Seasonal based-global sensitivity analysis of SAC-SMA model arameters

Abstract

This study implemented Sobol-based global sensitivity analysis to assess parameter importance in the Sacramento Soil Moisture Accounting (SAC-SMA) model under contrasting regimes: Flood and dry seasons. The Mu Cang Chai watershed in Nam Mu basin is used as a case study. Results show clear seasonal contrasts in dominant processes and parameter interactions. In the flood season, runoff is strongly governed by lower-zone storages, with LZFSM, LZFPM, and LZTWM presenting the greatest influence. Interactions involving UZFWM further highlight the importance of upper-to-lower zone linkages in shaping flood flows. Calibration for wet conditions should therefore prioritize these storages, while handling percolation and drainage parameters as interaction-driven controls. In the dry season, UZFWM, LZTWM, and UZTWM are dominant parameters, while interaction effects involving impervious areas (PCTIM) also indicate that small surface conditions can modulate drought runoff. These seasonal insights support more efficient calibration to improve SAC-SMA model interpretability.