In recent years, there has been a growing use of unmanned aerial vehicle (UAV) based light detection and ranging (LiDAR) data for mapping plant area index (PAI) in orchards. However, using LiDAR time-series collected throughout the growing season to assess PAI variations in response to phenology, represents an understudied area of investigation. Furthermore, establishing the optimal spatial resolution for mapping biophysical variables of tree crops from LiDAR point cloud data remains poorly defined. Here, we assess the capability of a UAV-based LiDAR system to characterize cherry trees throughout the growing season, with a focus on monitoring PAI and the vertical structure of individual trees. A time-series of 14 point cloud acquisitions with a density of 3300 points/m2 was collected between February and December 2022, covering all phenological stages of a cherry orchard in southern France. A voxel-based method was applied to create a three-dimensional grid within which PAI was estimated for each voxel. PAI was mapped by accumulating the individual voxel-based PAI values within each vertical voxel column. The results demonstrate that a voxel size of at least 0.7 m is required to retrieve reliable PAI estimates (RMSE = 0.58 m2.m−2, MAE = 0.48 m2.m−2, bias = 0.19 m2.m−2, rRMSE = 23 This research presents an automated LiDAR-based computational approach to assess within-field variability in orchards, including plant area index, vertical structure, and canopy condition.

Abstract Developed by the King Abdullah University of Science and Technology (KAUST) to support research in atmospheric science and remote sensing, KAUSTSat represented Saudi Arabia's (and the Middle East's) first research‐focused hyperspectral CubeSat mission for Earth observation. The primary payload consisted of the Simera HyperScape50, a miniaturized hyperspectral sensor operating in the visible to near‐infrared range. The sensor was equipped with a custom continuous variable filter that collected imagery at 30 m spatial resolution with a 120 km swath. The HyperScape50 allowed for up to 32 user‐defined spectral bands to be selected per acquisition from a total of 442 programmable channels between 442 and 884 nm, including a panchromatic band. These capabilities enable detailed observations of vegetation, soil, coastal zones, and other surface features relevant to applications in agriculture, biodiversity, resource management, and disaster response. In this paper, we provide an overview of the mission architecture, sensor design, acquisition strategy, and data structure. The hyperspectral data sets acquired over the 14‐month mission lifetime will also be presented, with a particular focus on the Arabian Peninsula and RadCalNet calibration sites. The KAUSTSat mission serves as a demonstration case of the viability of academic‐driven CubeSat platforms for delivering targeted, high‐quality environmental data, and represents a valuable reference for future small satellite Earth observation programs.

Hyperspectral unmixing aims to decompose each pixel in a hyperspectral image into a set of constituent endmembers and their corresponding abundances. Recent deep learning based approaches have demonstrated strong performance in capturing both spectral and spatial features. However, obtaining reliable per-pixel abundance ground truth in real hyperspectral scenes is generally infeasible, which motivates unsupervised and self-supervised unmixing strategies. In this work, we propose S3ViT, a self-supervised Spectral Vision Transformer designed for pixel-wise hyperspectral unmixing. The transformer captures spectral and spatial dependencies by applying self-attention over the full sequence of pixel tokens (1×1) augmented with learnable positional embeddings. It operates without ground-truth annotations by generating pseudo labels through an unsupervised process: first, Singular Value Decomposition (SVD) is used to estimate the number of endmembers based on a thresholded singular value spectrum; then, k-means clustering provides cluster-derived priors that are used to form a contextual token and initialize spectral prototypes, without being treated as true abundance supervision. To guide training, two initialization tokens, one from Vertex Component Analysis (VCA) and one from the k-means cluster-derived priors, are embedded alongside patch tokens into the transformer. The model learns to estimate abundance maps while enforcing the Abundance Non-negativity and Sum-to-One constraints through ReLU and Softmax layers. Endmember spectra are later estimated from pixels with high predicted abundances. We evaluated S3ViT on the Samson, Jasper Ridge and Washington DC Mall benchmark datasets and compared it to state-of-the-art geometrical and deep learning methods. Our model achieves superior or comparable performance in both RMSE and SAD metrics, with up to 31% improvement in SAD and 25% in RMSE. These results indicate that a compact pixel-token ViT, guided by weak spectral priors and optimized via reconstruction losses, can achieve competitive unmixing performance on standard benchmarks.

