Search

Winter wheat is one of the major cereal crops globally and one of the top three cereal crops in China. The precise forecasting of the yield of winter wheat holds significant importance in the realms of agricultural management and ensuring food security. The use of multi-temporal remote sensing data for crop yield prediction has gained increasing attention. Previous research primarily focused on utilizing remote sensing data from individual or a few growth stages as input parameters or integrated data across the entire growth period. However, a detailed analysis of the impact of different temporal combinations on the accuracy of yield prediction has not been extensively reported. In this study, we optimized the temporal sequence of growth stages using interpolation methods, constructed a yield prediction model incorporating the enhanced vegetation index (EVI) at different growth stages as input parameters, and employed a random forest (RF) algorithm. The results indicated that the RF model utilizing the EVI from all the temporal combinations throughout the growth period as input parameters accurately predicted the winter wheat yield with an R2 of the calibrated dataset exceeding 0.58 and an RMSE less than 1284 kg/ha. Among the 1023 yield models tested in this study with ten different growth stage combinations, the most accurate temporal combination comprised five stages corresponding to the regreening, erecting, jointing, heading, and filling stages, with an R2 of 0.81 and an RMSE of 1250 kg/ha and an NRMSE of 15%. We also observed a significant decrease in estimation accuracy when the number of growth stages was fewer than five and a certain degree of decline when the number exceeded five. Our findings confirmed the optimal number and combination of growth stages for the best yield prediction, providing substantial insights for winter wheat yield forecasting. [ABSTRACT FROM AUTHOR]

In addressing the challenges of quantitative precipitation estimation (QPE) using weather radar, the importance of enhancing the rainfall estimates for applications such as flash flood forecasting and hydropower generation management is recognized. This study employed dual-polarization weather radar data to refine the traditional Z–R relationship, which often needs higher accuracy in areas with complex meteorological phenomena. Utilizing tree-based machine learning algorithms, such as random forest and gradient boosting, this research analyzed polarimetric variables to capture the intricate patterns within the Z–R relationship. The results highlight machine learning's potential to improve the precision of precipitation estimation, especially under challenging weather conditions. Integrating meteorological insights with advanced machine learning techniques is a remarkable achievement toward a more precise and adaptable precipitation estimation method. [ABSTRACT FROM AUTHOR]

Accurate spatiotemporal monitoring and modeling of soil moisture (SM) is of paramount importance for various applications ranging from food production to climate change adaptation. This study deals with modeling SM with the random forest (RF) algorithm using datasets comprising multispectral data from Sentinel-2, Landsat-8/9, and hyperspectral data from the CoSpectroCam sensor (CSC, licensed to AgriWatch BV, Enschede, The Netherlands) mounted on an unmanned aerial vehicle (UAV) in Iran. The model included nine bands from Landsat-8/9, 11 bands from Sentinel-2, and 1252 bands from the CSC (covering the wavelength range between 420 and 850 nm). The relative feature importance and band sensitivity to SM variations were analyzed. In addition, four indices, including the perpendicular index (PI), ratio index (RI), difference index (DI), and normalized difference index (NDI) were calculated from the different bands of the datasets, and their sensitivity to SM was evaluated. The results showed that the PI exhibited the highest sensitivity to SM changes in all datasets among the four indices considered. Comparisons of the performance of the datasets in SM estimation emphasized the superior performance of the UAV hyperspectral data (R2 = 0.87), while the Sentinel-2 and Landsat-8/9 data showed lower accuracy (R2 = 0.49 and 0.66, respectively). The robust performance of the CSC data is likely due to its superior spatial and spectral resolution as well as the application of preprocessing techniques such as noise reduction and smoothing filters. The lower accuracy of the multispectral data from Sentinel-2 and Landsat-8/9 can also be attributed to their relatively coarse spatial resolution compared to the CSC, which leads to pixel non-uniformities and impurities. Therefore, employing the CSC on a UAV proves to be a valuable technology, providing an effective link between satellite observations and ground measurements. [ABSTRACT FROM AUTHOR]

Digital elevation models (DEMs) are widely used in digital terrain analysis, global change research, digital Earth applications, and studies concerning natural disasters. In this investigation, a thorough examination and comparison of five open-source DEMs (ALOS PALSAR, SRTM1 DEM, SRTM3 DEM, NASADEM, and ASTER GDEM V3) was carried out, with a focus on the Chongqing region as a specific case study. By utilizing ICESat-2 ATL08 data for validation and employing a random forest model to refine terrain variables such as slope, aspect, land cover, and landform type, a study was undertaken to assess the precision of DEM data. Research indicates that spatial resolution significantly impacts the accuracy of DEMs. ALOS PALSAR demonstrated satisfactory performance, reducing the corrected root mean square error (RMSE) from 13.29 m to 9.15 m. The implementation of the random forest model resulted in a significant improvement in the accuracy of the 30 m resolution NASADEM product. This improvement was supported by a decrease in the RMSE from 38.24 m to 9.77 m, demonstrating a significant 74.45% enhancement in accuracy. Consequently, the ALOS PALSAR and NASADEM datasets are considered the preferred data sources for mountainous urban areas. Furthermore, the study established a clear relationship between the precision of DEMs and slope, demonstrating a consistent decline in precision as slope steepness increases. The influence of aspect on accuracy was considered to be relatively minor, while vegetated areas and medium-to-high-relief mountainous terrains were identified as the main challenges in attaining accuracy in the DEMs. This study offers valuable insights into selecting DEM datasets for complex terrains in mountainous urban areas, highlighting the critical importance of choosing the appropriate DEM data for scientific research. [ABSTRACT FROM AUTHOR]

The European Mediterranean Basin (Euro-Med), a region particularly vulnerable to global warming, notably lacks research aimed at assessing and enhancing the widely used remote climate detection products known as General Circulation Models (GCMs). In this study, the proficiency of GCMs in replicating reanalyzed 1981–2010 temperature data sourced from the ERA5 Land was assessed. Initially, the least data-modifying interpolation method for achieving a resolution match of 0.1° was ascertained. Subsequently, a pixel-by-pixel evaluation was conducted, employing five goodness-of-fit metrics. From these metrics, we compiled a Comprehensive Rating Index (CRI). A Multi-Model Ensemble using Random Forest was constructed and projected across three emission scenarios (SSP1-RCP2.6, SSP2-RCP4.5, and SSP5-RCP8.5) and timeframes (2026–2050, 2051–2075, and 2076–2100). Empirical Bayesian Kriging, selected for its minimal data alteration, supersedes the commonly employed Bilinear Interpolation. The evaluation results underscore MPI-ESM1-2-HR, GFDL-ESM4, CNRM-CM6-1, MRI-ESM2-0, CNRM-ESM2-1, and IPSL-CM6A-LR as top-performing models. Noteworthy geospatial disparities in model performance were observed. The projection outcomes, notably divergent from IPCC forecasts, revealed a warming trend of 1 to over 2 °C less than anticipated for spring and winter over the medium–long term, juxtaposed with heightened warming in mountainous/elevated regions. These findings could substantially refine temperature projections for the Euro-Med, facilitating the implementation of policy strategies to mitigate the effects of global warming in vulnerable regions worldwide. [ABSTRACT FROM AUTHOR]

Wetlands are integral components of agricultural landscapes, providing a wide range of ecological, economic, and social benefits essential for sustainable development and rural livelihoods. Globally, they are vulnerable ecological assets facing several significant threats including water extraction and regulation, land clearing and reclamation, and climate change. Classification and mapping of wetlands in agricultural landscapes is crucial for conserving these ecosystems to maintain their ecological integrity amidst ongoing land-use changes and environmental pressures. This study aims to establish a robust framework for wetland classification and mapping in intensive agricultural landscapes using time series of Sentinel-2 imagery, with a focus on the Gwydir Wetland Complex situated in the northern Murray–Darling Basin—Australia's largest river system. Using the Google Earth Engine (GEE) platform, we extracted two groups of predictors based on six vegetation indices time series calculated from multi-temporal Sentinel-2 surface reflectance (SR) imagery: the first is statistical features summarizing the time series and the second is phenological features based on harmonic analysis of time series data (HANTS). We developed and evaluated random forest (RF) models for each level of classification with combination of different groups of predictors. Our results show that RF models involving both HANTS and statistical features perform strongly with significantly high overall accuracy and class-weighted F1 scores (p < 0.05) when comparing with models with either statistical or HANTS variables. While the models have excellent performance (F-score greater than 0.9) in distinguishing wetlands from other landcovers (croplands, terrestrial uplands, and open waters), the inter-class discriminating power among wetlands is class-specific: wetlands that are frequently inundated (including river red gum forests and wetlands dominated by common reed, water couch, and marsh club-rush) are generally better identified than the ones that are flooded less frequently, such as sedgelands and woodlands dominated by black box and coolabah. This study demonstrates that HANTS features extracted from time series Sentinel data can significantly improve the accuracy of wetland mapping in highly fragmentated agricultural landscapes. Thus, this framework enables wetland classification and mapping to be updated on a regular basis to better understand the dynamic nature of these complex ecosystems and improve long-term wetland monitoring. [ABSTRACT FROM AUTHOR]

