercial geospatial datasets representing natural, built, human, cultural, and financial capital revealed the contributions of specific elements of rural capital to people’s QoL. These included natural capital (farmland, symbolic natural sites, mountain peak view, lower temperature, and tranquility), built capital (highways, railway stations, shops, and restaurants), and financial capital (employment). Many of these are related to the reasons that migrants, including return and one-way migrants, chose their present home location in Hokuto City, indicating their intention to increase QoL by migration. Particularly, one-way migrant homes were located predominantly on higher up mountain slopes with lower temperatures, higher forest cover, near natural parks, and symbolic natural sites, and yet with easier access to railway stations and employment. These results provide a valuable evidence base for rural spatial planning for increased QoL and attracting migrants that considers ecological–social feedbacks, and hence supports deurbanization.

Dynamic monitoring of reservoir water storage in arid areas is important for water resources assessment, hydroelectric power generation and agricultural irrigation. However, existing reservoir water calculation methods often rely on in-situ measurements, which limits their application in data scarce regionals and for regional scale analyses. Hence, we propose a novel method to estimate the water storage of channel-type reservoirs in arid areas with unknown underwater topography, with the Bosten Lake watershed serving as a case study site. The method first divides reservoirs into three types based on their upstream and downstream topography: V-shape, U-shape, and flat-shape reservoirs. For the V-shape and U-shape reservoirs, the underwater topography was produced by fitting a linear fit and a polynomial based on the observed elevation above the water surface, respectively. Meanwhile, extrapolation or splining techniques were used to derive the underwater topography for the flat-shape reservoir. The proposed methods are able to measure the underwater topography of the Bosten Lake watershed accurately, with the coefficient of determination (R2) values of 0.83, 0.75 and 0.61 for the V-shape, U-shape, and flat-shape reservoirs, respectively. In addition, the fit of the in-situ water depths of unmanned ships was matched to the simulated water depths for the Xiaoshankou and Bayi reservoirs, yielding R2 values of 0.91 and 0.83 as well as root mean square error (RMSE) of 1.27 m and 1.18 m, respectively. Our approach may be applied in other areas where river underwater topography data is lacking or sparse, and provide important basis for rational water resources management in these areas.

Research on transformers in remote sensing (RS), which started to increase after 2021, is facing the problem of a relative lack of review. To understand the trends of transformers in RS, we undertook a quantitative analysis of the major research on transformers over the past two years by dividing the application of transformers into eight domains: land use/land cover (LULC) classification, segmentation, fusion, change detection, object detection, object recognition, registration, and others. Quantitative results show that transformers achieve a higher accuracy in LULC classification and fusion, with more stable performance in segmentation and object detection. Combining the analysis results on LULC classification and segmentation, we have found that transformers need more parameters than convolutional neural networks (CNNs). Additionally, further research is also needed regarding inference speed to improve transformers’ performance. It was determined that the most common application scenes for transformers in our database are urban, farmland, and water bodies. We also found that transformers are employed in the natural sciences such as agriculture and environmental protection rather than the humanities or economics. Finally, this work summarizes the analysis results of transformers in remote sensing obtained during the research process and provides a perspective on future directions of development.

Coastal social-ecological systems (SES) are essential for the wellbeing of coastal communities and the wider society. However, in many parts of the world coastal SES face rapid change, and ultimately degradation. In this paper we unravel the mechanisms and implications of change in coastal SESmobilising multiple sources of knowledge, including scientific, expert-based and traditional and local knowledge (TLK). We focus on the rapidly changing Nakatsu mudflat in Japan, and combine primary and secondary data elicited through a mixed-method participatory approach that mobilised local stakeholders with different types of engagement with (and knowl-edge of) the mudflat. Through 4 expert interviews and 40 questionnaire surveys we identified the main ecosystem services provided by the mudflat that are perceived to be essential to the wellbeing of the local community. Although practically all respondents identified food provision as an important mudflat ecosystem services, many also pointed to the importance of some cultural (e.g. aesthetic beauty, spirituality, education and knowledge) and supporting services (e.g. habitat provision, sediment formation/retention). Through 8 Focus Group Discussions (FGD) and concept mapping we identified and systematized the underlying direct and indirect drivers of ecosystem change in the Nakatsu mudflat. These include population ageing and shrinking, economic diversification, and technological change that have collectively eroded TLK practices associated with the sus-tainable use of the mudflat. We also identified the mechanisms mediating these drivers and how they unfold in reality. Our study demonstrates that participatory processes engaging multiple stakeholders with different types of knowledge can provide rich and useful information on coastal SES change, which might not be readily obvious from simple headline indicators such as the change in the extent of the SES.