Fractional Vegetation Cover (FVC) is a key ecological variable for monitoring ecosystem health, land degradation, and vegetation dynamics in dryland environments. While satellite and UAV observations enable scalable FVC estimation over large spatial extents, the accuracy and robustness of these models remain strongly dependent on high-quality field-based reference data for calibration and validation. Traditional in-situ methods, including visual estimates using transect-based surveys, remain widely used but are labor-intensive and inherently subjective. Digital photography has emerged as a practical alternative, typically analyzed using index-based computer vision techniques or deep learning models. However, these methods are highly sensitive to background variability and therefore rely on massive labeled datasets. Recent advances in multimodal large language models (MLLMs) suggest a potential paradigm shift, as these models combine visual perception with high-level reasoning and benefit from diverse pre-training that enables conceptual knowledge transfer across tasks. In this study, we evaluate the feasibility of using MLLMs for direct estimation of FVC from ground-level photographs without task-specific training. We collected and compiled a dataset of more than 1,100 quadrat pictures from across 26 dryland sites in Saudi Arabia, spanning a wide range of surface conditions from bare soil to sparsely vegetated rangelands. Each picture corresponded to a 1 m × 1 m quadrat with FVC estimated independently by two experts, whose average was used as reference data for assessment of model predictions. Six state-of-the-art multimodal large language models, including Qwen2.5-VL, Mistral-Small-3.2, LLaMA-4-Maverick, LLaMA-4-Scout, and two Gemma-3 variants, were evaluated using four prompt designs that varied in length, ecological context, and methodological detail. Across all models and prompts, MLLMs achieved a mean absolute error of approximately 7.8%, demonstrating competitive performance relative to traditional image-based methods. The best-performing model-prompt combinations achieved mean absolute error values below 5%, with low systematic bias. Short and ecologically explicit prompts consistently outperformed more complex prompt designs, achieving an average reduction in mean absolute error (MAE) of approximately 1.3–1.4 percentage points compared to visually guided or highly structured prompts (MAE ≈ 6.9% versus 8.2–8.4%). Overall performance was more sensitive to model choice than to prompt structure, with mean MAE varying from approximately 5.6% to 10.0% across models, compared to a narrower range across prompts. The highest accuracy was obtained using the Qwen2.5-VL model with an ecologically detailed prompt, which achieved a mean absolute error of 4.9%, near-zero bias, and an RMSE of 8.4%. Across all prompt designs, Qwen2.5-VL and Mistral-Small-3.2 consistently delivered the best overall performance, both maintaining mean MAE values below 6% and exhibiting stable behavior across prompt variations, indicating robustness to prompt design. These results demonstrate that MLLMs can provide accurate and scalable FVC estimates directly from field photographs, without requiring specialized training datasets. This approach offers a promising alternative for rapid field surveys and reference data generation, particularly in dryland ecosystems where background complexity and data scarcity limit the effectiveness of conventional methods.

Unoccupied Aerial Systems (UAS) have emerged as powerful tools for ultra-shallow (<2 m) bathymetric mapping using Structure-from-Motion (SfM) or spectrally derived bathymetry (SDB) based methods. However, the impact of equipment selection on the accuracy of bathymetric maps created through these methods is not well characterized. This study investigates how different UAS navigation and imaging sensors impact bathymetric accuracy and whether this differs for SfM- and SDB-based methods. We assessed five common UAS across two coral reef habitats. real-time kinematics or post-processing kinematics based positioning was critical for accurate SfM-based bathymetry (RMSE = 0.14 +/- 0.02 m). However, shallow water adaptations of Stumpf's band ratio trained on as few as 10 ground truth samples consistently outperformed SfM, regardless of the UAS configuration (RMSE = 0.11 +/- 0.003 m). Spectrally trained support vector machines and single-layer neural networks demonstrated the strongest performance (RMSE = 0.09 +/- 0.005 m) but required 200 ground truth measurements to reach such accuracies. Multispectral sensors improved the accuracy of spectrally derived bathymetric maps, but broadband RGB cameras also performed well (RMSE difference <= 0.06 m). SfM-derived bathymetry could be combined with spectral models to improve bathymetry reconstruction in texture-poor sand habitats. However, the overall accuracy of such combined models remained constrained by the accuracy of the original SfM reconstruction. Ultimately, Stumpf's band ratio using the red image band as the denominator provided the most practical and accurate method for UAS-based ultra-shallow bathymetry retrieval when ground truth data was sparse, while spectral machine learning models performed the best when ground truth data were abundant.

