Centre d'Études Spatiales de la Biosphère

Quantifying the evaporative flux from the land surface into the atmosphere is important for local irrigation management, regional water management and global weather prediction, but measuring evapotranspiration (ET) remains a challenge. Methods that directly measure ET, including eddy covariance and lysimeters, have high temporal resolution but limited spatial coverage, while indirect estimates, including remote sensing techniques, have broad spatial coverage but lower temporal frequency. Each principle for estimating ET comes with unique uncertainties, so a combined use of these methods under identical conditions, which is rarely done in field campaigns, should be promoted to assess its relative and overall performance. In order to quantify the relative uncertainty of different commonly-used ET estimation techniques, we performed an ET method intercomparison using data from the 2021 Land surface Interactions with the Atmosphere over the Iberian Semi-arid Environment (LIAISE) experiment (Boone et al., 2025). Fifteen ET estimation methods were evaluated across six crop types, spanning footprint extents from individual fields to the regional scale and encompassing multiple measurement principles. We also include an upscaled “mixed-agriculture” estimate from the eddy-covariance data to match the footprints of methodologies that cover more than one field (e.g. satellite remote sensing data and long-path scintillometry) We find that the total sampling uncertainty in the eddy covariance measurements is between 10% and 20% of the latent heat flux regardless of time of day. Moreover, we find that standard deviation among different ET methods for a single crop type (between 2.0 and 3.7 mm day-1) is greater than the standard deviation among the different crop types (1.9 mm day-1) in the LIAISE domain.

Reliable crop mapping is essential in land and water management studies to understand the spatial distribution and dynamics of agricultural practices, to model resource use and production, and to propose sustainable scenarios for agricultural and water management. This work presents high-resolution crop mapping for the Mediterranean area, which is particularly interesting due to the limited data availability and the high level of land use heterogeneity. The main European land use dataset, Corine Land Cover (CLC), lacks the specificity required for accurate agricultural classification, especially for crop differentiation, and does not provide frequent or timely updates, which are crucial for many applications. Other more recent EU-wide crop mapping efforts (d’Andrimont et al, 2021) still lack regional accuracy due to widely scattered training data. To overcome these limitations, a large-scale crop mapping initiative was implemented in Sardinia to test and validate an artificial intelligence–based approach for Mediterranean environments. In this context, irrigated agriculture is a key sector for the sustainable management of limited water resources. The method uses Sentinel‑2 time series and survey data from the LPIS (Land Parcel Identification System). The study relies on IOTA2, a land‑use map production chain first developed and tested at the French level, producing maps with 24 land‑use classes. The originality of the approach lies in the use of open‑source satellite images and an automated processing workflow based on supervised classifiers, making crop mapping faster and easily reproducible across years.Learning samples are derived from 2018 LPIS data, supplemented by CLC and CLCplus Backbone datasets for natural areas and the Urban Atlas for urban areas. Two nomenclatures are tested: a detailed versus a simplified one, with 32 and 26 thematic classes, respectively, both focusing on Mediterranean-relevant crop typologies. The two nomenclatures are evaluated with sampling rates of 10%, 50%, and 100% of training pixels. Results show that the simplified nomenclature achieves higher accuracy, with an Overall Accuracy (OA) of 0.77 compared to 0.61 for the detailed nomenclature, using 100% training pixels. Increasing the training sample rate improves classification quality in both nomenclatures: in the short nomenclature, OA values are 0.596, 0.613, and 0.774 for 10%, 50%, and 100% sampling rates. In the detailed nomenclature, the improvement is weaker, with OA values of 0.596, 0.601, and 0.610, indicating that increasing sample size does not resolve class confusion. Among agricultural classes, rice, citrus, vegetables, and grapevine achieve the highest classification scores, which are among the crops with the largest irrigation requirements. Nuts, cereals, and fruit trees perform poorly, mainly due to insufficient training samples. Overall, the proposed nomenclature significantly improves the crop classes available in the CLC by increasing crop specificity and differentiation. This study presents a framework for fully automatic crop‑map production in Mediterranean environments, ensuring fast reproducibility over the years thanks to the use of openly accessible satellite imagery and an automated processing chain. This can improve the accuracy and reliability of water accounting for the agricultural sector and help promote sustainable use of limited water resources in the Mediterranean areas.

