Assimilation of Doppler Weather Radar Data in Numerical Weather Prediction Models

Abstract

The main aim of the thesis is: newline Assimilation of Doppler Weather Radar Data in Numerical Weather Prediction Models newlineThe Problem discussed in thesis is as follows: newline To improve weather analysis and forecast for nowcast and very short range by assimilation of Indian Doppler Weather Radar Data in NWP Models newline newlineAssimilation of Doppler weather radars in numerical weather prediction (NWP) models is very important for accurate real time prediction of cyclones and convective activities. Quality control radial velocity and reflectivity data modify the dynamical and thermo dynamical structure of weather systems in numerical model and hence improve the analysis and forecast. newline newlineIn this thesis, Numerical models Advanced Regional Prediction System (ARPS) and Weather Research and Forecast (WRF) are used as forecast system for simulation and real time prediction of extreme weather such as cyclones, thunderstorms, squall line and cloud burst. In both the models radial velocity is assimilated using either ARPS Data Analysis System (ADAS) or a three dimensional variational (3DVAR) newlinemethod and reflectivity data is assimilated using cloud analysis technique. The ADAS uses a successive correction scheme, known as Bratseth analysis method and includes radial velocity adjustment. ARPS 3DVAR uses multiple analysis passes that can be used to analyze different data types with different filter scales to account for the variations in the observation spacing among different data sources. Reflectivity data is used to adjust moisture and temperature in precipitation regions. newline newlineIn general, assimilation of DWR data for cyclone cases remarkably improves the hydrometeor structures of the initial vortex, divergence/convergence conditions over the cyclone s inner-core area and precipitation pattern. Cyclone track errors are significantly reduced. However, impact is restricted up to maximum 12 hours Perhaps it arises due to the dominance of the large-scale characteristics over the local features present in the initial fields (GFS) with the progress of...