Multi-scale and multi-modal benchmarking of 22 global soil moisture products across India's diverse climate zones using in-situ networks and extended triple collocation

The evaluation of satellite- and model-based soil moisture (SM) products is critical for hydrological and agricultural applications, particularly in regions with diverse climatic conditions. This study presents a multi-scale and multi-modal benchmarking of 22 global operational SM products across India’s diverse climate zones using four in-situ observation networks spanning 22 sites. The benchmarking integrates active, passive, active-passive, GNSS reflectometry (GNSS-R), and model-based SM to assess product performance. The cumulative distribution function (CDF) matching approach was employed to address scale and distributional inconsistencies between in-situ and gridded datasets. Results reveal systematically improved performance of most products in arid regions (unbiased Root Mean Square Error, ubRMSE < 0.04 m 3 m -3 ), while increased variability was observed in tropical regimes, highlighting the influence of vegetation and surface heterogeneity. L-band products (SMAP, SMOS, SMOS-IC) demonstrate consistent performance across overpass times, whereas LPRM-AMSR2 shows reduced accuracy in humid tropical conditions (ubRMSE up to 0.1 m 3 m -3 ). Model-based products (ERA5, GLDAS, IMDAA) exhibit strong temporal consistency (R ≥ 0.75) in tropical regions but reduced agreement in semi-arid environments. Extended triple collocation (ETC) analysis was further applied to quantify product-specific uncertainties independent of in-situ errors. ETC-derived correlations range from 0.1 to 1, with error standard deviations up to 0.3 m 3 m -3 , revealing substantial inter-product variability. Among all datasets, GLEAM consistently outperforms others across climatic regimes. Notably, strong agreement between ETC-derived metrics and conventional validation statistics underscores the robustness of the evaluation framework. This study provides a unified benchmarking framework integrating scaling, validation, and uncertainty quantification, offering critical insights into the performance and limitations of global SM products across heterogeneous environments. The findings support improved product selection, data fusion strategies, and the development of next-generation high-resolution SM datasets.