Grassland degradation poses a serious threat to biodiversity, ecosystem services, and human well-being. In this study, we investigated grassland degradation in Zhaosu County, China, between 2001 and 2020, and analyzed the impacts of climate change and human activities using the Miami model. The actual net primary productivity (ANPP) obtained with CASA (Carnegie-Ames-Stanford Approach) modeling, showed a decreasing trend, reflecting the significant degradation that the grasslands in Zhaosu County have experienced in the past 20 years. Grassland degradation was found to be highest in 2018, while the degraded area continuously decreased in the last 3 years (2018-2020). Climatic factors for found to be the dominant factor affecting grassland degradation, particularly the decrease in precipitation. On the other hand, human activities were found to be the main factor affecting improvement of grasslands, especially in recent years. This finding profoundly elucidates the underlying causes of grassland degradation and improvement and helps implement ecological conservation and restoration measures. From a practical perspective, the research results provide an important reference for the formulation of policies and management strategies for sustainable land use.

National monitoring of forests is essential for tracking progress towards various global environmental goals, including those of the Kunming-Montreal Global Biodiversity Framework and the Paris Agreement. Inconsistent national definitions of "forest", however, can complicate the tracking of global progress towards achieving these goals. The FAO's (Food and Agricultural Organization of the UN) definition of "Forest" is well-known and broad enough to be applicable globally, but it is difficult for countries to produce national forest maps according to this definition using only a single source of remote sensing data. Here, we developed an approach to integrate multiple existing land use/land cover (LULC) maps and generate an integrated map of forests and "Other land with tree cover" that is more consistent with FAO definitions. The proposed approach is based on merging thematic information from the global "PALSAR-2 Forest/Non-forest map", a global forest/non-forest map, with that of a national map containing more detailed LULC classes. By applying the map integration approach at the national level in the Philippines as a case study, we identified 5.937 & PLUSMN; 0.217 Mha of "Missing forest" that were not included in the country's national LULC map, mainly forest patches in areas that were predominantly "Brush/shrub", "Grassland", or "Marshland/swamp" lands. We also identified 4.294 & PLUSMN; 0.258 Mha of land cor-responding to FAO's definition of "Other land with tree cover" that were previously unmapped; specifically, patches of tree cover on predominantly agricultural and urban lands. Based on these additional areas of "Forest" and "Other land with tree cover" identified, we further estimated an additional 145,480 GgCO2/year of carbon sinks. Our approach is generalizable enough to potentially be applied in other countries for more standardized forest and ecosystem services monitoring.

The NH 58 area in India has been experiencing an increase in landslide occurrences, posing significant threats to local communities, infrastructure, and the environment. The growing need to identify areas prone to landslides for effective disaster risk management, land use planning, and infrastructure development has led to the increased adoption of advanced geospatial technologies and statistical methods. In this context, this research article presents an in-depth analysis aimed at developing a landslide susceptibility zonation (LSZ) map for the NH 58 area using remote sensing, GIS, and logistic regression analysis. The study incorporates multiple geo-environmental factors for analysis, such as slope aspect, curvature, drainage density, elevation, fault distance, flow accumulation, geology, geomorphology, land use land cover (LULC), road distance, and slope angle. Utilizing 50% of the landslide inventory data, the logistic regression model was trained to determine correlations between causal factors and landslide occurrences. The logistic regression model was then employed to calculate landslide probabilities for each mapping unit within the NH 58 area, which were subsequently classified into relative susceptibility zones using a statistical class break technique. The model’s accuracy was verified through ROC curve analysis, resulting in a 92% accuracy rate. The LSZ map highlights areas near road cut slopes as highly susceptible to landslides, providing crucial information for land use planning and management to reduce landslide risk in the NH 58 area. The study’s findings are beneficial for policymakers, planners, and other stakeholders involved in regional disaster risk management. This research offers a comprehensive analysis of landslide-influencing factors in the NH 58 area and introduces an LSZ map as a valuable tool for managing and mitigating landslide risks. The map also serves as a critical reference for future research and contributes to the broader understanding of landslide susceptibility in the region.

