Wind speed estimation for the correction of wind-caused errors in historical precipitation data

Abstract

The wind has a significant impact on the accuracy of precipitation measurement in the case of collecting gauges. As widely known, the velocity field of wind suffers a deformation over and around the precipitation gauges, which causes deviations in the measured quantities. This error must be corrected if it is possible. Thanks to numerous researches, correction formulas give tools for adjusting precipitation data in the function of the wind speed and raindrop distribution (DSD) relationship, gauge parameters, and for the case of snow and temperature. The measured intensity of precipitation in historical data allows estimating the DSD, but in most cases, there are no simultaneously measured wind speed data coupled to the historical precipitation data. Characteristic data of wind speed can be estimated based on the wind speed statistics, and these data can be utilized for the statistical correction of the precipitation measurements. The statistical correction means that the rainfall data can be adjusted with the expected value of the wind speed for a more extended observation period, assuming a stationarity of wind speed statistics for the given location. After the statistical correction, the unique data will not be unbiased, but statistically they will be closer to the actual value, and the correction will be statistically correct in inherited perecipitation cheracteristics, as for example the IDF curves. For this correction, an investigation is necessary to find the adequate wind statistics for the rainfall correction. This paper shows the results of a study about the relation of statistics of wind speeds during precipitation, based on a 10-minute sampling period. The wind speed data were independent of the rain depth (or intensity) data. The result of the study shows that the distribution of wind speeds differs of the wind speed distribution measured in the precipitation events. This difference can be treated easily using the stable rate of the means of these distributions. This result gives a step toward correcting the wind-affected error of historical precipitation data.