The West Sahel is facing significant threats to its vegetation and wildlife due to the land degradation and habitat fragmentation. It is crucial to assess the regional vegetation greenness dynamics in order to comprehensively evaluate the effectiveness of protection in the natural reserves. This study analyzes the vegetation greenness trends and the driving factors in the Dosso Partial Faunal Reserve in Niger and nearby unprotected regions—one of the most important habitats for endemic African fauna—using satellite time series data from 2001 to 2020. An overall vegetation browning trend was observed throughout the entire region with significant spatial variability. Vegetation browning dominated in the Dosso Reserve with 17.7% of the area showing a significant trend, while the area with significant greening was 6.8%. In a comparison, the nearby unprotected regions to the north and the east were found to be dominated by vegetation browning and greening, respectively. These results suggest that the vegetation protection practice was not fully effective throughout the Dosso Reserve. The dominant drivers were also diagnosed using the Random Forest model-based method and the Partial Dependence Plot tool, showing that water availability (expressed as soil moisture) and land use/land cover change were the most critical factors affecting vegetation greenness in the study region. Specifically, soil moisture stress and specific land management practices associated with logging, grazing, and land clearing appeared to dominate vegetation browning in the Dosso Reserve. In contrast, the vegetation greening in the central Dosso Reserve and the nearby unprotected region to the east was probably caused by the increase in shrubland/forest, which was related to the effective implementation of protection. These findings improve our understanding of how regional vegetation greenness dynamics respond to environmental changes in the Dosso Reserve and also highlight the need for more effective conservation planning and implementation to ensure sustainable socio-ecological development in the West Sahel. [ABSTRACT FROM AUTHOR]

The transboundary region along the Heilongjiang River, encompassing the Russian Far East and Northeast China, possesses abundant agricultural natural resources crucial for global food security. In the face of the challenge of disruptions in the global food supply chain, the precise monitoring and exploitation of agricultural resources in the Heilongjiang Basin becomes imperative. This study employed deep learning to classify crop status in 2023 in the Heilongjiang Basin using Sentinel-2 satellite remote sensing images at a 10 m resolution. Various vegetation indices, including the Normalized Difference Vegetation Index (NDVI), the Normalized Difference Water Index (NDWI), the Enhanced Vegetation Index (EVI), the Modified Soil Adjusted Vegetation Index (MSAVI), and others, were computed and analyzed for different crops. The Google Earth Engine (GEE) platform was utilized for validation point sampling based on plot objects. The random forest (RF) classification method was successfully employed to classify and identify major crops in the study area (wheat, maize, rice, and soybean), as well as wetlands, tree cover, grassland, water, and constructed land, with an overall classification accuracy of 86%. Tree cover dominated the land cover, constituting 62%, while wheat, maize, rice, and soybeans accounted for 7% of the total area. Of these, soybeans occupied the largest area (57,646.60 hectares), followed by rice (53,209.53 hectares), maize (39,998.37 hectares), and wheat (8782.31 hectares). This study demonstrated that sample selection based on plot objects facilitates efficient sample labeling, providing insights into crop classification in other, potentially larger, areas. This method simultaneously distinguishes wetland, cultivated land, and forest features, supporting further integrated investigations for more natural resources. [ABSTRACT FROM AUTHOR]

Quaking aspen is an important deciduous tree species across interior western U.S. forests. Existing maps of aspen distribution are based on Landsat imagery and often miss small stands (<0.09 ha or 30 m2), which rapidly regrow when managed or following disturbance. In this study, we present methods for deriving a new regional map of aspen forests using one year of Sentinel-1 (S1) and Sentinel-2 (S2) imagery in Google Earth Engine. Using observed annual phenology of aspen across the Southern Rockies and leveraging the frequent temporal resolution of S1 and S2, ecologically relevant seasonal imagery composites were developed. We derived spectral indices and radar textural features targeting the canopy structure, moisture, and chlorophyll content. Using spatial block cross-validation and Random Forests, we assessed the accuracy of different scenarios and selected the best-performing set of features for classification. Comparisons were then made with existing landcover products across the study region. The resulting map improves on existing products in both accuracy (0.93 average F1-score) and detection of smaller forest patches. These methods enable accurate mapping at spatial and temporal scales relevant to forest management for one of the most widely distributed tree species in North America. [ABSTRACT FROM AUTHOR]

Soil organic carbon (SOC) measurements are an indicator of soil health and an important parameter for the study of land-atmosphere carbon fluxes. Field sampling provides precise measurements at the sample location but entails high costs and cannot provide detailed maps unless the sampling density is very high. Remote sensing offers the possibility to quantify SOC over large areas in a cost-effective way. As a result, numerous studies have sought to quantify SOC using Earth observation data with a focus on inter-field or regional distributions. This study took a different angle and aimed to map the spatial distribution of SOC at the intra-field scale, since this distribution provides important insights into the biophysiochemical processes involved in the retention of SOC. Instead of solely using spectral measurements to quantify SOC, topographic and spectral features act as predictor variables. The necessary data on study fields in South-East England was acquired through a detailed SOC sampling campaign, including a LiDAR survey flight. Multi-spectral Sentinel-2 data of the study fields were acquired for the exact day of the sampling campaign, and for an interval of 18 months before and after this date. Random Forest (RF) and Support Vector Regression (SVR) models were trained and tested on the spectral and topographical data of the fields to predict the observed SOC values. Five different sets of model predictors were assessed, by using independently and in combination, single and multidate spectral data, and topographical features for the SOC sampling points. Both, RF and SVR models performed best when trained on multi-temporal Sentinel-2 data together with topographic features, achieving validation root-mean-square errors (RMSEs) of 0.29% and 0.23% SOC, respectively. These RMSEs are competitive when compared with those found in the literature for similar models. The topographic wetness index (TWI) exhibited the highest permutation importance for virtually all models. Given that farming practices within each field are the same, this result suggests an important role of soil moisture in SOC retention. Contrary to findings in dryer climates or in studies encompassing larger areas, TWI was negatively related to SOC levels in the study fields, suggesting a different role of soil wetness in the SOC storage in climates characterized by excess rainfall and poorly drained soils. [ABSTRACT FROM AUTHOR]

This study addresses a crucial gap in understanding the impact of urban morphologies on the canopy urban heat islands (CUHI) effect. The selection of reference stations lacks a unified standard, and their surface air temperature (SAT) sequences are also inevitably influenced by urbanization. However, synchronous observational data from relocated meteorological stations could provide high-quality sample data for studying CUHI. Utilizing remote sensing techniques, the findings of this paper revealed that the observation environment of stations after relocation exhibited remarkable representativeness, with their observation sequences accurately reflecting the local climatic background. The differences in synchronized observation sequences could characterize the CUHI intensity (CUHII). Among the various factors, land use parameters and landscape parameters played particularly significant roles. Furthermore, the fitting performance of the random forest (RF) model for both training and testing data was significantly superior to that of the linear model and support vector regression (SVR) model. Additionally, the influence of local circulation on CUHI could not be overlooked. The mechanisms by which urban morphologies affect CUHII under different circulation backgrounds deserve further investigation. [ABSTRACT FROM AUTHOR]

Accurate mapping of forest habitats, especially in NATURA sites, is essential information for forest monitoring and sustainable management but also for habitat characterisation and ecosystem functioning. Remote sensing data and spatial modelling allow accurate mapping of the presence and distribution of tree species and habitats and are valuable tools for the long-term assessment of habitat status required by the European Commission. In order to serve the above, the present study aims to propose a methodology to accurately map the spatial distribution of forest habitats in three NATURA2000 sites of Cyprus by employing Sentinel-1 and Sentinel-2 data as well as topographic features using the Google Earth Engine (GEE). A pivotal aspect of the methodology identified was that the best band combination of the Random Forest (RF) classifier achieves the highest performance for mapping the dominant habitats in the three case studies. Specifically, in the Akamas region, eight habitat types have been mapped, in Paphos nine and six in Troodos. These habitat types are included in three of the nine habitat groups based on the EU's Habitat Directive: the sclerophyllous scrub, rocky habitats and caves and forests. The results show that using the RF algorithm achieves the highest performance, especially using Dataset 6, which is based on S2 bands, spectral indices and topographical features, and Dataset 13, which includes S2, S1, spectral indices and topographical features. These datasets achieve an overall accuracy (OA) of approximately 91–94%. In contrast, Dataset 7, which includes only S1 bands and Dataset 9, which combines S1 bands and spectral indices, achieve the lowest performance with an OA of approximately 25–43%. [ABSTRACT FROM AUTHOR]

Precipitation is the basic component of the Earth's water cycle. Obtaining high-resolution and high-precision precipitation data is of great significance. This paper establishes a precipitation retrieval model based on a random forest classification and regression model during the day and at night with FY-4B/AGRI Level1 data on China from July to August 2022. To evaluate the retrieval effect of the model, the GPM IMERG product is used as a reference, and the retrieval results are compared against those of the FY-4B/AGRI operational precipitation product. In addition, the retrieval results are analyzed according to different underlying surfaces. The results showed that compared with the FY-4B/AGRI operational precipitation product, the retrieval model can better identify precipitation and capture precipitation areas of light rain, moderate rain, heavy rain and torrential rain. Among them, the probability of detection (POD) of the day model increased from 0.328 to 0.680, and the equitable threat score (ETS) increased from 0.252 to 0.432. The POD of the night model increased from 0.337 to 0.639, and the ETS score increased from 0.239 to 0.369. Meanwhile, the precipitation estimation accuracy of the day model increased by 38.98% and that of the night model increased by 40.85%. Our results also showed that due to the surface uniformity of the ocean, the model can identify precipitation better on the ocean than on the land. Our findings also indicated that for the different underlying surfaces of the land, there is no significant difference in each evaluation index of the model. This is a strong argument for the universal applicability of the model. Notably, the results showed that, especially for more vegetated areas and areas covered by water, the model is capable of estimating precipitation. In conclusion, the precipitation retrieval model that is proposed herein can better determine precipitation regions and estimate precipitation intensities compared with the FY-4B/AGRI operational precipitation product. It can provide some reference value for future precipitation retrieval research on FY-4B/AGRI. [ABSTRACT FROM AUTHOR]