Foundation models offer a promising route to transferable remote sensing representations, but many current approaches depend on very large pretraining datasets and fixed sensor configurations, limiting their suitability for ecological and environmental applications, where observations often vary across platforms, spatial and spectral resolutions, and available modalities. We introduce FLORO, a multimodal geospatial foundation model designed to learn transferable representations from a small but highly diverse remote sensing corpus. FLORO is pretrained using masked autoencoding on a heterogeneous combination of Sentinel-1, Sentinel-2, SkySAT imagery, elevation, and UAV-derived data. To accommodate sensor variability, FLORO incorporates availability-aware inputs that indicate which spectral bands and auxiliary modalities are present in each sample, enabling a unified input space across heterogeneous sensor configurations. We evaluated FLORO on the PANGAEA benchmark under a frozen-encoder protocol across scene classification, segmentation, and regression tasks. Despite being pretrained on a smaller corpus than competing foundation models, FLORO achieved strong and stable transfer across optical, optical-SAR, and optical-elevation benchmarks spanning medium-resolution satellite, airborne, and ultra-high-resolution UAV imagery. FLORO obtained the second-best average segmentation performance across six PANGAEA benchmarks, trailing only a recently introduced foundation model pretrained on over two orders of magnitude more images, remained competitive on scene classification, and was robust in regression tasks, while qualitative results showed improved preservation of spatial structure in flood, urban, biomass, and canopy-height prediction settings. In a separate controlled experiment on EuroSAT-MS, geo-positional encoding further improved classification relative to absolute positional encoding.

IntroductionIn arid and hyper-arid regions, agriculture depends heavily on irrigation, making crop type monitoring important for water allocation, monitoring crop management policies, and providing the information required to forecast food supply. However, field labels are often scarce, and crop calendars can shift due to locally managed planting, harvest, and irrigation decisions, complicating mapping at field-scale.MethodsWe present a seasonal crop type mapping approach applied to Wadi Al-Dawasir, Saudi Arabia, generating maps for 2020–2024 from biweekly optical and radar satellite time series. The method learns representations from unlabeled imagery through self-supervised pretraining and fine-tunes a segmentation model using a small set of field observations with pseudo-label augmentation from unsupervised clustering. We mapped four classes: fallow, cereal, vegetables, and forage, and evaluated performance using overall accuracy, F1-score, and mean intersection-over-union (mIoU).ResultsThe configuration combining self-supervised pretraining, pseudo-label augmentation, and optical-radar inputs achieved an mIoU of 0.80, an F1-score of 0.88, and an overall accuracy of 0.97. In contrast, using optical information alone substantially reduced accuracy, with an mIoU of 0.35, an F1-score of 0.32, and an overall accuracy of 0.64. Fallow and forage produced the highest mapping accuracies, with mIoU values of 0.97 and 0.85, respectively, while vegetables had the lowest accuracy among the four crop groups, with an mIoU of 0.52.DiscussionThese results show that self-supervised temporal pretraining combined with pseudo-label augmentation can support efficient multi-season, field-scale crop mapping using limited labels in irrigation-driven arid regions. The lower accuracy for vegetables highlights the continued challenge of mapping heterogeneous crop groups with overlapping phenological patterns.

