D2Mamba: A mamba-based method for floodway obstructions segmentation from multispectral satellite imagery

Saltmarshes provide key ecosystem services, including atmospheric CO2 sequestration and nitrogen burial in sediments. In recent decades, these blue carbon ecosystems have faced significant degradation from natural and anthropogenic stressors. In this study, rewilding of a desiccated saltmarsh in Cadiz Bay (SW Spain) was assessed as a nature-based solution to restore carbon (Corg) and nitrogen (NT) storage. The rewilding process began in 2004 after breaching an external tidal wall. We evaluated changes in vegetated and unvegetated areas using Landsat satellite imagery (1994-2024) and quantified Corg and NT stocks and burial rates in wild and rewilded sediments, including vegetated saltmarsh (Sarcocornia sp.) and bare sediments colonized by microphytobenthos (MPB). Vegetated saltmarsh cover increased by 85% over 20 years, at an average recovery rate of 5hay-1, concurrent with a decrease in unvegetated tidal flats. Average Corg stocks in the top 1m ranged from 32 to 57t Corg ha-1, with higher values in vegetated sediments. However, only 5-12% of Corg was stored during the rewilding period. Corg burial rates averaged 69g Corg m-2 y-1, and NT stocks were 55% higher in rewilded sediments than in wild ones (3.6 vs. 1.6t NT ha-1). Despite vegetation recovery, burial rates of Corg and NT did not increase clearly, suggesting that long-term storage may be influenced by factors beyond rewilding. Less than 8% of sedimentary Corg originated from saltmarsh vegetation, indicating the dominance of allochthonous sources. These findings highlight the complexity of biogeochemical recovery in rewilded saltmarshes and underscore the need for long-term monitoring to determine how much time is truly required for Corg and NT recovery.

Ancient populations inhabited and transformed neotropical forests, yet the spatial extent of their ecological influence remains underexplored at high resolution. Here we present a deep learning and remote sensing based approach to estimate areas of pre-Columbian forest modification based on modern vegetation. We apply this method to high-resolution satellite imagery from the Sierra Nevada de Santa Marta, Colombia, as a demonstration of a scalable approach, to evaluate palm tree distributions in relation to archaeological infrastructure. Our findings document a non-random spatial association between archaeological infrastructure and contemporary palm concentrations. Palms were significantly more abundant near archaeological sites with large infrastructure investment. The extent of the largest palm cluster indicates that ancient human-managed areas linked to major infrastructure sites may be up to two orders of magnitude bigger than indicated by current archaeological evidence alone. These patterns are consistent with the hypothesis that past human activity may have influenced local palm abundance and potentially reduced the logistical costs of establishing infrastructure-heavy settlements in less accessible locations. More broadly, our results highlight the utility of palm landscape distributions as an interpretable signal within environmental and multispectral datasets for constraining predictive models of archaeological site locations.

ABSTRACT Wetlands provide various goods and services, including habitat, medicine, food, flood mitigation, and other local and global environmental benefits. However, the ecosystem functions of wetlands have been at risk due to human interference and climate change. This review article highlights the application of remote sensing techniques in wetland monitoring and mapping. It emphasizes the use of satellite imagery and the advancement of classification methods to delineate wetland boundaries and assess land‐use dynamics and vegetation health. We observed a significant study on wetland ecosystem mapping with Landsat and Sentinel satellite data. Despite advancements in various remote sensing tools and techniques for wetland monitoring, challenges persist, including cloud cover interference (primarily during rainy seasons), limited access to open data, the need for high‐resolution data and the requirement for more accurate classification methods. Moreover, this review highlights the gaps in current remote sensing applications and suggests future research directions to improve wetland ecosystem management and conservation plans.

In precision agriculture, the assessment and estimation of key crop parameters are crucial aspects for the optimisation of input usage and, as an ultimate goal, for the improvement of yield quality and quantity. In this context, a reliable prediction of yield by remotely sensed imagery is an enabling technology for optimisation. In this work, an innovative method for estimating yield in maize cultivation is presented, which exploits multi-temporal and multispectral Sentinel-2 satellite imagery with supervised Machine Learning (ML) techniques. For model training and validation, yield ground truth experimental data from combine harvesters was used, enabling the yield estimation at sub-field scale. The investigation, which was conducted on five case study plots, involved a preliminary comparison of four ML-based algorithms, trained with raw spectral bands. An assessment of the effect of the training dataset on the yield prediction accuracy was then performed. A set of Vegetation Indices (VIs) and Two Band Indices (TBIs) was also considered for this purpose. Finally, a multi-temporal analysis was conducted, in which the temporal evolution of crop spectral data over the maize growing season was exploited using imageries acquired in different epochs. The obtained results proved that an accurate estimation of maize yield can be reached using a Gaussian process regression model, exploiting multi-temporal features directly provided by the raw spectral bands. The model showed a high accuracy in the estimation of maize yield, even when fed with data acquired during only the maize vegetative phase, thus proving its capacity as a prediction tool. • Supervised machine learning techniques are used to estimate maize yield. • Combine harvester ground truth data enables prediction at subfield scale. • Multi-temporal imageries from Sentinel-2 improve the estimation. • Gaussian Process Regression algorithms reach accuracies up to R 2 > 0.9

