Research On Characteristics Of The Heavy Rainfalls In South China With C-band Weather Radar

By using the data of Doppler weather radar from Guangzhou Baiyun Airport and hourly precipitation data from automatic weather stations, we have a study on the spatial and temporal distribution from2009to2011in the first flood season in south China. Combining the radar echo data and radial wind data to analyze the three-dimensional structure and the temporal and spatial features of different rainfall pattern and provide a short-term precipitation forecast. Selecting three typical process of heavy rainfall during the first flood season in South China, with the data under the quality control, we estimate precipitation with the fixed Z-I Relations and the optimal Z-I Relations. Through the analysis, in the course of radar echo with strong distribution and rainfall distribution has a very good relationship, forecasting the radar echo for the next moment of radar echo data. And the precipitation are estimate by fixed Z-I relationship and the optimization method. The result shows that the two methods can reflect the rainfall distribution on the whole. But there are some errors in the estimation of precipitation and the optimization method are better than the fixed Z-I relationship. Here are some some conclusions:(1) With the process of remove the super refraction echo, isolated echo and median filter, the characteristics of original radar data are retained and noise are largely filtered. Using mixed scanning method the problem of the low elevation data have much noise and the high elevation data are limited by scan range are solved.(2) By analyzing11times rain with the circulation from2009to2011before flood season rainstorm process we have some conclusions.500hPa field situation can be roughly divided into four types such as two ridge one trough type, two trough and one ridge, a trough of a ridge and multi wave type.850hPa field mainly affects the system performance for the cyclone and shear line. Some precipitation process are affected by the interaction effect of two kinds of system, and not all of the strong precipitation process are associated with low-level jet. The location, direction of jet and intensity of the rainfall process is different by different process. Found on the radar echo and the statistics of the radial wind field that the radar echo characteristics of the eleven precipitation process can be divided into mixed precipitation echo in Stratiform Cloud Precipitation dominated, mixed echo in cumulus cloud precipitation, squall line system and super single thunderstorm. The four kinds of strong precipitation echo characteristics are not the same.(3) Through the statistics of radar radial wind field analysis we can see that the radial wind characteristics of radar can be used to illustrate the actual wind field changes with height. The characteristics are compared with the weather situation is depicted in detail, so we can use radar radial wind field to more clearly identify precipitation leading system and weather situation.(4) By analyzing the three typical rainfall process in Southern China in May,2010, we found that the circulation are not the same. The high latitude circulation respectively for the performance of the two troughs and one ridge, multi wave type and two ridge one trough the circulation situation. Rain belt location with shear line is closely related to the radar echo. The distribution of rain precipitation process and echo center of radar data are in good correlation.(5) The moving direction and speed of radar echo can be predicted by the radar echo of the next moment values in two adjacent time echo value. Three typical precipitation process in the radar is different, the position and distribution analysis of rain echo are not the same. Forecast the radar echo location of the three strong precipitation process by using extrapolation method, correlation coefficient of the strong precipitation center belt distribution prediction and the actual situation in precipitation the66.93%,58.83%and75.96%.(6) Combined with the Z-I relation to inversion of precipitation, we compare the estimation of radar rainfall distribution and precipitation distribution. The distribution during the heavy rainfall belt and the precipitation is very close, but in the scope and intensity of rainfall have a certain degree of underestimation. Comparing the optimal Z-I method for estimating rainfall intensity obtained than fixed Z-I relation has been greatly improved. But there are echoes for different types of the same precipitation process, using the same A, b value of precipitation estimation will still have a great effect on the estimation results.