Monitoring crop conditions is crucial for effective crop management and provides valuable insights into soil-plant-atmosphere interactions. While some studies have used unmanned aerial vehicle (UAV)-based light detection and ranging (LiDAR) data for mapping plant area index (PAI) in orchards, LiDAR-based time-series analysis to assess PAI variations with phenology throughout the growing season represents a significant gap in knowledge. Tracking PAI dynamics across phenological stages reflects canopy development and leaf expansion, which are directly linked to yield formation. Furthermore, the optimal spatial resolution for mapping biophysical variables of tree crops from LiDAR point clouds is yet to be determined. This study aimed to demonstrate the potential of UAV-derived LiDAR time-series to monitor the PAI and tree vertical profiles at high spatial resolution throughout the growing season of a cherry orchard located in southeastern France. A time series of 14 point cloud acquisitions with a density of 3300 points/m² was collected between February and December 2022, with at least one acquisition per month, covering all phenological stages of the cherry orchard. Field measurements were collected on May 30, and October 6, to measure the PAI at twilight using an LAI-2200C Plant Canopy Analyzer (LI-COR Biosciences, Lincoln, NE, USA), with 248 trees sampled. A voxel-based method was applied on the LiDAR point cloud data to create a three-dimensional grid within which PAI was estimated for each voxel. The results showed that a voxel size of at least 70 cm is required to retrieve reliable PAI estimates, while a voxel size of 100 cm produced the most accurate PAI estimates (RMSE = 0.5 m2.m-2, bias = 0.07, R2 = 0.59), when assessed against in-situ PAI measurements. The temporal variation of canopy PAI illustrated the progression of the phenological stages, including flowering, leaf development, ripening and senescence, and the response of the canopy to drought stress (reduction in PAI due to leaf rolling) during the summer. The maps of PAI successfully described the variations in leaf canopy density for different cherry varieties and allowed assessment of the vertical PAI profile at the individual tree level. The LiDAR-derived PAI maps and vertical profiles were able to detect trees exhibiting poor leaf development, which is an important health indicator for effective crop management in orchard settings. Future work should focus on applying UAV-derived observations to optimize crop models to enhancing decision-making tools for effective orchard management.

Integrating data from both unmanned aerial vehicle (UAV) and satellite platforms, UAV-satellite synergies can effectively leverage multi-scale crop information extraction. Although previous reviews have discussed UAV-satellite synergies, they have either adopted a broader scope beyond crop monitoring or focused mainly on specific aspects such as optical data or data fusion methods. This paper provides an updated and crop monitoring specific synthesis by systematically analyzing peer-reviewed studies and offering a structured framework of data sources, synergy strategies, applications, key challenges, and future directions. Our findings reveal a strong concentration of optical sensor combinations and model-based learning strategies, particularly for crop trait estimation and classification. However, the synergy field remains in its infancy, facing persistent challenges in multi-source data matching, strategy design, and practical deployment. Spatial, temporal, and signal mismatches hinder data compatibility and model transferability, while many current strategies rely on empirical routines with limited generalization. To address these limitations, we propose three future directions: enriching multi-source and multimodal observations, enabling modular and adaptive synergies for complex applications, and integrating emerging and cross-domain technologies such as UAV swarms, autonomous deployment systems, and precision agriculture infrastructures. These insights provide a theoretical foundation and practical guidance for advancing UAV-satellite synergies specially for agricultural remote sensing.

