Improving Bowen-ratio estimates of evaporation using a rejection criterion and multiple-point statistics

The application of the Bowen ratio method to estimate evaporation is heavily affected by uncertainties on the measured quantities. Time series collected with a hydro-meteorological monitoring station often contain measurements for which a reliable estimate of evaporation cannot be computed. Such measurements can be identified with standard error propagation methods. However, simply discarding some values might introduce a bias in the cumulative evaporation for long time intervals, also depending on the threshold of acceptance. In this paper, we propose the use of multiple-point statistics simulation to integrate the time series of reliable evaporation estimates. A test conducted on a two-year-long time series of data collected with a hydro-meteorological station in the Po plain (Italy) shows that the usage of a rejection criteria in conjunction with multiple-point statistics simulation is a promising and useful tool for the reconstruction of reliable evaporation time series. In particular, it is shown that if the rejected values are not replaced by simulation, then the cumulative evaporation curves are estimated with a bias comparable with estimates of cumulative annual evaporation. Moreover, the test gives some insights for the selection of the best rejection threshold.