Evaluation of soil moisture forecasts for landslide-prone areas in the Western Himalayas: A case study

This study evaluates the skill of the Joint UK Land Environment Simulator (JULES) for forecasting soil moisture relevant to landslide early warning in the Western Himalayan region. During 13–14 August 2023, Himachal Pradesh experienced extreme monsoon rainfall associated with a western disturbance and moisture incursion from the Arabian Sea, triggering multiple landslides in Shimla, Solan, and Mandi. JULES soil moisture forecasts were analyzed for four soil layers (0–10 cm, 10–35 cm, 35–100 cm, and 100–300 cm) using hindcasts from day − 5 to day − 1. The model captured a pronounced increase in soil moisture coincident with the event, with the day − 3 forecast showing a notable rise in near-surface soil moisture (0.40–0.42 m³ m⁻³), reaching at a threshold value of soil moisture > 0.4 m³ m⁻³, indicating potential early warning capability. Pearson correlation between rainfall and day − 3 top-layer soil moisture was 0.94 for Mandi, 0.78 for Shimla, and 0.60 for Solan are used to show hydrological linkage. The results demonstrate a strong coupling between forecasted soil moisture and rainfall prior to landslide occurrence, highlighting the utility of land surface model forecasts for landslide preparedness in data-scarce mountainous regions.