In line with Saudi Vision 2030 and the ambitions of the Saudi Green Initiative to plant 10 billion trees and protect 30% of Saudi Arabia’s land and sea areas, large parts of Saudi Arabia are being transitioned into nature reserves and undergoing regreening activities, while also fencing off large areas for vegetation restoration and protection. To effectively manage the regreening and restoration initiatives, it is imperative to frequently monitor biomass changes over time and quantify the accumulation of biomass to determine if the restoration and tree-planting initiatives have the intended outcomes. However, landscape responses to regreening and vegetation restoration are poorly understood in hyper-arid rangelands. Environmental processes within different habitat zones might differ, which further complicates biomass monitoring. Here, we present a framework that is currently being applied across multiple hyper-arid rangelands in Saudi Arabia to estimate biomass within newly established nature reserves. The initial work focused on developing a scaling approach between ground, unmanned aerial vehicle (UAV), and satellite image data, including PlanetScope and Sentinel-2 imagery. Field sites were identified using Google Earth imagery and a number of criteria, including a 5-year Sentinel-2 NDVI time-series to identify greening events, terrain characteristics based on DEM data, and differences in vegetation functional types (annual and perennial grass/herbs/forbs, shrubs and trees), soil types, and habitats. Field-based measurements at selected sites focused on determining biomass of different vegetation functional types, using a double-sampling approach, including a limited number of destructive samples, and a large number of biomass estimates based on the structural and dimensional characteristics of the destructive samples. UAV-based light detection and ranging (LiDAR) and multispectral data were obtained for each site to estimate biomass from the field-based samples using different machine and deep learning approaches (random forest, support vector machines, vision transformer, UNet with attention encoder). It was found that LiDAR data provided useful information on vegetation height and volume, which improved the biomass estimates, whereas the multispectral data enabled discrimination between photosynthetic and non-photosynthetic vegetation components. Based on the UAV-derived estimates of biomass, a scaling approach was applied to estimate biomass from both PlanetScope and Sentinel-2 data, with results demonstrating that the higher spatial resolution of the PlanetScope data improved the accuracy due to the very sparse vegetation in most parts of the hyper-arid rangelands of this study. While hyper-arid rangelands are generally underrepresented in research studies worldwide, this research provide a viable framework for assessing vegetation dynamics of biomass both seasonally and annually.

Current research on livestock movement ecology focuses on quantifying the factors that trigger alterations in movement behavior and understanding hidden mechanisms. Modern tracking technologies and robust statistical analysis models deliver new opportunities for investigating how individual animals cope with the joint effect of biotic and abiotic factors at different time scales. We applied multivariate Hidden Markov Models (HMMs) to characterize the fine-scale movement behavior (30-second intervals) of GPS-tracked domestic Zhongwei goats (Capra aegagrus hircus) for 124 days and analyzed the combined influence of biotic and abiotic factors and specific time of day on their seasonal movement behavioral transition in a predator-free, semi-arid mountain grassland in China. We classified the behaviors of goats into two states: foraging (low step length, varied turning angle) and travelling (long step lengths, small turning angles). The terrain slopes had the most impact on their movement behavioral transition in the full year, spring, autumn, and winter. However, in the summer with hotter temperatures, the specific time of day explains their movement behavior most. Forage resources indicated by the Normalized Difference Vegetation Index (NDVI), and terrain ruggedness measured by the Vector Ruggedness Measure (VRM), had less impact on their behavior transitions compared to terrain slope and specific time of day. Elevation and solar radiation could not explain their movement behavior in different seasons, nor could NDVI in winter or VRM in spring and autumn. Across different seasons, the probability of foraging behavior increased with the later times of day, steeper terrain slopes, and higher NDVI, while it decreased with increasing VRM. The impact of NDVI on the probability of foraging behavior was largest during the early onset of vegetation growth in spring, and lowest in winter coinciding with a lower availability of food resources. The movement speed was lower, and the daily foraging percentage was higher in spring and winter due to lower food resources and shorter daylight hours. In contrast, movement speed was higher, and the daily foraging percentage was lower in summer and autumn with more food resources and longer daylight hours. The percentage of time allocated to foraging increases hourly from 9:00 am to 8:00 pm across various seasons. HMMs were found to be useful for disentangling the movement behavior of goats. Our approach provides new insights into the seasonal and daily behavioral strategies of goats. Results demonstrate that in the mountain region, terrain slopes and specific times of the day more effectively trigger domestic goat behavioral transition from one state to the next compared with biotic factors, represented herein by NDVI, across different seasons. The early onset of vegetation growth and a shorter period of available high-quality forage in spring, significantly influenced goat behavioral transitions. Overall, these results are important for designing appropriate grazing management strategies that satisfy the ecological and socioeconomic demands of semi-arid grassland ecosystems.

