The impact of standard and hard‐coded parameters on the hydrologic fluxes in the Noah‐MP land surface model

Summary

This study employed Sobol' global sensitivity analysis to quantify the influence of 117 parameters (42 standard and 75 hard-coded) in the Noah-MP land surface model on hydrologic fluxes across diverse United States catchments. The analysis identified that soil surface resistance for direct evaporation—a hard-coded value—is the most sensitive parameter, and that calibration strategies must account for both plant and soil parameters to derive realistic model predictions. The findings have implications for land surface model calibration methodology, suggesting that including sensitive hard-coded parameters in calibration frameworks would improve model agility and output reliability.

UK applicability

The methodological approach and sensitivity analysis framework could be applied to calibrate land surface models for UK hydrology, though UK catchments have different precipitation and temperature regimes than the studied United States sites. The findings regarding parameter coupling (latent heat and runoff) may be transferable to temperate maritime climates, but site-specific sensitivity assessments would be needed.

Key measures

Sobol' global sensitivity indices for latent heat, total runoff, surface runoff, and component flux outputs; parameter sensitivity threshold set at 1% Sobol' index

Outcomes reported

The study identified 139 hard-coded parameters in the Noah-MP land surface model and evaluated their sensitivity to hydrologic output fluxes (latent heat and total runoff) across 12 catchments with contrasting hydrometeorological regimes. Global sensitivity analysis revealed that two-thirds of standard parameters and many hard-coded parameters significantly influence model outputs.

Topic tags