Climate change and other anthropogenic factors have caused a significant decline in seagrass cover globally. Identifying the specific causes of this decline is paramount if they are to be addressed. Consequently, we identified the causes of long-term change in seagrass/submerged aquatic vegetation (SAV) percentage cover and extent in a marine protected area on Jamaica's southern coast. Two random forest regression (RFr) models were built using 2013 hydroacoustic survey SAV percentage cover data (dependent variable), and auxiliary and 2013 Landsat 7 and 8 reflectance data as the predictors. These were used to generate 24 SAV percentage cover and benthic feature maps (SAV present, absent, and coral reef) for the period 1984–2021 (37 years) from Landsat satellite series reflectance data. These maps and rainfall data were used to determine if SAV extent/area (km2) and average percentage cover and annual rainfall changed significantly over time and to evaluate the influence of rainfall. Additionally, rainfall impact on the overall spatial patterns of SAV loss, gain, and percentage cover change was assessed. Finally, the most important spatial pattern predictors of SAV loss, gain, and percentage cover change during 23 successive 1-to-4-year periods were identified. Predictors included rainfall proxies (distance and direction from river mouth), benthic topography, depth, and hurricane exposure (a measure of hurricane disturbance). SAV area/extent was largely stable, with >70% mean percentage cover for multiple years. However, Hurricane Ivan (in 2004) caused a significant decline in SAV area/extent (by 1.62 km2, or 13%) during 2002–2006, and a second hurricane (Dean) in 2007 delayed recovery until 2015. Additionally, rainfall declined significantly by >1000 mm since 1901, and mean monthly rainfall positively influenced SAV percentage cover change and had a positive overall effect on the spatial pattern of SAV cover percentage change (across the entire bay) and gain (close to the mouth of a river). The most important spatial pattern predictors were the two rainfall proxies (areas closer to the river mouth were more likely to experience SAV loss and gain) and depth, with shallow areas generally having a higher probability of SAV loss and gain. Three hurricanes had significant but different impacts depending on their distance from the southern coastline. Specifically, a hurricane that made landfall in 1988 (Gilbert), resulted in higher SAV percentage cover loss in 1987–1988. Benthic locations with a northwestern/northern facing aspect (the predominant direction of Ivan's leading edge wind bands) experienced higher SAV losses during 2002–2006. Additionally, exposure to Ivan explained percentage cover loss during 2006–2008 and average exposure to (the cumulative impact of) Ivan and Dean (both with tracks close to the southern coastline) explained SAV loss during 2013–2015. Therefore, despite historic lows in annual rainfall, overall, higher rainfall was beneficial, multiple hurricanes impacted the site, and despite two hurricanes in three years, SAV recovered within a decade. Hurricanes and a further reduction in rainfall may pose a serious threat to SAV persistence in the future. [ABSTRACT FROM AUTHOR]

The production of "Nocciola Romana" hazelnuts in the province of Viterbo, Italy, has evolved into a highly efficient and profitable agro-industrial system. Our approach is based on a hierarchical framework utilizing aggregated data from multiple temporal data and sources, offering valuable insights into the spatial, temporal, and phenological distributions of hazelnut crops To achieve our goal, we harnessed the power of Google Earth Engine and utilized collections of satellite images from Sentinel-2 and Sentinel-1. By creating a dense stack of multi-temporal images, we precisely mapped hazelnut groves in the area. During the testing phase of our model pipeline, we achieved an F1-score of 99% by employing a Hierarchical Random Forest algorithm and conducting intensive sampling using high-resolution satellite imagery. Additionally, we employed a clustering process to further characterize the identified areas. Through this clustering process, we unveiled distinct regions exhibiting diverse spatial, spectral, and temporal responses. We successfully delineated the actual extent of hazelnut cultivation, totaling 22,780 hectares, in close accordance with national statistics, which reported 23,900 hectares in total and 21,700 hectares in production for the year 2022. In particular, we identified three distinct geographic distribution patterns of hazelnut orchards in the province of Viterbo, confined within the PDO (Protected Designation of Origin)-designated region. The methodology pursued, using three years of aggregate data and one for SAR with a spectral separation clustering hierarchical approach, has effectively allowed the identification of the specific perennial crop, enabling a deeper characterization of various aspects influenced by diverse environmental configurations and agronomic practices.The accurate mapping and characterization of hazelnut crops open opportunities for implementing precision agriculture strategies, thereby promoting sustainability and maximizing yields in this thriving agro-industrial system. [ABSTRACT FROM AUTHOR]

While protected areas (PAs) are an important conservation strategy to protect vulnerable ecosystems and species, recent analyses question their effectiveness in curbing deforestation and maintaining landscape connectivity. The spatial arrangement of forests inside and outside of PAs may affect ecosystem functioning and wildlife movement. The Osa Peninsula—and Costa Rica in general—are unique conservation case studies due to their high biodiversity, extensive PA network, environmental policies, and payment for ecosystem services (PES) programs. This study explores the relationship between forest management initiatives—specifically PAs, the 1996 Forest Law, and PES—and forest cover and landscape metrics in the Osa Conservation Area (ACOSA). The Google Earth Engine API was used to process Surface Reflectance Tier 1 Landsat 5 Thematic Mapper and Landsat 8 Operational Land Imager data for 1987, 1998, and 2019, years with relatively cloud-free satellite imagery. Land use/land cover (LULC) maps were generated with the pixel-based random forest machine learning algorithm, and Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), and functional landscape metrics were calculated. The LULC maps are the first to track land use change, from 1987 to 2019 and the first to separately classify mature and secondary forest in the region, and have already proven useful for conservation efforts. The results suggest that forest cover, NDVI, EVI, and structural connectivity increased from 1987 to 2019 across the study area, both within and surrounding the PAs, suggesting minimal deforestation encroachment and local leakage. These changes may have contributed to the increasing vertebrate abundance observed in the region. PAs, especially national parks with stricter conservation regulations, displayed the highest forest cover and connectivity. Forest cover increased in properties receiving PES payments. Following the Forest Law's 1996 deforestation ban, both forest conversion and reforestation rates decreased, suggesting the law curbed deforestation but did not drive reforestation across the region. Connectivity outside of PAs slightly declined following the adoption of the law, so the subsequent forest growth likely occurred in mostly previously unforested areas. Forest expansion alone does not ensure connectivity. We highlight the importance of developing policies, PES programs, and monitoring systems that emphasize conserving and restoring large, connected forest patches for biodiversity conservation and landscape resilience. Resumen: Aunque las áreas protegidas (APs) son una importante estrategia de conservación para proteger ecosistemas y especies vulnerables, algunos análisis recientes cuestionan su eficacia para frenar la deforestación y mantener la conectividad del paisaje. La distribución espacial de los bosques dentro y fuera de las AP puede afectar el funcionamiento de los ecosistemas y los movimientos de la fauna. La Península de Osa–y Costa Rica en general–constituyen casos de estudio únicos de conservación debido a su elevada biodiversidad, su extensa red de AP, sus políticas medioambientales y sus programas de Pago por Servicios Ambientales (PSA). Este estudio explora la relación entre APs, la Ley Forestal de 1996, PSA, cobertura y métricas del paisaje en el Área de Conservación Osa (ACOSA). Se utilizó la plataforma Google Earth Engine API para procesar datos de Reflectancia Superficial Tier 1 Landsat 5 Thematic Mapper y Landsat 8 Operational Land Imager para 1987, 1998 y 2019, años con imágenes satelitales relativamente libres de nubes. Se generaron mapas de uso del suelo con el algoritmo de aprendizaje automático basado en pixeles Random Forest, y se calcularon el índice de vegetación de diferencia normalizada (NDVI), el índice de vegetación mejorado (EVI) y las métricas de paisaje funcionales. Estos mapas, los primeros en clasificar por separado los bosques maduros y secundarios de la región, han demostrado su utilidad para los esfuerzos de conservación. Los resultados sugieren que la cobertura forestal, el NDVI, el EVI y la conectividad estructural aumentaron entre 1987 y 2019 en toda la región de estudio, tanto dentro de las AP como en sus alrededores, lo que sugiere una expansión mínima de la deforestación dentro y fuera de las AP. Estos cambios pueden haber contribuido al aumento de la abundancia de vertebrados observado en la región. Las AP, especialmente los parques nacionales con regulaciones de conservación más estrictas, mostraron la mayor cobertura forestal y conectividad. La cobertura forestal aumentó en aquellas propiedades que recibieron PSA. Tras la prohibición de la deforestación por la Ley Forestal de 1996, disminuyeron tanto las tasas de conversión forestal como las de reforestación, lo que sugiere que la ley frenó la deforestación, pero no impulsó la reforestación. La conectividad fuera de las AP disminuyó ligeramente tras la entrada en vigor de la ley, lo que sugiere que el crecimiento forestal posterior se produjo en zonas que antes no estaban forestadas. Por lo tanto, la expansión forestal por sí sola no garantiza la conectividad. Resaltamos la importancia de desarrollar políticas, programas PSA y sistemas de monitoreo que hagan hincapié en la conservación y restauración de grandes zonas forestales conectadas para apuntalar la conservación de la biodiversidad y la resiliencia del paisaje. [ABSTRACT FROM AUTHOR]

