Effects of soil-type datasets on regional terrestrial water cycle simulations under different climatic regimes

Abstract

Hydrological simulations play an important role in estimating terrestrial water budgets and monitoring extreme events such as floods. This study investigates how these simulations are affected by soil-type datasets and characterizes how these effects vary with climate. We study the differences between two ensemble simulations in China with the Noah-MP land surface model using two soil datasets from the Food and Agriculture Organization and Beijing Normal University. The differences in ensemble means are analyzed over a 10 year period from 2003 to 2012 with respect to estimated soil moisture, the partition of precipitation between evapotranspiration and runoff, and a flood magnitude index. Results show that the hydrological simulations using sandier soil types result in lower soil moisture, lower evapotranspiration, and higher subsurface runoff. Each of these effects varies uniquely with aridity. The changes in soil moisture decrease with increasing aridity, while the changes in water balance components (evapotranspiration and runoff) peak in the transitional zone between humid and arid regions. The flood magnitude, expressed as the maximum daily flow normalized by annual flow, is also substantially influenced by the input soil type. Soil types with more clay and less sand content yield significantly bigger floods, especially in arid regions.