Conditioning point and gridded weather data under aridity conditions for calculation of reference evapotranspiration

Abstract

Meteorological data are often collected in dry, desert regions where the local environments exhibit effects of aridity caused by the lack of evapotranspiration (ET). In contrast, the standard reference ET (ETo) calculations of FAO56 and ASCE assume that the surfaces underlying collected weather data are well-watered so that near-surface meteorological measurements reflect the cooling and humidifying effects of an evaporating surface. In this study, we develop a weather data conditioning process and algorithms to adjust for biases in meteorological data that exhibit aridity effects. The conditioning process is intended to adjust the weather data to better exhibit characteristics of data collected over a well-watered vegetated surface prior to use of the data to estimate reference ET. The procedure involves the extrapolation of air temperature, vapor pressure, and wind speed profiles to and from a regional blending height using standard surface energy balance equations and flux-profile relationships and employing ET estimated for the ambient, dry conditions of an arid weather station and reference ET that should co-exist with weather measurements used to calculation reference ET. Example applications are given with 24-hour weather data and hourly weather data in Idaho and Nevada. Results indicate that reference ETo can be overstated by as much as 25% in southern Idaho and 8% in eastern Nevada. The methodology is intended to be transferrable to other regions and climates and is self-aware of the need for conditioning of weather data according to differences in ambient ET and the reference ET estimates.