Monitoring salt marshes with remote sensing is necessary to evaluate their state and restoration. Determining appropriate techniques for this can be overwhelming. Our study provides insight into whether a pixel- or object-based Random Forest classification approach is best for mapping vegetation in north temperate salt marshes. We used input variables from drone images (raw reflectances, vegetation indices, and textural features) acquired in June, July, and August 2021 of a salt marsh restoration and reference site in Aulac, New Brunswick, Canada. We also investigated the importance of input variables and whether using landcover classes representing areas of change was a practical way to evaluate variation in the monthly images. Our results indicated that (1) the classifiers achieved overall validation accuracies of 91.1–95.2%; (2) pixel-based classifiers outperformed object-based classifiers by 1.3–2.0%; (3) input variables extracted from the August images were more important than those extracted from the June and July images; (4) certain raw reflectances, vegetation indices, and textural features were among the most important variables; and (5) classes that changed temporally were mapped with user's and producer's validation accuracies of 86.7–100.0%. Knowledge gained during this study will inform assessments of salt marsh restoration trajectories spanning multiple years. [ABSTRACT FROM AUTHOR]

Drought is a frequent global phenomenon. Solar-induced chlorophyll fluorescence (SIF), an electromagnetic signal, has been proven to be an efficient tool for monitoring and assessing gross primary productivity (GPP) and drought. To address the issue of the sparse resolution of satellite-based SIF, researchers have developed different downscaling algorithms. Recently, the most frequently used SIF products had a spatial resolution of 0.05 degrees. However, these spatial resolution SIF data are not conducive to regional agricultural drought monitoring. In this study, we utilized the global 'OCO-2' solar-induced fluorescence (GOSIF) products along with normalized difference vegetation index (NDVI) and land surface temperature (LST) products. With the powerful advantages offered by Google Earth Engine (GEE), we could conveniently acquire the necessary data. Additionally, employing the random forest (RF) method, we successfully acquired downscaled SIF data at an enhanced spatial resolution of 1 km. Using those downscaled SIF results with 1 km resolution, an SIF anomaly index was established and calculated to monitor drought. Results showed that the RF-based downscaled SIF result followed the same trend as the GOSIF value. Subsequently, correlation coefficients between SIF and GPP were calculated. The downscaled SIF demonstrated a higher correlation with GPP from MODIS compared to 0.05-degree GOSIF, with coefficients of 0.74 and 0.68 in May 2018, respectively. Moreover, the SIF anomaly index showed positive correlations with crop yield; the correlation coefficients were 0.93 for wheat and 0.89 for maize. The drought index had a negative correlation with areas affected by drought, with a correlation coefficient of −0.58. Finally, the SIF anomaly index was used to monitor drought from 2001 to 2020 in Henan Province. The 1 km SIF results obtained through the RF-based downscaled method were deemed reliable, thereby establishing the suitability of the SIF anomaly index for drought monitoring at a regional scale. [ABSTRACT FROM AUTHOR]

Mapping soil organic carbon (SOC) stock can serve as a resilience indicator for climate change. As part of the carbon dioxide (CO2) sink, soil has recently become an integral part of the global carbon agenda to mitigate climate change. We used hyperspectral remote sensing to model the SOC stock in the Sele River plain located in the Campania region in southern Italy. To this end, a soil spectral library (SSL) for the Campania region was combined with an aerial hyperspectral image acquired with the AVIRIS–NG sensor mounted on a Twin Otter aircraft at an altitude of 1433 m. The products of this study were four raster layers with a high spatial resolution (1 m), representing the SOC stocks and three other related soil attributes: SOC content, clay content, and bulk density (BD). We found that the clay minerals' spectral absorption at 2200 nm has a significant impact on predicting the examined soil attributes. The predictions were performed by using AVIRIS–NG sensor data over a selected plot and generating a quantitative map which was validated with in situ observations showing high accuracies in the ground-truth stage (OC stocks [RPIQ = 2.19, R2 = 0.72, RMSE = 0.07]; OC content [RPIQ = 2.27, R2 = 0.80, RMSE = 1.78]; clay content [RPIQ = 1.6 R2 = 0.89, RMSE = 25.42]; bulk density [RPIQ = 1.97, R2 = 0.84, RMSE = 0.08]). The results demonstrated the potential of combining SSLs with remote sensing data of high spectral/spatial resolution to estimate soil attributes, including SOC stocks. [ABSTRACT FROM AUTHOR]

Flood susceptibility mapping plays a crucial role in flood risk assessment and management. Accurate identification of areas prone to flooding is essential for implementing effective mitigation measures and informing decision-making processes. In this regard, the present study used high-resolution remote sensing products, i.e., synthetic aperture radar (SAR) images for flood inventory preparation and integrated four machine learning models (Random Forest: RF, Classification and Regression Trees: CART, Support Vector Machine: SVM, and Extreme Gradient Boosting: XGBoost) to predict flood susceptibility in Metlili watershed, Morocco. Initially, 12 independent variables (elevation, slope angle, aspect, plan curvature, topographic wetness index, stream power index, distance from streams, distance from roads, lithology, rainfall, land use/land cover, and normalized vegetation index) were used as conditioning factors. The flood inventory dataset was divided into 70% and 30% for training and validation purposes using a popular library, scikit-learn (i.e., train_test_split) in Python programming language. Additionally, the area under the curve (AUC) was used to evaluate the performance of the models. The accuracy assessment results showed that RF, CART, SVM, and XGBoost models predicted flood susceptibility with AUC values of 0.807, 0.780, 0.756, and 0.727, respectively. However, the RF model performed better at flood susceptibility prediction compared to the other models applied. As per this model, 22.49%, 16.02%, 12.67%, 18.10%, and 31.70% areas of the watershed are estimated as being very low, low, moderate, high, and very highly susceptible to flooding, respectively. Therefore, this study showed that the integration of machine learning models with radar data could have promising results in predicting flood susceptibility in the study area and other similar environments. [ABSTRACT FROM AUTHOR]

Cover crops play a pivotal role in mitigating climate change by bolstering carbon sequestration through biomass production and soil integration. However, current methods for quantifying cover crop biomass lack spatial precision and objectivity. Thus, our research aimed to devise a remote-sensing-based approach to estimate cover crop biomass across various species and mixtures during fallow periods in France. Leveraging Sentinel-2 optical data and machine learning algorithms, we modeled biomass across 50 fields representative of France's diverse cropping practices and climate types. Initial tests using traditional empirical relationships between vegetation indices/spectral bands and dry biomass revealed challenges in accurately estimating biomass for mixed cover crop categories due to spectral interference from grasses and weeds, underscoring the complexity of modeling diverse agricultural conditions. To address this challenge, we compared several machine learning algorithms (Support Vector Machine, Random Forest, and eXtreme Gradient Boosting) using spectral bands and vegetation indices from the latest available image before sampling as input. Additionally, we developed an approach that incorporates dense optical time series of Sentinel-2 data, generated using a Radial Basis Function for interpolation. Our findings demonstrated that a Random Forest model trained with dense time series data during the cover crop development period yielded promising results, with an average R-squared (r2) value of 0.75 and root mean square error (RMSE) of 0.73 t·ha−1, surpassing results obtained from methods using single-image snapshots (r2 of 0.55). Moreover, our approach exhibited robustness in accounting for factors such as crop species diversity, varied climatic conditions, and the presence of weed vegetation—essential for approximating real-world conditions. Importantly, its applicability extends beyond France, holding potential for global scalability. The availability of data for model calibration across diverse regions and timeframes could facilitate broader application. [ABSTRACT FROM AUTHOR]

Controlling groundwater table decline could mitigate land subsidence and induced environmental hazards in over-explored areas. Nevertheless, this becomes a challenge in the multi-layered porous system as (in)elastic deformation simultaneously occurs due to vast spatiotemporal variability in the groundwater table. In this study, SBAS-InSAR was used to estimate annual land deformation during 2017–2022 in a specific region of North China Plain, in which aquifers are composed of many layers of fine-grained compressible sediments and the groundwater table has experienced a prolonged decline. The random forest (RF) was applied to establish the nonlinear relationship between accumulated deformation and its potential driving factors, including the depth to the groundwater table (GWD) and its change rate, and the compressible sediment thickness. Results show that the marked subsidence and uplift co-exist in the region even though the groundwater table has risen widely since the South–North Water Diversion Project. The land subsidence is attributed to inelastic compaction of the thick compressible deposits in depression cone centers, where the GWD is over 40 m and 90 m in the shallow and deep aquifers, respectively. In contrast, the marked uplift is primarily attributed to fast rising of the groundwater table (e.g., −2.44 m/a). The RF predictions suggest that, to control the subsidence, the GWD should be less than 20 and 70 m in the shallow and deep aquifers, respectively, and the rising rate of the GWD should increase to 2–5 times of current rates in the depression cones. To mitigate the marked uplift, the rising rate of the GWD should reduce to 1/2–1/5 of the current rates in the shallow aquifers. The uneven deformations of sediments in the depression cone centers and uplift in their boundaries may exacerbate geohazards. Therefore, it is vital to implement appropriate governance of groundwater recovery in the multi-layered porous system. [ABSTRACT FROM AUTHOR]

Invasive plants are a serious problem in island ecosystems and are the main cause of the extinction of endemic species. Cuba is located within one of the hotspots of global biodiversity, which, coupled with high endemism and the impacts caused by various disturbances, makes it a region particularly sensitive to potential damage by invasive plants like Dichrostachys cinerea (L.) Wight & Arn. (marabú). However, there is a lack of timely information for monitoring this species, as well as about the land use and land cover (LULC) classes most significantly impacted by this invasion in the last few decades and their spatial distribution. The main objective of this study, carried out in Central Cuba, was to detect and monitor the spread of marabú over a 28-year period. The land covers for the years 1994 and 2022 were classified using Landsat 5 TM and 8 OLI images with three different classification algorithms: maximum likelihood (ML), support vector machine (SVM), and random forest (RF). The results obtained showed that RF outperformed the other classifiers, achieving AUC values of 0.92 for 1994 and 0.97 for 2022. It was confirmed that the area covered by marabú increased by 29,555 ha, from 61,977.59 ha in 1994 to 91,533.47 ha in 2022 (by around 48%), affecting key land covers like woodlands, mangroves, and rainfed croplands. These changes in the area covered by marabú were associated, principally, with changes in land uses and tenure and not with other factors, such as rainfall or relief in the province. The use of other free multispectral imagery, such as Sentinel 2 data, with higher temporal and spatial resolution, could further refine the model's accuracy. [ABSTRACT FROM AUTHOR]