This study analyses the potential of a change detection method—the near-real-time cumulative sum change point detection method (CuSum-NRT)—applied to Sentinel-1 C-band synthetic aperture radar (SAR) data to monitor crops and identify field work based on a monthly number of changes. The temporal evolution of the number of changes occurring on Sentinel-1 backscatter at both VV and VH polarisations averaged at field scale was analysed over five years (2017–2021) and compared with NDVI derived from the Sentinel-2 multispectral instrument (MSI) sensor over more than 1000 fields in the southwest of France. The monthly number of changes detected did not show a significant difference between months with soil work and months with no soil work, so further analysis on the dates of changes should be conducted. The number of changes based on VV was found to poorly correlate with the VV backscatter (Rglobal = 0.25), and that based on VH was found to moderately correlate with the VH backscatter (Rglobal = 0.61): CuSum provides additional information compared to backscatter alone. The results also showed that the number of changes detected using CuSum-NRT is correlated to NDVI, mostly positively for the VH polarisation (Rmax = 0.73, p-value < 0.05) and negatively for the VV polarisation (Rmin = −0.69, p-value < 0.005). Furthermore, the analysis of crop groups (cereals, oilseeds, protein crops, fodder, vegetables, others) displayed statistically significant differences in terms of the annual number of changes occurring on both VH and VV polarisations, which has potential applications in crop classification.

<strong class="journal-contentHeaderColor">Abstract.</strong> Incorporating human influences within water resources modelling, in the context of global change, has proven to be a fruitful approach for improving the assessment of the impact of climate and water demand changes on hydrology and water demand satisfaction. In this study, we use an integrated water resources management model that details water withdrawal for irrigation, drinking water supply, cattle watering and industry, as well as releases from wastewater treatment plants, drinking water network leakages, and industrial activities, within a catchment subject to human influence. Using this modelling approach and collaborating with relevant stakeholders through workshops, we developed a series of future water demand scenarios to examine the sustainability of water use in the future. Our findings indicate that climate change will be the primary driver of changes in water resources and water demand satisfaction. Moreover, we found that low flows and water demand satisfaction will greatly decline in the future. A single climate projection indicates a less drastic deterioration of the system in certain areas of the catchment. We found that adapting water uses could help mitigate the negative impacts, though it is not fully satisfactory. The irrigation sector is set to be the most impacted in terms of water demand satisfaction. The study presents a methodological framework that helps to provide water sector managers with tailor-made results to support the design of effective adaptation measures.

This paper presents a fully automated, open-source workflow for mapping and analysing Surface Urban Heat Island Intensity (SUHII) in urban areas. The method integrates freely available Earth Observation datasets, including Landsat imagery and SRTM elevation data, and ensures generalizability across diverse landscapes by analysing homogeneous altitudinal zones. This avoids over- or underestimation of thermal anomalies due to elevation differences. The pipeline requires only a city name and a local folder path as inputs, producing standardized warm-season Land Surface Temperature and SUHII maps in GeoTIFF format with full metadata. Developed in modular R and Python environments, the code supports easy adaptation, dataset integration, and deployment in web or desktop applications. Designed to lower technical barriers, the platform merges technical and experiential knowledge through citizen participation, enabling non-experts to explore urban thermal anomalies and their social and health implications. This approach fosters equitable climate adaptation and environmental justice by making climate-relevant spatial data accessible and actionable for citizens, researchers, and decision-makers. • A bottom-up project born from the encounter between science and activism, to foster knowledge and empower society. • A standardised, automated workflow to map Surface Urban Heat Island Intensity by leveraging globally available open data. • Mid-latitude cities are among the most vulnerable to Surface Urban Heat Island effects.

Abstract Over the past four decades, seasonal snow cover has declined rapidly in temperate alpine regions. However, the fine-scale dynamics of snowmelt preceding the ongoing warming period remain largely unknown, limiting our understanding of the long-term influence of past snow cover on alpine ecosystems. Here we rely upon the spatial similarities in melt-out patterns and a temperature-based model of fractional snow cover area, to reconstruct fine-scale snow cover changes over the past 250 years in instrumented catchments of the southwestern Alps. We provide evidence that, until the 1980s, prolonged snow cover in many late-lying snowfields delayed ecosystem development and explain why current vegetation cover, soil organic matter content, and mineral weathering are significantly lower in these areas than in surrounding ecosystems. These findings highlight the long-term legacy of snow cover on alpine landscapes and underscore the need to re-evaluate its effects on ecosystem structure, functioning, and responsiveness to ongoing changes.

Abstract. Surface reflectance (SR) is essential for many remote sensing applications, but retrieving it from aerial images is challenging due to the lack of in-flight radiometric calibration and varying acquisition conditions. We propose a novel method for radiometric cross-calibration of aerial imagery using satellite Top-of-Atmosphere (TOA) reflectance. The method involves estimating at-sensor reflectance at the airborne altitude from satellite TOA reflectance, followed by spectral band adjustment and spatial alignment between satellite and airborne imagery. A linear radiometric model is derived to relate the Digital Number (DN) to the at-sensor reflectance from a selected subset of robust aerial-satellite pixel correspondences. The radiometric calibration parameter was retrieved using linear regression. The method is particularly suitable for airborne campaigns that lack onboard or in-situ radiometric calibration equipment. An ablation study is presented to analyze the selection of reliable reference pixels.

