Refining GRACE TWS Data for Enhanced Water Management in Semi-Arid Regions

Satellite-based hydrology, particularly leveraging the GRACE and GRACE-FO missions, has revolutionized our understanding of terrestrial water storage (TWS) at global scales. These missions enable the monitoring of water resources and the detection of climate-driven hydrological changes, which are crucial in regions facing increasing water stress. However, GRACE data’s coarse spatial resolution hampers its integration with other fine resolution ancillary datasets to disaggregate individual hydrological components present significant challenges, especially for semi-arid regions where localized water management is paramount. Currently, there is growing attention to the need for higher-resolution datasets that can support granular decompositions of TWS in order to inform actionable policy and adaptive water usage strategies. Recent studies highlight substantial progress in remote sensing and the integration of ancillary datasets, yet the debate continues regarding the best approaches to overcome the intrinsic limitations of GRACE data.Recent advances in machine learning offer promising avenues to address these knowledge gaps. By integrating GRACE TWS data with climate reanalysis outputs and remote sensing products—such as soil moisture, NDVI, and land surface temperature—researchers have begun applying sophisticated models, including artificial neural networks and ensemble learning methods, to improve spatial granularity. Comparisons with in situ groundwater data and the application of robust statistical validation techniques have further strengthened confidence in these new methodologies. Nonetheless, the field is still evolving, and significant questions remain on optimizing model selection, predictor variables, and spatial transferability for different hydroclimatic settings.This Research Topic aims to accelerate the development and assessment of advanced computational methods for refining GRACE TWS data, with a special focus on semi-arid environments like the Mediterranean basin. The main objectives are to: (1) enhance spatial and/or temporal resolution of GRACE TWS datasets, (2) improve attribution of water storage changes to distinct hydrological components, and (3) support data-driven strategies for water management and climate adaptation at the regional scale. Key questions include the identification of suitable machine learning frameworks, the evaluation of various predictor datasets, the verification of outputs against ground observations, and the generalization of methods across diverse semi-arid settings.The scope of this Research Topic is defined by research that advances the disaggregation of TWS data and its application to regional water management and climate vulnerability studies in semi-arid landscapes. Contributions are limited to approaches that integrate GRACE or GRACE-FO data with complementary remote sensing or ground-based datasets and focus on analytical, modelling, or validation innovations. To gather further insights in this area, we welcome articles addressing, but not limited to, the following themes:• Machine learning algorithms and statistical models for GRACE TWS downscaling and component separation• Integration of multi-source remote sensing data (e.g., soil moisture, NDVI, LST) for enhanced water storage estimation• Validation of model outputs against in situ groundwater and hydrological observations• Comparative studies of spatial and temporal disaggregation techniques• Application of refined water storage products for policy, management, and climate adaptation planning• Uncertainty assessment and sensitivity analysis of disaggregation methods• Case studies on Mediterranean and other global semi-arid environments Topic editors: Alexandra Gemitzi, Muhammad Zohaib Source: https://www.frontiersin.org/research-topics/77432/refining-grace-tws-data-for-enhanced-water-management-in-semi-arid-regions