Using the SWAN (Severe Weather Automatic Nowcasting) maximum reflectivity mosaic product and the lightning positioning observations (LPOs) from the ADTD (Advanced Direction and Time of Arrival Detection) system obtained during the 2018–2020 warm season (May to September), adding multi-characteristic LPO parameters in addition to lightning density, the retrieval relationship between lightning and maximum proxy reflectivity, deemed FRST, is constructed by using random forest. The FRST is compared with two empirical relationships from the GSI (Gridpoint Statistical Interpolation) assimilation system, and the results show that the FRST retrieved result better reflects the frequency distribution structure and peak interval of maximum reflectivity. The correlation coefficient between the FRST retrieved result and the observed maximum reflectivity is 0.7037, which is 3.38 (3.12) times greater than that of empirical GSI relationships. The root mean square error and the mean absolute error are 50.85% (28.05%) and 57.15% (35.19%) lower than those for the empirical GSI relationships, respectively. The equitable threat score (ETS) and bias score (BIAS) for FRST are better than those of the empirical GSI relationships in all three maximum reflectivity intervals. [ABSTRACT FROM AUTHOR]

Improved agricultural production systems, together with increased grain yield, are essential to feed the growing global population in the 21st century. Global gridded crop models (GGCMs) have been extensively used to assess crop production and yield simulation on a large geographical scale. However, GGCMs are less effective when they are used on a finer scale, significantly limiting the precision in capturing the yearly maize yield. To address this issue, we propose a relatively more advanced approach that downsizes GGCMs by combining machine learning and crop modeling to enhance the accuracy of maize yield simulations on a regional scale. In this study, we combined the random forest algorithm with multiple data sources, trained the algorithm on low-resolution maize yield simulations from GGCMs, and applied it to a finer spatial resolution on a regional scale in China. We evaluated the performance of the eight GGCMs by utilizing a total of 1046 county-level maize yield data available over a 30-year period (1980–2010). Our findings reveal that the downscaled models created for maize yield simulations exhibited a remarkable level of accuracy (R2 ≥ 0.9, MAE < 0.5 t/ha, RMSE < 0.75 t/ha). The original GGCMs performed poorly in simulating county-level maize yields in China, and the improved GGCMs in our study captured an additional 17% variability in the county-level maize yields in China. Additionally, by optimizing nitrogen management strategies, we identified an average maize yield gap at the county level in China ranging from 0.47 to 1.82 t/ha, with the south maize region exhibiting the highest yield gap. Our study demonstrates the high effectiveness of machine learning methods for the spatial downscaling of crop models, significantly improving GGCMs' performance in county-level maize yield simulations. [ABSTRACT FROM AUTHOR]

Generating orchards spatial distribution maps within a heterogeneous landscape is challenging and requires fine spatial and temporal resolution images. This study examines the effectiveness of Sentinel-1 (S1) and Sentinel-2 (S2) satellite data of relatively high spatial and temporal resolutions for discriminating major orchards in the Khairpur district of the Sindh province, Pakistan using machine learning methods such as random forest (RF) and a support vector machine. A Multicollinearity test (MCT) was performed among the multi-temporal S1 and S2 variables to remove those with high correlations. Six different feature combination schemes were tested, with the fusion of multi-temporal S1 and S2 (scheme-6) outperforming all other combination schemes. The spectral separability between orchards pairs was assessed using Jeffries-Matusita (JM) distance, revealing that orchard pairs were completely separable in the multi-temporal fusion of both sensors, especially the indistinguishable pair of dates-mango. The performance difference between RF and SVM was not significant, SVM showed a slightly higher accuracy, except for scheme-4 where RF performed better. This study concludes that multi-temporal fusion of S1 and S2 data, coupled with robust ML methods, offers a reliable approach for orchard classification. Prospectively, these findings will be helpful for orchard monitoring, improvement of yield estimation and precision based agricultural practices. [ABSTRACT FROM AUTHOR]

The integration of optical and SAR datasets through ensemble machine learning models shows promising results in urban remote sensing applications. The integration of multi-sensor datasets enhances the accuracy of information extraction. This research presents a comparison of two ensemble machine learning classifiers (random forest and extreme gradient boost (XGBoost)) classifiers using an integration of optical and SAR features and simple layer stacking (SLS) techniques. Therefore, Sentinel-1 (SAR) and Landsat 8 (optical) datasets were used with SAR textures and enhanced modified indices to extract features for the year 2023. The classification process utilized two machine learning algorithms, random forest and XGBoost, for urban impervious surface extraction. The study focused on three significant East Asian cities with diverse urban dynamics: Jakarta, Manila, and Seoul. This research proposed a novel index called the Normalized Blue Water Index (NBWI), which distinguishes water from other features and was utilized as an optical feature. Results showed an overall accuracy of 81% for UIS classification using XGBoost and 77% with RF while classifying land use land cover into four major classes (water, vegetation, bare soil, and urban impervious). However, the proposed framework with the XGBoost classifier outperformed the RF algorithm and Dynamic World (DW) data product and comparatively showed higher classification accuracy. Still, all three results show poor separability with bare soil class compared to ground truth data. XGBoost outperformed random forest and Dynamic World in classification accuracy, highlighting its potential use in urban remote sensing applications. [ABSTRACT FROM AUTHOR]

Since the early 1950s, the development of human settlements and over-exploitation of agriculture in the China side of the Amur River Basin (CARB) have had a major impact on the water environment of the surrounding lakes, resulting in a decrease of aquatic vegetation. According to the United Nations Sustainable Development Goals, a comprehensive understanding of the extent and variability of aquatic vegetation is crucial for preserving the structure and functionality of stable aquatic ecosystems. Currently, there is a deficiency in the CARB long-sequence dataset of aquatic vegetation distribution in China. This shortage hampers effective support for actual management. Therefore, the development of a fast, robust, and automatic method for accurate extraction of aquatic vegetation becomes crucial for large-scale applications. Our objective is to gather information on the spatial and temporal distribution as well as changes in aquatic vegetation within the CARB. Utilizing a hybrid approach that combines the maximum spectral index composite and Otsu algorithm, along with the integration of convolutional neural networks (CNN) and random forest, we applied this methodology to obtain an annual dataset of aquatic vegetation spanning from 1985 to 2020 using Landsat series imagery. The accuracy of this method was validated through both field investigations and Google Images. Upon assessing the confusion matrix spanning from 1985 to 2020, the producer accuracy for aquatic vegetation classification consistently exceeded 87%. Further quantitative analysis unveiled a discernible decreasing trend in both the water and vegetation areas of lakes larger than 20 km2 within the CARB over the past 36 years. Specifically, the total water area decreased from 3575 km2 to 3412 km2, while the vegetation area decreased from 745 km2 to 687 km2. These changes may be attributed to a combination of climate change and human activities. These quantitative data hold significant practical implications for establishing a scientific restoration path for lake aquatic vegetation. They are particularly valuable for constructing the historical background and reference indices of aquatic vegetation. [ABSTRACT FROM AUTHOR]

A proliferation of invasive species is displacing native species, occupying their habitats and degrading biodiversity. One of these is the invasive goldenrod (Solidago spp.), characterized by aggressive growth that results in habitat disruption as it outcompetes native plants. This invasiveness also leads to altered soil composition through the release of allelopathic chemicals, complicating control efforts and making it challenging to maintain ecological balance in affected areas. The research goal was to develop methods that allow the analysis of changes in heterogeneous habitats with high accuracy and repeatability. For this reason, we used open source classifiers Support Vector Machine (SVM), Random Forest (RF), and satellite images of Sentinel-2 (free) and PlanetScope (commercial) to assess their potential in goldenrod classification. Due to the fact that invasions begin with invasion footholds, created by small patches of invasive, autochthonous plants and different land cover patterns (asphalt, concrete, buildings) forming heterogeneous areas, we based our studies on field-verified polygons, which allowed the selection of randomized pixels for the training and validation of iterative classifications. The results confirmed that the optimal solution is the use of multitemporal Sentinel-2 images and the RF classifier, as this combination gave F1-score accuracy of 0.92–0.95 for polygons dominated by goldenrod and 0.85–0.89 for heterogeneous areas where goldenrod was in the minority (mix class; smaller share of goldenrod in canopy than autochthonous plants). The mean decrease in the accuracy analysis (MDA), indicating an informativeness of individual spectral bands, showed that Sentinel-2 bands coastal aerosol, NIR, green, SWIR, and red were comparably important, while in the case of PlanetScope data, the NIR and red were definitely the most important, and remaining bands were less informative, and yellow (B5) did not contribute significant information even during the flowering period, when the plant was covered with intensely yellow perianth, and red-edge, coastal aerosol, or green II were much more important. The maximum RF classification values of Sentinel-2 and PlanetScope images for goldenrod are similar (F1-score > 0.9), but the medians are lower for PlanetScope data, especially with the SVM algorithm. [ABSTRACT FROM AUTHOR]