Urban areas, with their abundance of materials that absorb solar radiation and high population density, are particularly vulnerable to the heat island effect. Mitigating this phenomenon requires accurate estimates of surface albedo and emissivity, yet such measurements are often unavailable at a citywide scale. High-resolution optical and thermal-infrared satellite sensors offer a solution, overcoming the spatial and temporal limitations of in-situ observations. However, satellite measurements are constrained by limited acquisition geometries and spectral bands, which prevents the accurate computation of neded quantities such as urban albedo maps.This contribution describes the SuaBe project (https://stereoiv-suabe.eu/en), which focuses on designing and implementing a fast, and robust algorithm, that uses the 3D radiative transfer code DART (https://dart.omp.eu/#/) to invert remote sensing images of cities as a 3D distribution of optical properties and temperature, using a geometric urban database. This approach enables us to obtain a 3D model of a city to simulate urban surface albedo, emissivity, and net radiation maps at any date as long as the optical properties of urban elements remain constant. These optical properties can be up-dated with the inversion of recently acquired satellite imagery.As a case study, we will apply it to Brussels to retrieve surface albedo and emissivity maps at a neighbourhood scale. These results can be considered an asset to be used by urban planners and decision-makers to identify what urban areas should be considered priority candidates for an intervention to mitigate heat pollution, which in turn, shall allow authorities or civil organisations to maximise benefits from limited financial resources. Since the SuaBe’s methodology is based on a robust and rigorous physical model, it can be seamlessly implemented in any other city worldwide, provided that a geometric urban database is available.

Abstract. Saharan dust deposits frequently color alpine glaciers orange. Together with other light-absorbing particles, mineral dust reduces snow albedo, increases snow melt rate, and lowers the surface mass balance of glaciers. Since the surface mass balance drives the evolution of alpine glaciers, assessing the impact of impurities helps to understand their current and future evolution. The location of impurities within the snowpack and their effect on snow albedo can be estimated through physical modelling. In this study, we quantified the impact of dust, taking into account mineral dust and black carbon in snow, on the Argentière Glacier over the period 2019–2022. Our results show that during the three years preceding 2022, the contribution of mineral dust to the annual decrease in surface mass balance was between 0.19–0.28 m w.e., while it reached the double in 2022 with 0.47 m w.e. [0.41, 0.50] (median, [Q10–Q90]), and up to 1.00 m w.e. [0.78, 1.12] at specific locations. The impact of dust in snow was unevenly distributed over the glacier, especially in 2022. The highest simulated impacts occurred where firn layers from previous years were exposed after the total melt of the snowpack of the previous winter. The gravitational redistribution of the snow from avalanches was not taken into account, which can reduce the impact of dust at specific locations. Increasing the modelled scavenging efficiency of black carbon can double the impact of dust alone at the glacier scale. In general, the contribution of mineral dust to the melt represents between 6 and 12 % of Argentière Glacier summer melt depending on the year. Hence, we recommend to account for impurities to simulate the distributed surface mass balance of glaciers.

Abstract. Advances in Earth observation capabilities mean that there is now a multitude of spatially resolved data sets available that can support the quantification of water and carbon pools and fluxes at the land surface. However, such quantification ideally requires efficient synergistic exploitation of those data, which in turn requires carbon and water land-surface models with the capability to simultaneously assimilate several such data streams. The present article discusses the requirements for such a model and presents one such model based on the combination of the existing Data Assimilation Linked Ecosystem Carbon (DALEC) land vegetation carbon cycle model with the Biosphere Energy-Transfer HYdrology (BETHY) land-surface and terrestrial vegetation scheme. The resulting D&amp;amp;B model, made available as a community model, is presented together with a comprehensive evaluation for two selected study sites of widely varying climate. We then demonstrate the concept of land-surface modelling aided by data streams that are available from satellite remote sensing. Here we present D&amp;amp;B with four observation operators that translate model-derived variables into measurements available from such data streams, namely fraction of photosynthetically active radiation (FAPAR), solar-induced chlorophyll fluorescence (SIF), vegetation optical depth (VOD) at microwave frequencies and near-surface soil moisture (also available from microwave measurements). As a first step, we evaluate the combined model system using local observations and finally discuss the potential of the system presented for multi-stream data assimilation in the context of Earth observation systems.