This study analyzed the spatial-temporal change pattern and underlying factors in production-living-ecological space (PLES) of Nanchong City, China, over the past 20 years using historical land use data (2000, 2010, 2020). A land use transfer matrix was calculated from the historical land use maps, and spatial analysis was conducted to analyze changes in the land use dynamics degree, standard deviation ellipse, and center of gravity. The results showed that there was a rapid spatial evolution of the PLES in Nanchong from 2000 to 2010, followed by a stabilization in the second decade. The transfer of ecological-production space occurred mainly in the Jialing and Yilong River basins, while the reduction of production space and the increase of living space were most prominent in the intersection of three districts (Shunqing, Jialing, and Gaoping districts). The return of production-ecological space was observed in the south and northeast of Yingshan, and there was little notable transfer of other types. The distribution of production space in Nanchong evolved in a north-south to east-west trend, with the center of gravity moving from Yilong to Peng’an County. The living space and production space expanded in a north-south direction, and the center of gravity position was in Nanbu, indicating a more balanced growth or decrease in the last 20 years. The changes in the spatial-temporal pattern of PLES in Nanchong were attributed to the intertwined factors of national policies, economic development, population growth, and the natural environment. This study introduced a novel approach towards rational planning of land resources in Nanchong, which may facilitate more sustainable urban planning and development.

Assessing heat-related health risks is important for sustainable urban development. Although fine-scale infor-mation (e.g., at the community/neighborhood or city block level) is ideal for identifying and mitigating these risks, previous studies have preferred to work at the administrative unit level. High-resolution Local Climate Zone (LCZ) maps, i.e., maps of urban "zones" with different microclimates, could help to standardize the analyzing units. In this study, we proposed an LCZ-based risk assessment approach for this purpose. First, an LCZ map of the study site (Changzhou, China) was generated using multisource big data and machine-learning techniques. Next, Crichton's Risk Triangle framework, based on the hazard-exposure-vulnerability risk compo-nents, was employed to estimate heat-related health risks. Finally, the relationship between LCZ types and heat -related health risk levels was quantitatively analyzed in detail. The results indicated that at least 60% of LCZ1-5 (compact high-/mid-/low-rise, open high-/mid-rise areas) were designated as high-risk areas, while heat hazard mitigation and climate adaptation strategies in urban planning would benefit more from LCZ 6 (open low-rise). This study, based on the LCZ concept, shows the risk difference at the community level, and can be used for informing and implementing area-level urban planning strategies. It could contribute to global heat-related health risk analysis, since the LCZ is a globally consistent system for urban microclimate analysis.

Illegal sand mining has been identified as a significant cause of harm to riverbanks, as it leads to excessive removal of sand from rivers and negatively impacts river shorelines. This investigation aimed to identify instances of shoreline erosion and accretion at illegal sand mining sites along the Chambal River. These sites were selected based on a report submitted by the Director of the National Chambal Sanctuary (NCS) to the National Green Tribunal (NGT) of India. The digital shoreline analysis system (DSAS v5.1) was used during the elapsed period from 1990 to 2020. Three statistical parameters used in DSAS—the shoreline change envelope (SCE), endpoint rate (EPR), and net shoreline movement (NSM)—quantify the rates of shoreline changes in the form of erosion and accretion patterns. To carry out this study, Landsat imagery data (T.M., ETM+, and OLI) and Sentinel-2A/MSI from 1990 to 2020 were used to analyze river shoreline erosion and accretion. The normalized difference water index (NDWI) and modified normalized difference water index (MNDWI) were used to detect riverbanks in satellite images. The investigation results indicated that erosion was observed at all illegal mining sites, with the highest erosion rate of 1.26 m/year at the Sewarpali site. On the other hand, the highest accretion was identified at the Chandilpura site, with a rate of 0.63 m/year. We observed significant changes in river shorelines at illegal mining and unmined sites. Erosion and accretion at unmined sites are recorded at −0.18 m/year and 0.19 m/year, respectively, which are minor compared to mining sites. This study’s findings on the effects of illegal sand mining on river shorelines will be helpful in the sustainable management and conservation of river ecosystems. These results can also help to develop and implement river sand mining policies that protect river ecosystems from the long-term effects of illegal sand mining.