DATA ERROR EFFECTS ON NET-RADIATION AND EVAPOTRANSPIRATION ESTIMATION

The objective of this paper is to evaluate the potential error in esti mating the net radiation and reference evapotranspiration resulting fr om errors in the measurement or estimation of weather parameters. A me thodology for estimating the net radiation using hourly weather variab les measured at a typical agrometeorological station (e.g., solar radi ation, temperature and relative humidity) is presented. Then the error propagation analysis is made for net radiation and for reference evap otranspiration. Data from the Raimat weather station. which is located in the Catalonia region of Spain, are used to illustrate the error re lationships. The results show that temperature, relative humidity and cloud cover errors have little effect on the net radiation or referenc e evapotranspiration. A 5 degrees C error in estimating surface temper ature leads to errors as big as 30 W m(-2) at high temperature. A 4% s olar radiation (R-s) error can cause a net radiation error as big as 2 6 W m(-2) when R-s approximate to 1000 W m(-2) However, the error is l ess when cloud cover is calculated as a function of the solar radiatio n. The absolute error in reference evapotranspiration (ETo) equals the product of the net radiation error and the radiation term weighting f actor [W = Delta(Delta 1 + gamma)] in the ETo equation. Therefore, the ETo error varies between 65 and 85% of the R-n error as air temperatu re increases from about 20 degrees to 40 degrees C. (C) 1998 Elsevier Science B.V. All rights reserved.