Application Research Of Doppler Weather Radar Data Analysis And Assimilation In Mesoscale Synoptic System Numerical Simulation

Meiyu Front Heavy Rain Processes are important events which happen in late spring and early summer at the area of Yangdzi and Huaihe river. Recently many research demonstrate that these rainstorms are trigged by mesoscale systems which are combined by many cumulus-scale cloud subsystems. The mesoscale systems develop under meso-a or larger scale weather systems and may produce long-lasting precipitations if suitable conditions keep steady.Based on summarizing the research work on Meiyu rainstorms and mesoscale synoptic systems recently, this paper presents a method to research Meiyu rainstorms in valley of Yangdzi and Huaihe rivers with 4D variant assimilating technique by inputting estimate precipitation field which is received from CINRAD network and rain gauge network into numeric model MM5. Some results have been reached by comparison two Meiyu rainstorm cases of 2003.6.26 12:00-2003.6.27 12:00(UTC) and 2003.7.8 12:00-2003.7.9 12:00.(1) It is found that the severe storms are mainly happened in meso-p scale cyclone on Meiyu front and local convective precipitation are relative to meso-γ cyclone based on the analysis of observations and model results that use conventional weather station data as the input of numeric model MM5 from these two cases.(2) It is the first time to estimate large scale precipitation by use of multi CINRAD network radars and all available and proven radar-raingauge techniques. The precipitation field derived from radarnetwork can describe not only the structure of large scale synoptic system, but the features of mesoscale weather system with higher temporal and spacial resolution.(3) Precipitation estimation based on radar raingauge techniques can get rainfall field with relatively steady error especially when aggregated different results from such as average adjustment; combined Kalman filtering and optimum interpolation and combined kalman filtering and calculus of variation techniques. This could be applied to 4DVAR assimilation so as to reconstruct reliable and steady precipitation data in higher temporal and spacial resolution. Evaluation from the above two cases demonstrate that among above three methods of radar measment precipitation techniques, both combined kalman filtering and optimum interpolation and combined kalman filtering and calculus of variation are better than average adjustment, combined kalman filtering and optimum interpolation is further better than combined kalman filtering and calculus of variation.(4) Wind profile can be retrieved from single Doppler radar under the linear field assumed, which is basically quasi-geostrophic that shows the feature of large scale motion. The wind profile data from multi Doppler radars are then interpolated into model grids hourly by use of MM5 objective software LITTLE_R to adjust the whole wind field. Results show that wind profiles derived from weather radar have high resolution in time and in vertical direction and are able to assimilate meso-scale model from local feature. This may change the structure of atmospheric stability and cause the distribution of relative factors more reliable. The wind profile derived from CINRAD radar shows the property of wind changing in boundary layer even the low level southwest jet and plays an important role in simulating vapor transmission in meso-p cyclone after assimilating it into the model. This promoted the ability of simulating mesoscale weathersystem in boundary layer especially for southwest jet. On the other hand, the numeric model simulates the storm more clear after using radar wind profile data and the gravity wave is excitated which is one of the most important factor for severe storm forming.(5) This paper focus on the 4DVAR used high temporal and spacial resolution data of radar quantitative estimate of rainfall. Results are as follow based on the two cases of Meiyu front storm:4DVAR assimilation produces a model harmonious initial precipitation in physics from radar estimated precipitation, increases initial cloud water and forms precipitation at early steps that...