Monitoring seasonal grass cover dynamics in a multi-use savanna rangeland is important for sustaining the coexistence of livestock and wildlife. Moreover, population growth is driving increased livestock production, which further limits resources for both livestock and wildlife. To better understand the effect of grazing on grass cover dynamics, we developed a multi-scale remote sensing approach to study the monthly variation in grass cover in two types of conservation areas: a wildlife sanctuary and a communal livestock grazing and wildlife conservancy. The study was carried out in a semi-arid region in Kenya during an exceptionally dry year of 2022 when grazing resources were limited. The Excess Greenness color index was first used to develop a model predicting green fractional vegetation cover (fCover) of field photographs. This model was then applied to upscale fCover to the landscape level using very-high-resolution Pléiades satellite data. The resulting fCover maps were subsequently used to predict grass cover from medium-resolution Sentinel-2 multispectral satellite imagery using Random Forest machine learning. The final model showed high predictive power of grass cover in May (R 2 = 0.96, root mean square error (RMSE) = 4.95%), while predictions were less accurate yet promising for January (R 2 = 0.67, RMSE = 7.1%). The monthly grass cover maps demonstrate differences between the two conservation areas; the grazing area experienced low grass cover throughout the year, whereas grass cover in the wildlife sanctuary was more driven by rainfall. The results demonstrate the usability of digital cameras as the basis for vegetation cover models. Furthermore, this method can be used for adaptive land management to monitor within-season resources for both livestock and wildlife. • Fractional grass cover can be upscaled from field to regional levels • Training data from one month was used to predict monthly variations in grass cover • Areas with higher grazing pressure had lower grass cover throughout the year • The wildlife sanctuary experienced a more natural fluctuation in grass cover

Abstract Cloud occlusion is a pervasive issue causing data gaps in optical satellite imagery, limiting spatiotemporal analyses of marine biogeochemical variables. In this study, we develop a dynamic masking learning strategy integrated into deep learning frameworks to address this challenge. During model training, cloud-mimicking occlusion masks are applied to target frames to simulate real-world data degradation, forcing the model to reconstruct complete images from impaired inputs—without reliance on external pseudo‑labels, multi‑source fusion, or auxiliary variables. We implement this strategy with a U-Net to perform end‑to‑end reconstruction of Himawari‑8 chlorophyll‑a (Chl-a) concentration datasets over the Bohai and Yellow Sea (BYS) for 2015–2022. Validation results demonstrate high‑fidelity reconstruction performance, with a mean squared error (MSE) of 1.962 and coefficient of determination (R²) of 0.942. Comparative experiments show the U-Net consistently outperforms the Data-Interpolating Convolutional Autoencoder (DINCAE) under identical masking protocols. Relative to the publicly available National Oceanic and Atmospheric Administration (NOAA) Data Interpolating Empirical Orthogonal Functions (DINEOF) gap-filled product, our reconstructed dataset exhibits superior quality—preserving a broader dynamic range and higher percentiles, while maintaining the inherent variability that the NOAA product severely underestimates. The complete, high-quality reconstructed Chl-a dataset enables robust spatiotemporal analyses. Seasonal-Trend decomposition using Loess (STL) and Pettitt change-point testing identify a statistically significant downward trend commencing 12 May 2021, alongside consistent seasonal variability and a prominent nearshore-to-offshore concentration gradient. This framework exhibits strong generality and is readily adaptable to other optical remote sensing reconstruction tasks for geophysical variables.