The application of lidar data to assist with forest inventory is common around the world. However, the determination of tree species is still somewhat elusive. Lidar data collected using UAS (uncrewed aircraft systems) platforms offer high density point cloud data for areas from a few to several hundred hectares. General point cloud metrics computed using these data captured differences in the crown structure that proved useful for species classification. For our study, we manually adjusted plot and tree locations to align field trees and UAS lidar point data and computed common descriptive metrics using a small cylindrical sample of points designed to capture the top three meters and leader of each tree. These metrics were used to train a random forest classifier to differentiate between two conifer species, Douglas fir and western hemlock, common in the Pacific Northwest region of the United States. Our UAS lidar data had a single swath pulse density of 90 pulses/m2 and an aggregate pulse density of 556 pulses/m2. We trained classification models using both height and intensity metrics, height metrics alone, intensity metrics alone, and a small subset of five metrics, and achieved overall accuracies of 91.8%, 88.7%, 78.6%, and 91.5%, respectively. Overall, we showed that UAS lidar data captured morphological differences between the upper crowns of our two target species and produced a classification model that could be applied over large areas. [ABSTRACT FROM AUTHOR]

Surface urban heat islands (SUHIs) can extend beyond the urban boundaries and greatly affect the thermal environment of continuous regions over an agglomeration. Traditional urban-rural dichotomy depending on the built-up and non-urban lands is challenged in characterizing regional SUHIs, such as how to accurately quantify the intensity, spatial pattern, and scales of SUHIs, which are vulnerable to SUHIs, and what the optimal scale for conducting measures to mitigate the SUHIs. We propose a machine-learning-assisted solution to address these problems based on the thermal similarity in the Yangtze River Delta of China. We first identified the regional-level SUHI zone of approximately 42,328 km2 and 38,884 km2 and the areas that have no SUHI effects from the annual cycle of land surface temperatures (LSTs) retrieved from Terra and Aqua satellites. Defining SUHI as an anomaly on background condition, random forest (RF) models were further adopted to fit the LSTs in the areas without the SUHI effects and estimate the LST background and SUHI intensity at each grid point in the SUHI zone. The RF models performed well in fitting rural LSTs with a simulation error of approximately 0.31 °C/0.44 °C for Terra/Aqua satellite data and showed a good generalization ability in estimating the urban LST background. The RF-estimated daytime Aqua/SUHI intensity peaked at approximately 6.20 °C in August, and the Terra/SUHI intensity had two peaks of approximately 3.18 and 3.81 °C in May and August, with summertime RF-estimated SUHIs being more reliable than other SUHI types owing to the smaller simulation error of less than 1.0 °C in July–September. This machine-learning-assisted solution identified an optimal SUHI scale of 30,636 km2 and a zone of approximately 23,631 km2 that is vulnerable to SUHIs, and it provided the SUHI intensity and statistical reliability for each grid point identified as being part of the SUHI. Urban planners and decision-makers can focus on the statistically reliable RF-estimated summertime intensities in SUHI zones that have an LST annual cycle similar to that of large cities in developing effective strategies for mitigating adverse SUHI effects. In addition, the selection of large cities might strongly affect the accuracy of identifying the SUHI zone, which is defined as the areas that have an LST annual cycle similar to large cities. Water bodies might reduce the RF performance in estimating the LST background over urban agglomerations. [ABSTRACT FROM AUTHOR]

Cocoa is the economic backbone of Côte d'Ivoire and Ghana, making them the leading cocoa-producing countries in the world. However, cocoa farming has been a major driver of deforestation and landscape degradation in West Africa. Various stakeholders are striving for a zero-deforestation cocoa sector by implementing sustainable farming strategies and a more transparent supply chain. In the context of tracking cocoa sources and contributing to cocoa-driven deforestation monitoring, the demand for accurate and up-to-date maps of cocoa plantations is increasing. Yet, access to limited reference data and imperfect data quality can impose challenges in producing reliable maps. This study classified full-sun-cocoa-growing areas using limited reference data relative to the large and heterogeneous study areas in Côte d'Ivoire and Ghana. A Sentinel-2 composite image of 2021 was generated to train a random forest model. We undertook reference data refinement, selection of the most important handcrafted features and data sampling to ensure spatial independence. After refining the quality of the reference data and despite their size reduction, the random forest performance was improved, achieving an overall accuracy of 85.1 ± 2.0% and an F1 score of 84.6 ± 2.4% (mean ± one standard deviation from ten bootstrapping iterations). Emphasis was given to the qualitative visual assessment of the map using very high-resolution images, which revealed cases of strong and weak generalisation capacity of the random forest. Further insight was gained from the comparative analysis of our map with two previous cocoa classification studies. Implications of the use of cocoa maps for reporting were discussed. [ABSTRACT FROM AUTHOR]

The goal of this study was to estimate the areas under willow cultivation by farmers, as well as their growth and health status. Due to the extremely small patch size of land cover types in the study area, Sentinel-2 data were used to conduct supervised classification based on the random forest machine learning technique, and a large training dataset was produced from PlanetScope satellite imagery. The results of image classification using Google Earth Engine indicated that the Sentinel data were suitable for identifying willow-cultivated areas. It was found that these areas declined from 875.32 ha in 2017 to 288.41 ha in 2022. The analysis of the growth and health conditions of willow-cultivated plots also revealed that the temporal variations in the NDVI in these plots decreased significantly in 2022 as compared to previous years (p < 0.05). An in-depth analysis revealed a significant positive correlation between NDVI, precipitation, and temperature. It was found that the most efficient components explaining the process of browning the vegetation in the planted willow plots were the increasing temperature and decreasing precipitation. This research may be used to document the national and global monitoring efforts for climate change adaptation. [ABSTRACT FROM AUTHOR]

Sea ice motion (SIM) plays a crucial role in setting the distribution of the ice cover in the Arctic. Limited by images' spatial resolution and tracking algorithms, challenges exist in obtaining coastal sea ice motion (SIM) based on passive microwave satellite sensors. In this study, we developed a method based on random forest (RF) models to obtain Arctic SIM in winter by incorporating wind field and coastal geographic location information. These random forest models were trained using Synthetic Aperture Radar (SAR) SIM data. Our results show good consistency with SIM data retrieved from satellite imagery and buoy observations. With respect to the SAR data, compared with SIM estimated with RF model training using reanalysis surface wind, the results by additional coastal information input had a lower root mean square error (RMSE) and a higher correlation coefficient by 31% and 14% relative improvement, respectively. The latter SIM result also showed a better performance for magnitude, especially within 100 km of the coastline in the north of the Canadian Arctic Archipelago. In addition, the influence of coastline on SIM is quantified through variable importance calculation, at 22% and 28% importance of all RF variables for east and north SIM components, respectively. These results indicate the great potential of RF models for estimating SIM over the whole Arctic Ocean in winter. [ABSTRACT FROM AUTHOR]

The mapping of invasive plant species is essential for effective ecosystem control and planning, especially in protected areas. One of the widespread invasive plants that threatens the species richness of Natura 2000 habitats in Europe is the large-leaved lupine (Lupinus polyphyllus). In our study, this species was identified at two Natura 2000 sites in southern Poland using airborne HySpex hyperspectral images, and support vector machine (SVM) and random forest (RF) classifiers. Aerial and field campaigns were conducted three times during the 2016 growing season (May, August, and September). An iterative accuracy assessment was performed, and the influence of the number of minimum noise fraction (MNF) bands on the obtained accuracy of lupine identification was analyzed. The highest accuracies were obtained for the August campaign using 30 MNF bands as input data (median F1 score for lupine was 0.82–0.85), with lower accuracies for the May (F1 score: 0.77–0.81) and September (F1 score: 0.78–0.80) campaigns. The use of more than 30 MNF bands did not significantly increase the classification accuracy. The SVM and RF algorithms allowed us to obtain comparable results in both research areas (OA: 89–94%). The method of the multiple classification and thresholding of frequency images allowed the results of many predictions to be included in the final map. [ABSTRACT FROM AUTHOR]

Mangrove forests are a biodiverse ecosystem known for a wide variety of crucial ecological services, including carbon sequestration, coastal erosion control, and prevention of saltwater intrusion. Given the ecological importance of mangrove forests, a comprehensive and up-to-date mangrove extent mapping at broad geographic scales is needed to define mangrove forest changes, assess their implications, and support restoration activities and decision making. The main objective of this study is to evaluate mangrove classifications derived from a combination of Landsat-8 OLI, Sentinel-2, and Sentinel-1 observations using a random forest (RF) machine learning (ML) algorithm to identify the best approach for monitoring Guyana's mangrove forests on an annual basis. Algorithm accuracy was tested using high-resolution planet imagery in Collect Earth Online. Results varied widely across the different combinations of input data (overall accuracy, 88–95%; producer's accuracy for mangroves, 50–87%; user's accuracy for mangroves, 13–69%). The combined optical–radar classification demonstrated the best performance with an overall accuracy of 95%. Area estimates of mangrove extent ranged from 908.4 to 3645.0 hectares. A ground-based validation exercise confirmed the extent of several large, previously undocumented areas of mangrove forest loss. The results establish that a data fusion approach combining optical and radar data performs marginally better than optical-only approaches to mangrove classification. This ML approach, which leverages free and open data and a cloud-based analytics platform, can be applied to mapping other areas of mangrove forests in Guyana. This approach can also support the operational monitoring of mangrove restoration areas managed by Guyana's National Agricultural and Research Extension Institute (NAREI). [ABSTRACT FROM AUTHOR]

