Estimating seasonal variations in snow water equivalent in High Mountain Asia using differences in precipitation retrieved by top-down and bottom-up methods

Study region High Mountain Asia (HMA) Study focus Snow water equivalent (SWE) is critical in understanding snowpack and melt processes in mountainous regions. However, the high spatial heterogeneity of snow depth and density in these regions introduces significant uncertainty in current SWE estimations. This study proposes a statistical method based on precipitation differences (DM) between top-down and bottom-up retrieval approaches to estimate SWE dynamics across High Mountain Asia (HMA), with its accuracy assessed against existing SWE products and in situ observations. New hydrogeological insights from the region Over the HMA domain, DM exhibits low combined accuracy errors (CA; mean/median of 1.10/0.87 relative to GLDAS and 1.10/1.53 relative to ERA5) and moderate overall correlations (CC of 0.35 with ERA5 and 0.26 with GLDAS), with substantially higher agreement in high-dynamic regions (CC up to 0.8). The temporal uncertainty of DM is generally within ±5 days during snow accumulation periods (Delay Index; DI ∼ 0–5), indicating its ability to track snow evolution and assess potential water storage for runoff forecasting. Station-based evaluations further confirm that DM reliably reproduces major accumulation and ablation events, with consistently high CC and DI values comparable to ERA5. This study provides a practical pathway for characterizing water storage and seasonal water availability in data-scarce mountain environments, improving regional hydrological assessment in HMA.