Rapid urbanization in sub-Saharan Africa poses significant challenges to local food security through the conversion of productive agricultural lands to built-up areas. Kinondoni District in Dar es Salaam has experienced substantial urban growth over the past three decades, yet the implications for food security remain poorly understood.This study analyzed land use and land cover (LULC) changes in Kinondoni District between 1993 and 2023 to assess the impact of urbanization on agricultural land availability and its implications for local food security. Multi-temporal satellite imagery analysis was conducted using Landsat data from four epochs (1993, 2003, 2013, and 2023). Land use was classified into seven classes: water bodies, mangrove forests, bare areas, built-up areas, shrublands, and cultivated lands. Change detection analysis quantified the magnitude and direction of land use transitions over the 30-year period using supervised Maximum Likelihood Classification with overall accuracy ranging from 85-92% and Kappa coefficients of 0.82-0.89. The analysis revealed dramatic urbanization, with built-up areas expanding from 124.20 ha (2.23%) in 1993 to 1,905.14 ha (34.3%) in 2023 a 1,433% increase. Concurrently, cultivated land decreased by 96.6%, from 172.90 ha (3.11%) to 5.88 ha (0.1%), representing a loss of 167.02 ha. Mangrove forests declined by 40.1% (737.62 ha), shrublands decreased by 73.9% (546.41 ha), and bare areas reduced by 77.9% (402.11 ha). Water bodies remained relatively stable, increasing slightly by 3.3% (72.22 ha). The majority of agricultural land (85.2%) was directly converted to built-up areas. The near-complete elimination of cultivated land in Kinondoni District indicates a critical threat to local food production capacity and urban food security. The rapid conversion of agricultural and natural lands to urban infrastructure necessitates urgent policy interventions, including peri-urban agricultural planning, vertical farming initiatives, and regional food supply chain development to ensure food security for Kinondoni's growing population.

Daily estimation of Water–Salt–Carbon coordination in high-standard cotton fields using uav and PlanetScope satellite image fusion

Future risks of cyanobacterial blooms in lakes unveiled by open access data and integrated machine learning models.

Remote sensing plays an increasingly critical role in water quality monitoring due to its capacity for consistent observations on both large and small water bodies. However, current remote sensing approaches face limitations in aligning satellite observations with in-situ measurements, largely due to the dynamic vertical behavior of algae and the temporal constraints of satellite overpasses. Consequently, many studies rely on large water bodies, space–time substitution, or opportunistic imaging of blooms, which restricts the applicability of remote sensing for routine monitoring tasks such as periodic chlorophyll-a (Chl-a) estimation. With near-daily global coverage, PlanetScope imagery presents new opportunities to overcome these constraints. In this study, we propose a novel field-sampling augmentation framework that integrates satellite observations with in-situ data by modeling the diurnal vertical migration of algae through an Algal Behavior Function (ABF). This function enables the temporal adjustment of in-situ measurements, generating refined field-to-satellite matchups that enhance the robustness of Chl-a estimation models. We applied this method using PlanetScope imagery from 2022 to 2023 and co-located sonde measurements, incorporating vertical profile and timestamp information to correct for field-to-satellite temporal mismatches at two lakes in Ohio (Grand Lake St. Marys, samples = 84, Del-Co reservoirs, samples = 333). The augmented model improved Chl-a prediction accuracy (RMSE reduce) by 5.8%-18.0% compared to baseline models without refinement, with notable gains during non-bloom periods, offering potential for earlier bloom detection. Furthermore, the ABF demonstrated moderate geographic transferability: models using ABFs derived from a reservoir successfully improved Chl-a predictions at two additional lakes located 156 km (western Lake Erie) and 383 km (Saginaw Bay, Lake Huron) away, with accuracy gains (RMSE reduce) of 28.5%-35.3%. Collectively, these results position ABF as a practical, sensor-agnostic pre-processing step that can be embedded in operational workflows to improve high-resolution Chl-a retrievals, enable earlier harmful algal bloom alerts, and support cross-basin trend analyses for management.

Metadata, wavelet, and time aware diffusion models for satellite image super resolution

Abstract This paper proposes a parametric deconstruction-based model for evaluating the imaging capabilities of remote sensing satellites. It categorizes satellites into two types: optical imaging and synthetic aperture radar (SAR) imaging. The model identifies key influencing factors and establishes an impact assessment framework based on parametric deconstruction. It evaluates the effectiveness of target detection, recognition, and positioning for different remote sensing imagers individually. Correction terms are incorporated to account for spatial environmental conditions and imaging data downlink constraints. The “Zhuhai-1” remote sensing imaging constellation is used as a case study to validate the model. Results confirm the model’s effectiveness and robustness. This model provides a quantitative reference for comprehensively assessing remote sensing satellite imaging capabilities and supports the development of remote sensing imaging technologies.