Soil moisture (SM) plays a central role in water cycle dynamics and land-atmosphere interactions, acting across local and regional scales. Few studies have explored the use of the ground-based global navigation satellite system reflectometry (GNSS-R) interference pattern technique (IPT) for SM estimation. In these studies, SM was estimated from the GPS elevation angle where lower reflectivity occurs (notch), which is difficult to determine in real GNSS-R interference power (IP) acquisitions. This study introduces the use of IP amplitude at vertical polarization (V-pol), readily extracted from the IP oscillations, as an alternative for SM estimation beneath vegetation cover. An empirical model was developed for estimating SM in irrigated grassland using a GNSS-R receiver with a linearly polarized antenna. The experiment, conducted between June 6 and August 8, 2022, covered the grassland's growth phase and preharvesting and postharvesting. The study incorporated normalized difference water index (NDWI) from the Sentinel-2 satellite to account for vegetation's impact on IP amplitude. Results indicated that the IP amplitude at V-pol accurately estimates SM (RMSE =0.04 m3/m3). Moreover, the results show that the vegetation layer mainly attenuates the IP amplitude with a nonsignificant scattered contribution to the IP, allowing for the simplification of the empirical model by ignoring the scattered contribution of vegetation. The simplified empirical model can be numerically resolved to estimate the NDWI if the SM is known. In summary, this study highlights the effectiveness of the ground-based IPT for close-range sensing of SM and a biomass proxy, such as NDWI.

Key research in movement ecology is investigating shifts in animal behavior and identifying the factors that induce alterations in movement behavior and mechanics. The impact of natural environments and human activities on the underlying behavioral processes of domestic goats are still being elucidated. We applied seasonal multivariate Hidden Markov Models (HMMs) to characterize the fine-scale movements (30- second intervals) of GPS-tracked Zhongwei goats for 124 days and determine how grazing intensity, seasonal food resources, terrain factors and daylight hours affect movement behavior in the mountain grassland in China. We classified the goats’ activities as two basic behavioral states of foraging (low step length, varied and undirected turning angle) and travelling (long step lengths, low and directed turning angles). Grazing intensity, a management factor, exerted the most significant influence on goats across different seasons. Additionally, factors such as daylight hour and slope had a more pronounced impact on their movement activities compared to the normalized difference vegetation index (NDVI). Elevation and solar radiation were found not explain much of the variability in movement behavior of goats. Their probability of foraging behavior was most likely to increase with grazing intensity, slope, diurnal hours and NDVI. In addition, the percentage time allocation of foraging was higher in spring and winter with lower food resources periods and shorten daylight hours, than summer and autumn with larger food resources and long daylight hours. The foraging percentage increased from morning to afternoon. HMMs are found useful for disentangling movement behavior and understanding how goats respond to seasonal grazing intensity, time of daylight, NDVI and slope. Our findings underscore the importance of accounting for interactions between movement behavior and gazing management, not only the environmental factors and behavioral rhythms, when assessing the movement characteristics and behavioral transitions of goats. These results are important for designing grazing management strategies that satisfy ecological and socioeconomic demands on mountain grassland ecosystems.

Biological invasions pose significant threats to ecosystem structure and function, disrupt ecosystem services, cause high economic losses, and negatively impact human well-being. However, accurate prediction of species distribution is a critical challenge in ecological and biodiversity conservation. This study compares the predictive performance of 10 machine learning algorithms, including random forests, maximum entropy, support vector machines, and others, by integrating global occurrence records with climatic, edaphic, and human activity variables to identify the most robust model for predicting the global distribution of the invasive weed, Conyza sumatrensis (Retz.) E.Walker. Different algorithms yielded large variations in the predicted area of C. sumatrensis. Among these, random forests had the highest performance accuracy metrics and high agreement of predictions, aligning well with global occurrence records, and are used to explain and predict the potential distribution of C. sumatrensis. Distributions of C. sumatrensis are mainly influenced by temperature variables, adapted to a wide range of precipitation and various soil conditions, and facilitated by human activities. Currently, C. sumatrensis is distributed widely across all continents (6.20 Mkm2). The suitable habitat for C. sumatrensis is projected to have an increase of 8.03-8.78 % by 2041-2060 and 0.84-3.29 % by 2081-2100. In addition, the global extent of suitable environmental conditions for the establishment and spread of C. sumatrensis was anticipated to expand in urban and farmland by 2081-2100. The results provide an early warning of specific land-use types at higher risk of C. sumatrensis extent, offering valuable insights for managers to develop targeted prevention and control strategies. Additionally, to enhance predictive accuracy, our study underscores the critical role of selecting suitable algorithms and integrating human activity factors into invasive species distribution models.