Long-term exposure to high concentrations of fine particles can cause irreversible damage to people's health. Therefore, it is of extreme significance to conduct large-scale continuous spatial fine particulate matter (PM2.5) concentration prediction for air pollution prevention and control in China. The distribution of PM2.5 ground monitoring stations in China is uneven with a larger number of stations in southeastern China, while the number of ground monitoring sites is also insufficient for air quality control. Remote sensing technology can obtain information quickly and macroscopically. Therefore, it is possible to predict PM2.5 concentration based on multi-source remote sensing data. Our study took China as the research area, using the Pearson correlation coefficient and GeoDetector to select auxiliary variables. In addition, a long short-term memory neural network and random forest regression model were established for PM2.5 concentration estimation. We finally selected the random forest regression model (R2 = 0.93, RMSE = 4.59 μg m−3) as our prediction model by the model evaluation index. The PM2.5 concentration distribution across China in 2021 was estimated, and then the influence factors of high-value regions were explored. It is clear that PM2.5 concentration is not only related to the local geographical and meteorological conditions, but also closely related to economic and social development. [ABSTRACT FROM AUTHOR]

The variation of lake water storage is an important indicator for studying both climate change and ecological environment changes. Previous studies have mainly focused on the lake storage changes in recent decades, and predicting future lake storage changes on the Tibetan Plateau under climate change scenarios remains a crucial gap. We addressed this gap by establishing prediction models for water storage changes in nine lakes using historical water storage and climate data from the past 29 years and predicting the water storage changes for the next 80 years under three scenarios based on Phase 6 of the Coupled Model Intercomparison Project (CMIP6) data. The Quantile-mapping (QM) method was applied to correct the precipitation data of CMIP6 with assimilated data. The results indicated that the prediction model performed well, with high correlation (R2 > 0.7 for the training set) and low mean absolute error (MSE < 0.1 km3). The results suggest that most lakes will experience a slight increase in water storage until 2050, followed by a rapid rise until 2100 under all three SSP (Shared Socioeconomic Pathways) scenarios, including SSP126, SSP245, and SSP585. By the end of the century, the total projected increase in lake water storage is estimated to be 189.676 ± 16.266 km3, 191.762 ± 10.683 km3, and 186.212 ± 6.441 km3 until 2100, respectively. [ABSTRACT FROM AUTHOR]

Accurately predicting the condition rating of a bridge deck is crucial for effective maintenance and repair planning. Despite significant research efforts to develop deterioration models, the efficacy of Random Forest, eXtreme Gradient Boosting (XGBoost), and Artificial Neural Network (ANN) in predicting the condition rating of the nation's bridge decks has remained unexplored. This study aims to assess the effectiveness of these algorithms for deck condition rating prediction at the national level. To achieve this, the study collected bridge data, which includes National Bridge Inventory (NBI), traffic, and climate regions gathered using Geospatial Information Science (GIS) and remote sensing techniques. Two datasets were collected: bridge data for a single year of 2020 and historical bridge data covering a five-year period from 2016 to 2020. Three models were trained using 319,404 and 1,246,261 bridge decks in the single-year bridge data and the five-year historical bridge data, respectively. Results show that the use of historical bridge data significantly improves the performance of the models compared to the single-year bridge data. Specifically, the Random Forest model achieved an overall accuracy of 83.4% and an average F1 score of 79.7%. In contrast, the XGBoost model achieved an overall accuracy of 79.4% and an average F1 score of 77.5%, while the ANN model obtained an overall accuracy of 79.7% and an average F1 score of 78.4%. Permutation-based variable importance reveals that NBI, traffic, and climate regions significantly contribute to model development. In conclusion, the Random Forest, XGBoost, and ANN models, trained using updated historical bridge data, provide useful tools for accurately predicting the condition rating of bridge decks in the United States, allowing infrastructure managers to efficiently schedule inspections and allocate maintenance resources. [ABSTRACT FROM AUTHOR]

Flooding is a natural disaster that coexists with human beings and causes severe loss of life and property worldwide. Although numerous studies for flood susceptibility modelling have been introduced, a notable gap has been the overlooked or reduced consideration of the uncertainty in the accuracy of the produced maps. Challenges such as limited data, uncertainty due to confidence bounds, and the overfitting problem are critical areas for improving accurate models. We focus on the uncertainty in susceptibility mapping, mainly when there is a significant variation in the predictive relevance of the predictor factors. It is also noted that the receiver operating characteristic (ROC) curve may not accurately depict the sensitivity of the resulting susceptibility map to overfitting. Therefore, reducing the overfitting problem was targeted to increase accuracy and improve processing time in flood prediction. This study created a spatial repository to test the models, containing data from historical flooding and twelve topographic and geo-environmental flood conditioning variables. Then, we applied random forest (RF) and extreme gradient boosting (XGB) algorithms to map flood susceptibility, incorporating a variable drop-off in the empirical loop function. The results showed that the drop-off loop function was a crucial method to resolve the model uncertainty associated with the conditioning factors of the susceptibility modelling and methods. The results showed that approximately 8.42% to 9.89% of Marib City and 9.93% to 15.69% of Shibam City areas were highly vulnerable to floods. Furthermore, this study significantly contributes to worldwide endeavors focused on reducing the hazards linked to natural disasters. The approaches used in this study can offer valuable insights and strategies for reducing natural disaster risks, particularly in Yemen. [ABSTRACT FROM AUTHOR]

This study demonstrates a framework for using high-resolution satellite imagery to automatically map and monitor outbreaks of red needle cast (Phytophthora pluvialis) in planted pine forests. This methodology was tested on five WorldView satellite scenes collected over two sites in the Gisborne Region of New Zealand's North Island. All scenes were acquired in September: four scenes were acquired yearly (2018–2020 and 2022) for Wharerata, while one more was obtained in 2019 for Tauwhareparae. Training areas were selected for each scene using manual delineation combined with pixel-level thresholding rules based on band reflectance values and vegetation indices (selected empirically) to produce 'pure' training pixels for the different classes. A leave-one-scene-out, pixel-based random forest classification approach was then used to classify all images into (i) healthy pine forest, (ii) unhealthy pine forest or (iii) background. The overall accuracy of the models on the internal validation dataset ranged between 92.1% and 93.6%. Overall accuracies calculated for the left-out scenes ranged between 76.3% and 91.1% (mean overall accuracy of 83.8%), while user's and producer's accuracies across the three classes were 60.2–99.0% (71.4–91.8% for unhealthy pine forest) and 54.4–100% (71.9–97.2% for unhealthy pine forest), respectively. This work demonstrates the possibility of using a random forest classifier trained on a set of satellite scenes for the classification of healthy and unhealthy pine forest in new and completely independent scenes. This paves the way for a scalable and largely autonomous forest health monitoring system based on annual acquisitions of high-resolution satellite imagery at the time of peak disease expression, while greatly reducing the need for manual interpretation and delineation. [ABSTRACT FROM AUTHOR]

Climate-driven changes in fire regimes are expected across the pan-Arctic region. Trends in arctic fires are thought to be generally increasing; however, fire mapping across the region is far from comprehensive or systematic. We developed a new detection workflow and built a dataset of unrecorded tundra fires in Canada using Landsat data. We built a reference dataset of spectral indices from previously mapped fires in northern Canada to train a Random Forest model for detecting new fires between 1986 and 2022. In addition, we used time series information for each pixel to reduce false positives and narrow the large search space down to a finite set of regions that had experienced changes. We found 209 previously undetected fires in the Arctic and sub-Arctic regions, increasing the mapped burned area by approximately 30%. The median fire size was small, with roughly 3/4 of the fires being <100 ha in size. The majority of newly detected fires (69%) did not have satellite-derived hotspots associated with them. The dataset presented here is commission error-free and can be viewed as a reference dataset for future analyses. Moreover, future improvements and updates will leverage these data to improve the detection workflow outlined here, particularly for small and low-severity fires. These data can facilitate broader analyses that examine trends and environmental drivers of fire across the Arctic region. Such analyses could begin to untangle the mechanisms driving heterogeneous fire responses to climate observed across regions of the Circumpolar North. [ABSTRACT FROM AUTHOR]

Powdery mildew is one of the most significant rubber tree diseases, with a substantial impact on the yield of natural rubber. This study aims to establish a detection approach that coupled continuous wavelet transform (CWT) and machine learning for the accurate assessment of powdery mildew severity in rubber trees. In this study, hyperspectral reflectance data (350–2500 nm) of healthy and powdery mildew-infected leaves were measured with a spectroradiometer in a laboratory. Subsequently, three types of wavelet features (WFs) were extracted using CWT. They were as follows: WFs dimensionally reduced by the principal component analysis (PCA) of significant wavelet energy coefficients (PCA-WFs); WFs extracted from the top 1% of the determination coefficient between wavelet energy coefficients and the powdery mildew disease class (1%R2-WFs); and all WFs at a single decomposition scale (SS-WFs). To assess the detection capability of the WFs, the three types of WFs were input into the random forest (RF), support vector machine (SVM), and back propagation neural network (BPNN), respectively. As a control, 13 optimal traditional spectral features (SFs) were extracted and combined with the same classification methods. The results revealed that the WF-based models all performed well and outperformed those based on SFs. The models constructed based on PCA-WFs had a higher accuracy and more stable performance than other models. The model combined PCA-WFs with RF exhibited the optimal performance among all models, with an overall accuracy (OA) of 92.0% and a kappa coefficient of 0.90. This study demonstrates the feasibility of combining CWT with machine learning in rubber tree powdery mildew detection. [ABSTRACT FROM AUTHOR]

