Thèse Caractérisation des Déterminants de l'Hétérogénéité Spatiale de l'Humidité du Sol et Développement d'Un Indice de Représentativité de Sites Instrumentés Européens à Partir d'Une Évaluation M H/F

Établissement : Université de Toulouse École doctorale : SDU2E - Sciences de l'Univers, de l'Environnement et de l'Espace Laboratoire de recherche : CESBIO - Centre d'Etudes Spatiales de la BIOsphère Direction de la thèse : Arnaud MIALON ORCID 0000000179700701 Début de la thèse : 2027-09-01 Date limite de candidature : 2026-11-23T23:59:59 Dans le contexte du changement climatique, cette thèse s'attachera à améliorer le suivi de l'humidité du sol et les objectifs principaux de la thèse sont les suivants :

- Élaborer et valider des procédures d'étalonnage des mesures de l'humidité du sol spécifique à chaque site sur différents écosystèmes du réseau européen ICOS.
- A partir de campagnes expérimentales multi-capteurs sur différents écosystèmes, identifier les facteurs, universels et spécifiques à chaque site, qui influencent la représentativité de l'humidité du sol et les échelles caractéristiques associées.
- Mieux comprendre dans quelle mesure la représentativité spatiale des observations de l'humidité du sol peut s'expliquer par les caractéristiques des écosystèmes mesurées sur les sites ICOS ?
- Développer un indice générique de représentativité spatiale de l'humidité du sol transposable à tous sites instrumentés des réseaux de mesures internationaux (ICOS, ISMN, FluxNET). Cet indice fournira une estimation de la qualité des données des réseaux instrumentés pour l'étalonnage et la validation du pixel satellite de l'humidité du sol.

Soil moisture is a key variable for monitoring ecosystems under increasing pressure from climate change. The striking example of summer 2026, marked by successive heatwaves and severe drought across France and parts of Europe, highlights how interactions between soil drought and atmospheric conditions can amplify the intensity and persistence of heatwaves, while exacerbating their impacts on water resources, ecosystem vulnerability to wildfire, and agricultural production.
Satellite observations enable large-scale monitoring of soil moisture with regular revisit times. They can identify areas at risk of wildfire and drought and can play an important role in water-management policies. However, improving the accuracy of satellite-derived soil moisture products requires local reference data for their calibration and validation. With the establishment of ICOS ERIC (Integrated Carbon Observation System), which is an European research infrastructure monitoring greenhouse gases across ecosystems throughout Europe, standardized protocols and high-precision instruments have been implemented at labelled sites to monitor a wide range of biogeophysical, gaseous and energy fluxes, and climatic variables, including soil water content (Op de Beeck et al., 2018). Similarly, the International Soil Moisture Network (ISMN) is a collaborative initiative that collects and harmonizes a large database of in situ soil moisture observations for validating remote-sensing products. However, systematic site-specific soil moisture calibration is not routinely implemented, leading to substantial uncertainties in the calibration/validation (Cal/Val) of satellite soil moisture products.
One of the scientific challenges identified by ESA's FRM4SM project (Fiducial Reference Measurements for Soil Moisture) concerns the spatial representativeness of point-scale soil moisture measurements and scaling issues between local observations and satellite products (Gibon et al., 2024). Addressing this challenge first requires a better understanding of the drivers of soil water content scaling, including topography, soil properties, hydrology, and ecosystem functioning.
Therefore, prior to the proposed PhD project, several multi-sensor measurement campaigns are planned for 2027 at French and Swedish ICOS sites, combining systematic site-specific soil calibration with different sampling strategies to optimize the spatial and temporal resolution of soil moisture monitoring.
The objectives of the proposed PhD project are to:
1) quantify the extent to which the relationship between local soil moisture measurements and their spatial variability can be generalized, and determine at which spatial scales the generalizations remain valid;
2) develop and evaluate objective indicators of the spatial representativeness of instrumented sites;
3) integrate these indicators into the reference protocols promoted by FRM4SM and into international networks such as ISMN, FLUXNET and ICOS.

-Develop and validate site-specific calibration procedures for soil moisture measurements across contrasting ecosystems within the European ICOS network.
-Identify universal and site-specific factors influencing soil moisture representativeness and the associated characteristic spatial scales, using multi-sensor experimental campaigns across contrasting ecosystems
- Quantify the extent to which the spatial representativeness of soil moisture observations can be explained by ecosystem characteristics measured across ICOS sites.
- Develop a transferable spatial representativeness index for soil moisture that can be applied across instrumented sites within international observation networks such as ICOS, ISMN, FLUXNET). This index will provide a quantitative assessment of the data quality of instrumented networks for the calibration and validation of satellite soil moisture pixels.
The PhD project will begin with the analysis of long-term soil water content time series from sites in France, Sweden, and Germany, primarily within the ICOS network but also from the ISMN, in order to encompass contrasting climates, soils, and ecosystem types. These data will be used to investigate relationships between soil moisture and site-specific characteristics (soil texture, bulk density, rooting depth, available water capacity, slope and aspect) and/or environmental factors (vegetation type and climate).
Next, multiresolution approaches, including spectral, wavelet, and cross-wavelet analyses, will be used to identify common spatial structures and characteristic scales of soil moisture heterogeneity across sites. For example, spatial soil moisture measurements collected along transects or within a map grid could be analysed using these mathematical tools to detect the dominant scales of spatial variability and characterize how these structures change across the landscape. Cross-wavelet analysis could further be used to compare soil moisture patterns with other environmental variables or across sites, helping to identify common spatial structures and the factors controlling soil moisture heterogeneity. Eventually, the study of spatial coherence could be completed with SVAT (Soil Vegetation Atmosphere Transfer) model simulations driven by high-resolution satellite observations, by comparing these simulations with soil moisture patterns derived from distributed in-situ sensors. The SVAT simulations could also help to test and infer relevant indicators for soil moisture heterogeneity.