Abstract

Horizontal water transport from hills to valleys widely dictates vegetation growth and modulates terrestrial water and energy budgets. However, land cover on hillslopes remains oversimplified in many current land surface models (LSMs) despite its critical role in controlling terrestrial water and energy fluxes. To this end, we focused on Africa to investigate how explicit representation of hillslope water dynamics and land cover conditions in LSM influences terrestrial water and energy budgets. A sub-grid catchment-based strategy was implemented in the LSM MATSIRO, in which the hillslope water dynamics and land cover heterogeneity were modeled in discretized height bands of representative hillslopes. Through several experiments that differed in terms of complexity in representing water and vegetation heterogeneity along hillslope, the impact of representing them in LSM is evaluated and validated. Increasingly accurate representation of the heterogeneities amplified the variabilities of the simulated water and energy budgets, particularly in equatorial Africa. The hill-to-valley water dynamics aside, more realistic representation of the hillslope vegetation distribution aligned with known differences in soil water content (e.g., denser vegetation exists along riversides in arid regions). Such representations changed the soil water contents, runoffs, and evapotranspiration of almost 5%, 7% and 18% of the total African land areas studied, respectively. By better capturing hillslope-scale rainfall interception and transpiration with the improved model, the results are more consistent with observed discharges and leaf area index. Overall, an LSM that considers hillslope water dynamics and land cover heterogeneity substantially modulates and better captures the terrestrial water and energy budgets.

Key points

• Water dynamics and land coverheterogeneity along hillslope areexplicitly considered in the landsurface model MATSIRO
• It reveals more soil water in valleys andgreater (less) evapotranspiration(runoff) on densely vegetated parts ofhillslope
• The appropriate resolution of hillslopevegetation in land surface modelwidely modulates and better capturesthe water and energy budgets