Lightweight Cryptographic Approach: Pixel-Wise Adaptive Diffusion and Spatial Block Permutation for Securing Satellite Images

The Ofin River basin has historically supported Simulium blackfly breeding, vectors of Onchocerca, yet recent Programme reports have noted a sharp decline in monitored populations. With illegal artisanal mining (galamsey) expanding around the basin, this study assessed how associated ecological changes may influence habitat suitability for blackfly breeding across three riverine communities (Adwuman, Buabenso, and Kyekyewere). Water discharge and quality were assessed through field measurements and laboratory analyses of pH, electrical conductivity, turbidity, colour, total suspended solids (TSS), and total dissolved solids (TDS). Satellite imagery from 2008, 2017, and 2022/23 was analysed using Iso Cluster Unsupervised Classification algorithm (ISODATA), Principal Raster Components Analysis (PRCA) and the Normalized Difference Vegetation Index (NDVI) to quantify transitions and vegetation health using ArcGIS Pro. Water discharge rates varied (Adwuman: 181.57m3/s, Buabenso: 78.93m3/s, Kyekyewere: 111.95m3/s) and quality analysis showed differences in key parameters. Adwuman's pH was 6.98, conductivity (145.5µS/cm), turbidity (3392.5 NTU), colour (3375Hz) and TSS (3630mg/L). Buabenso had a pH of 6.98, conductivity 146.75µS/cm, turbidity 3525 NTU, colour at 3812.5Hz and TSS of 3857.5mg/L. Kyekyewere recorded the lowest pH (6.95) and conductivity (145.25µS/cm), but the highest turbidity (3725 NTU), colour (4175Hz) and TSS (4342.5mg/L). Forest cover declined by 10.72, 7.41, and 8.80 percentage points in Adwuman, Buabenso, and Kyekyewere, respectively, while light vegetation increased by 15.71, 15.00, and 18.93 points. Water coverage expanded by 10.81, 6.12, and 5.26 percentage points across the communities, indicating hydrological alteration. NDVI revealed widespread declines in vegetation health and density, particularly near mining zones. The combined effects of extreme sedimentation, vegetation degradation and riparian disturbance suggest ecological conditions that are increasingly unsuitable for blackfly breeding in the Ofin River basin. The disruptions also threaten food security, clean water access and ecosystem integrity, with implications for achieving SDG 2, 3, 6, and 15. Strengthened River management, reforestation of degraded riparian areas, enforcement against illegal mining and community-based monitoring are needed to restore ecological function and safeguard both biodiversity and public health.

Understanding the spatiotemporal dynamics of Land Use and Land Cover (LULC), caused by rapid urbanization, is crucial to mitigating the negative impacts of urban growth, threatening agricultural sustainability and ecological balance. This research was conducted on the Lower Turag River Basin in the northwestern region of Dhaka, Bangladesh and aims to quantify the built-up area changes along with population changes from 2001 to 2021, evaluating the nature of urban expansion by calculating the ratio between Land Consumption Rate and Population Growth Rate (LCRPGR) following the Sustainable Development Goal (SDG) metadata used for the indicator 11.3.1. Here, satellite imagery, including Landsat 5, Landsat 8, and Shuttle Radar Topography Mission (SRTM) data, was used in the Google Earth Engine (GEE) platform, where the Random Forest algorithm was incorporated to classify LULC into five distinct categories. The high kappa values (>0.87) ensured the accuracy and the reliability of the classification. The results show that built-up areas increased by 139.39%, while the total population grew by 163.71% between 2001 and 2021, largely replacing agricultural land and natural vegetation. The urban expansion outpaced population growth, with a land consumption to population growth ratio of 1.775 for the period of 2001 to 2011 and 0.513 for the period of 2011 to 2021, indicating critical urban expansion. Besides, the LULC patterns found that the expansion occurred towards the north of the Turag basin with high growth of population, which increased the land consumption. Furthermore, the findings of this study will help policymakers and researchers to understand the spatio-temporal changes of urban expansion, population growth, and land consumption with a sustainable policy framework.

