Blending Extrapolation and R3R Forecast

It is well known that convective weather can significantly affect aviation safety. The reliable and successful forecast of convection can help air traffic controllers to manage airport, plan air routes, avoid potential flight delay or even safety problems, therefor reduce possible losses and greatly improve economic efficiency and airlines service quality. Nowcasting is a short-term forecast and simply extrapolates remote sensing observations. It is a relatively reliable forecast of convection for aviation industry. However, the performance of the nowcasting based on extrapolation will rapidly decrease with the increase of forecast leading time because this method only consider advective transport and cannot take into account the time evolution of the convection. It is expected that the numerical model solidly based on physics and mathematics, taking both advection and time evolution of convection account, will finally solve the problems in all. However, at present, the skill of numerical model forecasting convection underperforms the skill of nowcasting within two hours, although it could outperforms nowcasting methods after that leading time.

Enterprise Integrated Aviation Weather System (eIAWS) developed by I.M. Systems Group Inc. (IMSG) is a powerful real time system. It is particularly designed for and serves to the aviation industry. It has been installed in a few airports and a couple of regional aviation traffic control centers in China. eIAWS has two subsystems among others. One is the nowcasting subsystem (imNowcastor); another is Real time Regional Rapidly Refreshing (R3R) model subsystem. imNowcaster can provide users with nowcasting up to 6 hours in every 10 minutes and R3R subsystem can provide forecast with high spatial resolution every 10 minutes too. By combining nowcasting from imNowcaster and forecast from R3R, eIAWS provides users a blended result, which is aimed at extending leading time of current nowcasting and improving nowcasting performance beyond two hours.