Estimating surface moisture availability for evaporation on bare soil from routine meteorological data and its parameterization without soil moisture

Abstract

Surface moisture availability is the most important parameter for estimating evaporation from bare soil using the bulk transfer method. In this study, we developed a simple method for the estimation and parameterization of surface moisture availability over a period ranging from 10 days to several months on bare soil using only routine meteorological data without using sensible and latent heat flux and soil moisture data. If the daily mean soil temperature is observed with other routine meteorological data of daily values (air temperature, solar radiation, precipitation, wind speed, humidity, etc.), surface moisture availability can be estimated by using a simple soil temperature model without using the sensible and latent heat flux observation data. The simple soil temperature model is composed of the heat balance equation at the ground surface, bulk transfer method and modified version of the force-restore model for estimating daily mean soil temperature. The estimation period of surface moisture availability ranges from 10 days to several months according to precipitation conditions that reflect soil wetness. The antecedent precipitation index (API) is an effective indicator by which the soil wetness condition can be judged without using soil moisture. This enables classification of proper estimation periods of surface moisture availability. The surface moisture availability is estimated by minimizing the sum of the squares of the residual between the observed value of the daily mean soil temperature and the calculated value under the estimation period as classified by API. Furthermore, for the purpose of estimating daily mean soil temperature and cumulative evaporation on bare soil using only routine meteorological data, we propose that the surface moisture availability can be parameterized using the ratio of precipitation to potential evaporation with estimation periods ranging from 10 days to several months.