Correction of precipitation based on off-site weather information

Owing to aerodynamic errors, correction of precipitation measurements, liquid, mixed or solid, is often carried out by means of an explicit mathematical statistical model. The magnitude of correction, e.g., calculated as a correction factor to the observed amount of precipitation, is the dependent variable, and wind speed, temperature and a measure of precipitation intensity are independent variables entering the correction model. Usually these independent variables are accessible on site where the precipitation measurements are recorded. However, in many standard precipitation networks the gauge measurements are not accompanied by on site information of these variables, and information from remote stations must be used. The aim of the paper is to describe how the statistical models can be used in evaluating if the set of remote information of the independent variables can be used for the estimation of a reliable correction factor on site. The technique is then applied to an example of precipitation network in Denmark where the typical situation for the standard gauges is that only the amount of precipitation is available. It is expected that not only can the methods be adapted to other countries, but for landscapes similar to Denmark even the conclusions can be applied directly: extrapolation of all independent variables from remote sites should be conducted with caution, wind speed information can be extrapolated from remote sites not farther away than approximately 50 km, while information on rain intensity and temperature can be safely extrapolated across longer distances. Published by Elsevier Science B.V.