The Investigation Of Warning And Forecasting Methods For Local Region In The Summer Of Huaihe River

Forecasting and early warning of disastrous rainstorm is one of the most difficult problems in weather forecasting. For many years, the forecasting and warning of heavy rain are trying to predict the strength of the heavy rain as soon as possible and to provide the scientific reference and emergency measures to the policy maker.This is especially important for the local residents’ safety of their agricultural products and their even life. However the prediction of the extreme weather is always the most difficult in the weather forecast system. The research of predicting heavy rain usually make efforts to give more accurate warning and forecasting information by using the various modern observational and forecasting methods. Since there are continuous precipitations which could lead to widespread floodings in the summer of Huaihe River. Therefore the forecasting and warning of the heavy rain for local area are urgely needed by the local reduction work as for flood and rainstorm. To meet this end, the methods and technologies of forecasting and warning could provide the reduction work with technical support.Continuous heavy rain is usually the combination of multiple different life periods of meso-small scale strong convection rainstorm systems which are often short term weather process. Doppler weather radar is an important mean to research the mechanism of occurrence and development of these weather systems by using its echo products that provides the dynamic information of the storm system. Systems of surveillance radar nets so far have been completed by State Meteorological Administration in the area where the possibility of storm is high. How to forecasting and warning especially identifying and judging the heavy rain by using the elaborate radar echo data is still the key point.Weather Radar products include reflectivity, radial velocity and spectral width. The reflectivity data is usually used to identify a variety of convective weather systems and to understand the mechanism of these weather systems’occurrence, development3-D structure, movement path, water vapor and precipitation. However, as to the forecasting of heavy rain, it is better to use the Euler method to identify and trace the storm system in the observational range than using Lagrange point which though can identify the evolution characteristics of the storm system. So in this paper, the3-D reflectivity data of weather radar is used to investigate the weather forecast in the next one hour for a local region in Jianhu, so as to warning the heavy rain. Considering the Weather Radar can only support the detection feather of the clouds in the local region, so the probabilistic methods and numerical simulation is also used to investigate the forecast of heavy rain. After the construction of three-dimensional multi-volume scan radar echo composite indicator, the return period of different level precipitation is calculated by using probilistic methods. Then by referring the outcome of the numerical simulation from WRF, supporting the local government with the warning and forecasting information of the heavy rain.Doppler radar volume scan echo data, recording pluviometer data and automatic weather station data is analyzed in this paper. Optimization technique and probability-fitting technique are used to predict quantitative precipitation in Jianhu, Jiangsu. By using optical flow method, the best parameters of different optical flow algorithm were determined. Then the motion vector field of radar echo data in the region deemed as the upstream of the region being predicted is calculated. After calculating the direction of the radar echo’s movement, we took every scan level into consideration and endowed them different weights due to the different effect they affected the precipitation. By linear calculating, we could get a comprehensive factor of radar echo data which can be used to predict the precipitation in the next hour. The result shows that the correlation between the radar echo factors has been raised from54%to71%, which is a considerable improvement regarding the heavy rain forecast. Then the refined radar echo data are classified as two different classifications and the precipitation in the next hour under each radar echo factor are predicted. To inform the possibility of extreme precipitation in the next hour at a local region, two probabilistic methods including P-Ⅲ probability curve and Generalized Pareto distribution are used. Combined with the radar echo factor, these two methods can provide the possibility and intensity of the extreme precipitation in the next hour, thus giving the forecasting and warning information in the next hour for the local region.