@COMMENT This file was generated by bib2html_grace.pl <https://sourceforge.net/projects/bib2html/> version 0.94
@COMMENT written by Patrick Riley <https://sourceforge.net/users/patstg/>
@COMMENT This file was prepared using the NASA Astrophysics Data System (ADS)
@COMMENT https://ui.adsabs.harvard.edu/
@ARTICLE{2026ESE....10.6829A,
       author = {{Ali}, Shoaib and {Ran}, Jiangjun and {Tangdamrongsub}, Natthachet and {Khorrami}, Behnam and {Ferreira}, Vagner and {Shi}, Haiyun and {Zhang}, Wenmin},
        title = "{Weight-Supported Random Forest Downscaled GRACE (-FO) Data Uncovers Groundwater Depletion Linked to Winter Wheat Cultivation in the North China Plain}",
      journal = {Earth Systems and Environment},
     keywords = {North china plain, GRACE (-FO), GWSA, RF$_{SW}$, Downscaling, Winter wheat, Earth Sciences, Physical Geography and Environmental Geoscience},
         year = 2026,
        month = oct,
       volume = {10},
       number = {6},
        pages = {6829-6852},
     abstract = "{Groundwater is a critical resource for sustainable development,
        particularly in arid regions facing water scarcity. The Gravity
        Recovery and Climate Experiment (GRACE) and its Follow-On,
        GRACE-FO, offer valuable data on groundwater storage anomalies
        (GWSA). However, while their coarse resolution has been improved
        using machine learning approaches such as the global random
        forest (RF$_{G}$) model, the aspatial nature of the RF$_{G}$
        model limits its ability to capture spatial heterogeneity when
        downscaling GRACE (-FO) data. Downscaling GWSA data to higher
        resolutions is crucial for assessing small-scale groundwater
        variations. To address this, a novel spatially weighted random
        forest (RF$_{SW}$) model has been proposed to downscale GWSA to
        a high resolution (0.1{\textdegree}) across the North China
        Plain (NCP) from 2003 to 2023. We found that the RF$_{SW}$ model
        outperforms the RF$_{G}$ model, reducing RMSE by 44.44\% and
        residuals by 43.57\%. The downscaled GWSA data strongly
        correlate with in-situ measurements from 559 monitoring wells
        (correlation coefficients: 0.52--0.85), revealing significant
        groundwater depletion in the Piedmont Plain (PP) and East-
        Central Plain (ECP) sub-regions, with the most severe losses in
        Shijiazhuang (17.08), Xingtai (16.67), and Handan (16.02 mm/yr),
        respectively. The winter wheat area doubling from 2.5 million to
        5.8 million hectares, reducing GWSA from 180 mm to 480 mm. This
        improved downscaling technique enhances our understanding of
        local groundwater dynamics and their relationship to
        agricultural practices. This method's high-resolution GWSA data
        can inform more targeted and effective water management
        strategies in water-stressed regions worldwide. The figure
        presents a detailed graphical representation of the methodology
        employed to downscale GRACE (-FO) derived GWSA data for
        evaluating groundwater storage distribution. The analysis
        commences with the GRACE (-FO) TWSA dataset, which exhibits gaps
        in monthly data. The Seasonal-Trend decomposition based on the
        Loess (STL) method is employed to reconstruct the missing
        values, resulting in a continuous TWSA dataset. The estimated
        continuous TWSA calculates GWSA by deducting SMSA and SWEA from
        TWSA. An RF$_{sw}$ model is utilized to downscale GWSA to a
        higher resolution of 0.1{\textdegree} by incorporating
        explanatory climatic (rainfall, LST, airT, ET) and hydrological
        (SMS, SWE, Qs, Qsb, Snowcov), along with additional NDVI. The
        validation of the downscaled GWSA against in situ data reveals a
        high correlation coefficient (CC = 0.85) and similar trends.
        Landsat-5/7/8 and Sentinel-2 multispectral imagery are utilized
        to map winter wheat, demonstrating the spatial distribution of
        wheat cultivation across the NCP region. The resulting
        distribution maps reveal significant spatial expansion of winter
        wheat, particularly in the groundwater-dependent areas. This
        workflow facilitates a more accurate comprehension of
        groundwater storage estimation and the downscaling of GWSA.}",
          doi = {10.1007/s41748-025-00976-6},
       adsurl = {https://ui.adsabs.harvard.edu/abs/2026ESE....10.6829A},
      adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}