| Precipitation is one of the most important meteorological elements. Accurate and quantificational measurement of precipitation is of great significance for disaster prevention and mitigation and the precipitation forecast. And rain gauges are a little far from each other, so they can’t reflect the spatial distribution of precipitation accurately. However, radar has features of short time, wide range, high resolution, etc., therefore, radar quantificationally estimating precipitation is gradually becoming an important method for accurate measurement of precipitation. In this article, taking Taiyuan City as research area, using the data of C Band of Doppler weather radar in the city and precipitation in 1 hour from weather stations in Taiyuan and surrounding cities, and basing on two kinds of factors, the distances form weather stations to radar centers and the values of radar echo which are corresponding to the weather stations, establish three kinds of models of radar quantificationally estimating precipitation of self-adaptive relation of Z-I to estimate the inversion of the surface precipitation in Taiyuan. Meanwhile, use a set of fixed model of the relation of Z-I which is fitted by all fitting stations (Fixed model) and the new radar model (Z=300I1.4) to invert the precipitation in Taiyuan during the same rainfall process. At last, compare and analysis the inversion results from five models. The major work includes:1. In this article, use the relations of Z-I of every gird from the self-adaptive optimized algorithm, and estimate the surface precipitation in Taiyuan, to get the spatial distribution of hourly precipitation of two rainfall processes in Taiyuan;2. Compare the results of surface precipitation in research area from three models of self-adaptive relation of Z-I which are established in the article, and analysis the accuracies of three models.3. Compare and analysis the pros and cons among the models which are self-adaptive relation of Z-I model, the new radar relation Z=300I1.4, and the fixed modelThrough the above work of fitting the surface precipitation and comparison and analysis of each model, we can conclude that:1. The precipitation of research area inverted by the optimized model of self-adaptive relation of Z-I is closer to the observed value of rain gauge, and the spatial and temporal distribution of precipitation is similar to the distribution of surface precipitation interpolated by the rainfall values measured by rain gauge. Meanwhile, the accuracy of the radar quantificationally estimating surface precipitation in Taiyuan is much improved than directly using the new radar model (Z=300I1.4).2. It is of great difference for the accuracy of the results of surface precipitation in research area estimated by the models which are three optimized models of self-adaptive relation of Z-I, the fixed model and the new radar model. The effects of three optimized models of self-adaptive relation of Z-I are best, followed by the fixed model. And the estimated errors of these four models are far less than the error of the model of directly using the new radar relation Z=300I1.4. Therefore, for one research area, if use the proper relation of Z-I which is adapted to local situation to measure precipitation quantificationally, we can effectively improve the accuracy, which is of great importance in the rain forecasting and disaster prevention and mitigation.3. Among three models of self-adaptive relation of Z-I, the accuracies of surface precipitation estimated by the distance model and the echo model are slightly higher than the accuracy of comprehensive model, and the distance model and the echo model have better space-time stability as well, which shows that it has better accuracy and applicability for using these two models of self-adaptive relation of Z-I to estimate the surface precipitation. Using self-adaptive relation of Z-I, we can implement automation and accuracy of precipitation monitoring business much more easily. |
PDF Full Text Request