During the current study, the causes, impacts, and mitigation efforts related to Glacial Lake Outburst Floods (GLOFs) in Bilhanz and Badswat, Ishkoman Valley, were thoroughly examined between July to September 2024. These events led to significant landscape alterations, transforming once fertile agricultural and forest lands into debris covered areas, as evidenced by satellite and digital camera imagery. A large lake formed in Badswat, destroying extensive agricultural and forest land, which forced the relocation of many residents. Most displaced individuals received housing support from the Aga Khan Agency for Habitat (AKAH), while the government provided limited assistance. The Saudi government also made notable contributions. Between 2018 and 2021, successive GLOF events damaged 186 hectares of agricultural land in both villages in addition to damages and losses to livestock and other life supporting systems. Climate change and deforestation were identified as the primary drivers of these disasters. A multivariate analysis revealed that gender, age, and education had significant impacts on awareness and perceptions of environmental degradation and climate change. Older respondents, particularly those aged 50-60, were the most influential in shaping future disaster preparedness and mitigation efforts. Although education had a smaller effect size, it still played a crucial role in raising environmental awareness. Mitigation strategies proposed include afforestation, carbon capture, renewable energy, and waste reduction. Many respondents emphasized the importance of community involvement and education in addressing environmental issues. However, opinions vary on the most effective methods for controlling soil pollution and improving overall environmental health.

Publicly available satellite imagery supports reliable waterbody extraction, which is crucial for hydrological studies, environmental monitoring, resource management, disaster management, urban heat reduction planning, etc. This study focuses on the extraction of inland waterbodies of Hakaluki Haor using Sentinel-1 and Sentinel-2 imagery and compares their results during the pre-monsoon (March) and post-monsoon (November) seasons from 2019 to 2024. The waterbodies were delineated using a threshold-based approach for Sentinel-1 and the Normalized Difference Water Index (NDWI) for Sentinel 2. The results reveal seasonal and inter-annual variability, with minimal pre-monsoon water observed in 2021–2022 and exceptionally high post-monsoon inundation in 2020, when Sentinel-2 detected approximately 122.4 km² compared to 95 km² of water extent from Sentinel-1. The analysis further shows that during the pre-monsoon season, waterbody extraction overlapping area between these two sensors varies from 63-91% due to scattered and fragmented waterbodies, whereas during post-monsoon, this variation is significantly reduced to 81-95%. It indicates that during post-monsoon, due to continuous waterbodies, the sensor choice has less effect on waterbody extraction in a similar context.

The human impact, combined with the geomorphological, lithological, and climatic specificities of northern Morocco, makes the soils highly vulnerable to the risk of water erosion. The primary purpose of this research is to map and assess the soil water erosion vulnerability in the Boussekour watershed, located in the Rif Mountain chain in northern Morocco, by applying the erosion potential method (EPM) developed by Gavrilović. It is an approach that involves integrating, within a Geographic Information System (GIS) environment, six parameters involved in the erosive phenomenon: temperature, rainfall, slope, soil erodibility, current erosion, and soil protection. The input data necessary for applying this model were derived from satellite images, digital elevation models (DEMs), granulometric and physico-chemical analyses of soil samples, and local rainfall data. Results show that the high erosion class (from 20 to 40 Mg∙ha−1∙yr−1) is prevalent and acts on 45% of the watershed, while the low (from 5 to 10 Mg∙ha−1∙yr−1) and very low (from 0 to 5 Mg∙ha−1∙yr−1) erosion classes account for just 6.98% and 1.44% of the area, respectively. This result demonstrates that the Boussekour watershed is profoundly threatened by soil water erosion, and an effort ought to be made to promote operative soil conservation efforts.

The article presents an analytical review of modern mathematical models for predicting the spread of landscape grass fires in steppe zones, where such fires are characterized by high dynamics and destructive force due to the peculiarities of the grassy cover of open landscapes. The main classes of models are considered: semi-empirical, based on a combination of physical principles and empirical data; simulation, implementing a space-time simulation of the fire front.; as well as machine learning-based approaches, including regression algorithms and convolutional neural networks for analyzing large amounts of data (weather parameters, satellite images, historical events). The methodological foundations, advantages, limitations and areas of application of models are analyzed, taking into account the specifics of steppe (grass) fires — a high rate of spread, strong dependence on wind, the degree of drying of grass and the topography of open landscapes (slope, exposure of slopes). Special attention is paid to the integration of global digital SRTM terrain models to accurately account for slope slope and exposure, which corrects the fire propagation rate in the models, as well as data from unmanned aerial systems for rapid assessment of vegetation indices (NDVI, biomass, fuel moisture) with high resolution, which improves the accuracy of fuel maps and forecasting. The trends of pyrological activity growth in the steppe regions of Russia under the influence of climatic changes have been revealed. The prospects of combined use of models for operational forecasting and extinguishing management in the conditions of the Russian Ministry of Emergency Situations, including hybrid approaches with integration of SRTM and UAS to minimize risks, are described.