As global biodiversity faces increasing threats from climate change, habitat loss, and human activities, effective methods for assessing and monitoring biodiversity are crucial. Drylands are particularly vulnerable to the impacts of climate change, and integrating remote sensing technology with ecological research can help protect these environments. Identification of individual trees and shrubs is fundamental to assess biodiversity and can also improve carbon stocks estimates. However, identifying individual trees using medium resolution satellite images is often not feasible. The use of advanced technologies, such as machine learning and satellite imagery in environmental management plays a key role in biodiversity conservation and can potentially fill this gap. The use of readily available Maxar satellite imagery makes conservation approaches accessible and cost-effective, which is crucial for widespread adoption, especially in resource-limited settings or for large-scale studies. This study aims to improve the identification of vegetation in dryland ecosystems by integrating deep learning methods to remote sensing. The primary objective was to distinguish vegetation from rocks or shadows in these areas, which is often challenging due to the dark appearance of vegetation and the similar visual features of the landscape, such as landforms textures, water bodies or man-made objects.. To address this challenge, a Vision Transformer (ViT) deep learning model was developed to estimate near infrared (NIR) spectral bands from high resolution Maxar Satellite images. By enhancing the spectral richness, the model aids in the differentiation of vegetation. Maxar satellite imagery was primarily used because of its accessibility through Google services, making it ideal for planning initial surveys to identify areas of interest for more detailed study. The accuracy of the model was validated against high-resolution SkySat NIR imagery, and achieved an R2 of 0.92. The obtained NIR band helped to clearly distinguish vegetation from non-vegetative surfaces such as soil, rocks, and water, which were not as discernible in RGB imagery alone. The use of a deep learning method to estimate a synthetic NIR band proved to be cost-effective and efficient for large-scale identification of individual trees and shrubs in drylands, overcoming the limitations of medium-resolution satellite imagery. The findings of the study are crucial to the conservation of biodiversity and offer a practical approach for environmentalists and researchers. Future work includes expanding the dataset to include various dryland environments, and integrating additional data sources such as soil data and topographic features for a more comprehensive analysis.

The Arabian Peninsula, a predominantly hyperarid region, is facing increasing demands for freshwater driven by its growing population. The pressure on the resource would be reduced if local evaporation led to more rainfall within the domain. This study assesses moisture recycling in the region by evaluating moisture transport, precipitation, and evaporation in four state-of-the-art reanalyses against in situ and remote sensing data. The reanalyses reliably capture the main spatial and temporal features of the transport, in particular its two-layered structure. The upper layer, above 900 hPa, is marked by an anticyclonic circulation which has moisture enter in the northwest and exit in the southwest with little net effect. The lower layer is shaped by an influx of moisture from sea breezes during the day and an outflux during the night. However, the reanalyses underestimate daytime sea-breeze imports compared to radiosonde data, leading to a net average outward fl ow of moisture from the domain. This diurnal imbalance is likely not due to desert agriculture because the reanalyses do not explicitly consider irrigated cropland. Certain reanalyses also report nonphysical residual evaporation across the region, contradicting remote sensing data. Agreement between reanalyses and observations is better regarding precipitation because it is controlled in part through synoptic conditions. Despite the uncertainties, the magnitude of incoming transport relative to evaporation suggests that internal moisture recycling is negligible. As a consequence, increased evaporation due to incremental land-use changes and irrigation is unlikely to enhance regional rainfall.