Conventional maps of soil parent material (SPM) types obtained by field survey and manual mapping or predictions from other map data have limited accuracy. Digital soil mapping of SPM types necessitates accurate acquisition of SPM distribution information, which is still a challenge in hilly areas. This study developed a high-accuracy method for SPM identification in hilly areas at the county scale. Based on geographic information system technology, seven feature variables were extracted from the geological map, geomorphic map, digital elevation model, and remote sensing image data of Shanggao County, Jiangxi Province, China. Different feature combination schemes were designed to develop SPM identification models based on random forest (RF), support vector machine (SVM), and maximum likelihood classification (MLC) algorithms. The best SPM identification results were obtained from the RF algorithm using the combination of geological type, geomorphic type, elevation, and slope. Confusion matrices were constructed based on a field survey of 586 validation samples, and the results were evaluated in terms of overall accuracy, precision, recall, F1 score, and Kappa coefficient. The overall accuracy and Kappa coefficient of the results from the optimal RF model were 83.11% and 0.79, respectively, which were 26.11% and 0.31 higher than those of the conventional map, respectively. Its precision and recall for various SPM types were greater than 75%. A comprehensive comparison of the accuracy, uncertainty, and plotting performance of the SPM recognition results reveals that the RF algorithm outperforms the SVM algorithm and the MLC algorithm. Geological type was the largest contributor to SPM identification, followed by geomorphic type, elevation, and slope. The importance of different feature variables varied for distinct SPM types. The accuracy of SPM identification was not improved by selecting more feature variables, such as land use type, normalised difference vegetation index, and topographic wetness index. This study demonstrates the feasibility of high-accuracy county-level SPM mapping in hilly areas based on the RF algorithm using geological type, geomorphic type, elevation, and slope as feature variables. As hilly areas have typical topographic features and SPM types, the proposed method of SPM mapping can be useful for application in other similar areas. There are a few limitations in this study with regard to data quality and resolution, feature variable selection, classification algorithm generalisation, and study area representativeness, which may affect the outcomes and need to be solved. [ABSTRACT FROM AUTHOR]

Based on the community radiative transfer model, ensemble tree-based random forest algorithm, and extreme gradient boosting tree algorithm, this study established a random forest retrieval model (RF) and an extreme gradient boosting tree retrieval model (XGBoost) for the microwave land surface emissivity by using ERA5 reanalysis data and the observed brightness temperature of 10.65 GHz vertical polarization from FY3C Microwave Radiation Imager-I. In addition, an optimized Bayesian regularized neural network retrieval model (M2_30NN) has also been established on the basis of the original neural network land surface emissivity retrieval model (M1_20NN). The results show that compared with the simulated brightness temperature of the original land surface emissivity, the simulated brightness temperature of the land surface emissivity from each retrieval model is not only significantly improved in the correlation coefficient with the observed brightness temperature (5.92% (M1_20NN), 4.23% (M2_30NN), 14.21% (RF), 13.07% (XGBoost)), but also in the evaluation indexes of root mean square error, mean absolute error and explained variance regression score in the training datasets. Furthermore, in terms of testing datasets and spatiotemporal independence test datasets, the retrieval results of RF and XGBoost models can capture the spatial distribution patterns that are consistent with observations well, and also show great numerical improvement compared with the original model. In general, the XGBoost retrieval model is the best, followed by the RF retrieval model. [ABSTRACT FROM AUTHOR]

Knowledge of the soil water content (SWC) is important for many aspects of agriculture and must be monitored to maximize crop yield, efficiently use limited supplies of irrigation water, and ensure optimal nutrient management with minimal environmental impact. Single-location sensors are often used to monitor SWC, but a limited number of point measurements is insufficient to measure SWC across most fields since SWC is typically very heterogeneous. To overcome this difficulty, several researchers have used data acquired from unmanned aerial vehicles (UAVs) to predict the SWC by using machine learning on a limited number of point measurements acquired across a field. While useful, these methods are limited by the relatively small number of SWC measurements that can be acquired with conventional measurement techniques. This study uses UAV-based data and thousands of SWC measurements acquired using geophysical methods at two different depths and before and after precipitation to predict the SWC using the random forest method across a vineyard in the central United States. Both multispectral data (five reflectance bands and eleven vegetation indices calculated from these bands) and thermal UAV-based data were acquired, and the importance of different reflectance data and vegetation indices in the prediction of SWC was analyzed. Results showed that when both thermal and multispectral data were used to estimate SWC, the thermal data contributed the most to prediction accuracy, although multispectral data were also important. Reflectance data contributed as much or more to prediction accuracy than most vegetation indices. SWC measurements that had a larger sample size and greater penetration depth (~30 cm sampling depth) were more accurately predicted than smaller and shallower SWC estimates (~18 cm sampling depth). The timing of SWC estimation was also important; higher accuracy predictions were achieved in wetter soils than in drier soils, and a light precipitation event also improved prediction accuracy. [ABSTRACT FROM AUTHOR]

It is a difficult undertaking to reliably estimate global terrestrial evapotranspiration (ET) using the Visible Infrared Imaging Radiometer Suite (VIIRS) at high spatial and temporal scales. We employ deep neural networks (DNN) to enhance the estimation of terrestrial ET on a global scale using satellite data. We accomplish this by merging five algorithms that are process-based and that make use of VIIRS data. These include the Shuttleworth–Wallace dual-source ET method (SW), the Priestley–Taylor-based ET algorithm (PT-JPL), the MOD16 ET product algorithm (MOD16), the modified satellite-based Priestley–Taylor ET algorithm (MS-PT), and the simple hybrid ET algorithm (SIM). We used 278 eddy covariance (EC) tower sites from 2012 to 2022 to validate the DNN approach, comparing it to Bayesian model averaging (BMA), gradient boosting regression tree (GBRT) and random forest (RF). The validation results demonstrate that the DNN significantly improves the accuracy of daily ET estimates when compared to three other merging methods, resulting in the highest average determination coefficients (R2, 0.71), RMSE (21.9 W/m2) and Kling–Gupta efficiency (KGE, 0.83). Utilizing the DNN, we generated a VIIRS ET product with a 500 m spatial resolution for the years 2012–2020. The DNN method serves as a foundational approach in the development of a sustained and comprehensive global terrestrial ET dataset. The basis for characterizing and analyzing global hydrological dynamics and carbon cycling is provided by this dataset. [ABSTRACT FROM AUTHOR]

Wildfires present a significant threat to ecosystems and human life, requiring effective prevention and response strategies. Equally important is the study of post-fire damages, specifically burnt areas, which can provide valuable insights. This research focuses on the detection and classification of burnt areas and their severity using RGB and multispectral aerial imagery captured by an unmanned aerial vehicle. Datasets containing features computed from multispectral and/or RGB imagery were generated and used to train and optimize support vector machine (SVM) and random forest (RF) models. Hyperparameter tuning was performed to identify the best parameters for a pixel-based classification. The findings demonstrate the superiority of multispectral data for burnt area and burn severity classification with both RF and SVM models. While the RF model achieved a 95.5% overall accuracy for the burnt area classification using RGB data, the RGB models encountered challenges in distinguishing between mildly and severely burnt classes in the burn severity classification. However, the RF model incorporating mixed data (RGB and multispectral) achieved the highest accuracy of 96.59%. The outcomes of this study contribute to the understanding and practical implementation of machine learning techniques for assessing and managing burnt areas. [ABSTRACT FROM AUTHOR]

Timely and accurate detection of diseases plays a significant role in attaining optimal growing conditions of olive crops. This study evaluated the use of two machine learning algorithms, Random Forest (RF) and XGBoost (XGB), in conjunction with the feature selection methods Recursive Feature Elimination (RFE) and Mutual Information (MI), for detecting stress in olive trees using hyperspectral data. The research was conducted in Halkidiki, Northern Greece, and focused on identifying stress caused by biotic and abiotic factors through the analysis of hyperspectral images. Both the RF and XGB algorithms demonstrated high efficacy in stress classification, achieving roc-auc scores of 0.977 and 0.955, respectively. The study also highlighted the effectiveness of RFE and MI in optimizing the classification process, with RF and XGB requiring a reduced number of hyperspectral features for an optimal performance of 1.00 on both occasions. Key wavelengths indicative of stress were identified in the visible to near-infrared spectrum, suggesting their strong correlation with olive tree stress. These findings contribute to precision agriculture by demonstrating the viability of using machine learning for stress detection in olive trees, and underscores the importance of feature selection in improving classifier performance. [ABSTRACT FROM AUTHOR]

The mapping of small-scale irrigation areas is essential for food security and water resource management studies. The identification of small-scale irrigation areas is a challenge, but it can be overcome using expert knowledge and satellite-derived high-spatial-resolution multispectral information in conjunction with monthly normalized difference vegetation index (NDVI) time series, and additional terrain information. This paper presents a novel approach to characterize small-scale irrigation schemes that combine expert knowledge, multi-temporal NDVI time series, multispectral high-resolution satellite images, and the random forest classifier in the Zamra catchment, North Ethiopia. A fundamental element of the approach is mapping small-scale irrigation areas using expert decision rules to incorporate the available water resources. We apply expert decision rules to monthly NDVI composites from September 2020 to August 2021 along with the digital elevation model (DEM) data on the slope, drainage order, and distance maps to derive the sample set. The samples were based on the thresholds obtained by expert knowledge from field surveys. These data, along with the four spectral bands of a cloud-free Planet satellite image composite, 12 NDVI monthly composites, slope, drainage order, and distance map were used as input into a random forest classifier which was trained to classify pixels as either irrigated or non-irrigated. The results show that the analysis allows the mapping of small-scale irrigation areas with high accuracy. The classification accuracy for identifying irrigated areas showed a user accuracy ranging from 81% to 87%, along with a producer accuracy ranging from 64% to 79%. Furthermore, the classification accuracy and the kappa coefficient for the classified irrigation schemes were 80% and 0.70, respectively. As a result, these findings highlight a substantial level of agreement between the classification results and the reference data. The use of different expert knowledge-based decision rules, as a method, can be applied to extract small-scale and larger irrigation areas with similar agro-ecological characteristics. [ABSTRACT FROM AUTHOR]