Statistical and neural classifiers in estimating rain rate from weather radar measurements
Abstract. Weather radars are used to measure the electromagnetic radiation backscattered by cloud raindrops. Clouds that backscatter more electromagnetic radiation consist of larger droplets of rain and therefore they produce more rain. The idea is to estimate rain rate by using weather radar as an alternative to rain-gauges measuring rainfall on the ground. In an experiment during two days in June and August 1997 over the Italian-Swiss Alps, data from weather radar and surrounding rain-gauges were collected at the same time. The statistical KNN and the neural SOM classifiers were implemented for the classification task using the radar data as input and the rain-gauge measurements as output. The proposed system managed to identify matching pattern waveforms and the rainfall rate on the ground was estimated based on the radar reflectivities with a satisfactory error rate, outperforming the traditional Z/R relationship. It is anticipated that more data, representing a variety of possible meteorological conditions, will lead to improved results. The results in this work show that an estimation of rain rate based on weather radar measurements treated with statistical and neural classifiers is possible.