Vegetation phenology, encompassing critical events like leaf emergence and maturity, serves as an important indicator of adaptive plant responses to environmental factors. In the context of Saudi Arabia, existing crop phenology retrieval methods encounter several challenges related to local farm management operations. These can include unstable crop calendars with planting and harvesting at any time throughout the year, uncertainty in sub-field management with independent control of areas within a center-pivot field, and diverse crop rotations between fodder and non-fodder crops. To address these challenges, we present an innovative framework utilizing machine learning and Sentinel-2 NDVI time series data for mapping phenology stages of key crops at a national scale. The framework is composed of three modules that are implemented step-wise, including: (1) a within-field dynamic clustering module (termed WithinFDy) that monitors fields for potential subdivision based on pixel-level NDVI temporal dynamics; (2) a phenology estimation module (termed PhenoEst) that segments NDVI time series into growing seasons and extracts essential phenology stages (e.g., planting and harvesting dates) for each season; and (3) a crop type discrimination module (termed CropDis) that utilizes extracted phenology information as input features to discriminate between different crop types. Evaluated on 1,000 randomly selected fields in northern Saudi Arabia, our framework achieved overall accuracies of 93.38%, 96.40%, and 94.39% for WithinFDy, PhenoEst, and CropDis modules, respectively. When applied nationwide in 2020, the framework revealed valuable insights. In terms of field management, 21.8% of the fields were divided into two distinct subfields, featuring different planting and harvesting dates - and sometimes crop type, while 73.2% showed consistent practices across the entire field. For seasonal dynamics, 53.4%, 36.3%, and 8.7% of fields supported crops for one, two, and three seasons annually, respectively. Main planting and harvesting activities occurred during winter seasons (November to February), with another peak observed in June. Approximately 30% of fields were under production for 5 to 6 months, and 15.7% were under production year-round. The dominant crop types in 2020 were fodder crops (e.g. alfalfa and Rhodes grass), followed by winter crops like winter wheat. Our methodology represents a substantial advancement over previous approaches, expanding applicability beyond crops with regular growth patterns. The results not only enrich agricultural datasets in Saudi Arabia but also hold promise for enhancing food and water security studies globally.

The AquaCrop model is a powerful tool for crop monitoring, providing a daily estimation of soil-crop-atmosphere dynamics. The model requires a substantial number of input variables and parameters, highlighting the need for identifying those that significantly influence model outputs. Sensitivity analysis is a vital method for this purpose. A key objective of this study is to examine the performance of the AquaCrop model in simulating wheat yield and irrigation water requirement in drylands under two scenarios: first running the model employing a minimal amount of in situ data, and second using all available in situ data. A second focus is to analyze the sensitivity to all crop and soil related input variables and parameters. To do this, a pilot-scale study was undertaken, focusing on a commercial farm in the Al-Jouf province of Saudi Arabia. The farm comprised 200 center-pivot fields of mainly wheat crops. In situ data was collected to calibrate the model for two consecutive growing seasons (2019-2020 and 2020-2021). Using the variance-based Sobol technique, the sensitivity of the AquaCrop model outputs, particularly wheat yield and irrigation water requirement, to crop and soil related input variables and parameters was examined, as were the influential and non-influential inputs on these outputs. Results showed that the second scenario (all data) outperformed the first (minimal data), demonstrating more accurate wheat yield predictions with rRMSE values of 17% and 21% for the 2019-2020 and 2020-2021 growing seasons, respectively. Regarding irrigation water requirement estimations, the second scenario also exhibited lower rRMSE values of 20% and 19% for the same growing seasons. Results also demonstrated that the sensitivity indices of variables and parameters varied with model outputs and growing seasons. By synthesizing inputs sensitivities under different conditions, the influential input variables and parameters were distinguished. Overall, six variables and parameters held significant influence on the analyzed model outputs based on their total-order sensitivity indices. These included duration from sowing to senescence (senescence), duration from sowing to harvesting (maturity), duration from sowing to yield formation (HIstart), base temperature below which growth does not progress (Tbase), minimum air temperature below which pollination failure begins (Tmin_up), and shape factor describing reduction in biomass production (fshabe_b). It was revealed that most variables and parameters were non-influential, which might allow them to be fixed within their ranges to optimize model calibration. The research represents the performance assessment and sensitivity analysis of the AquaCrop model over a desert farming system and offers guidelines for model calibration by delivering information on influential and non-influential input variables and parameters.