Monday, 28 August 2017
Zurich DEFG (Swissotel Chicago)
Radar-based nowcasting systems are widely used for quantitative precipitation forecasting (QPF) within 6 hours. Weather Radar Center (WRC) of Korea Meteorological Administration (KMA) produces the precipitation forecasts using the McGill Algorithm for Precipitation nowcasting by Lagrangian Extrapolation (MAPLE) based on a nationwide composite of radar reflectivity.
Two intrinsic problems were raised in terms of operational nowcasting system: One is the significant spatial distortion of precipitation system near boundary regions in MAPLE. Second issue is that the motion vectors of precipitation system are seriously subject to weighting factor of MAPLE.
In this study, we examined the sensitivity of MAPLE to different boundary conditions. The smaller domain was used as the control simulation (CTRL) and composed large domain with margin (EXP1). Furthermore, CNTL has a different boundary condition without margin (EXP2). The experiment with larger domain without margin (EXP2) shows a better skill scores in evaluation of precipitation forecasting. We also improved the performance of precipitation nowcasting by optimizing the cost functions of Variational Echo Tracking (VET) in MAPLE. These weighting factors were determined to the optimum combination. The spatial variation of motion vectors was reduced by optimization of weighting factor. The relatively smaller variation results in better forecast skills and lower false alarm ratio.
Two intrinsic problems were raised in terms of operational nowcasting system: One is the significant spatial distortion of precipitation system near boundary regions in MAPLE. Second issue is that the motion vectors of precipitation system are seriously subject to weighting factor of MAPLE.
In this study, we examined the sensitivity of MAPLE to different boundary conditions. The smaller domain was used as the control simulation (CTRL) and composed large domain with margin (EXP1). Furthermore, CNTL has a different boundary condition without margin (EXP2). The experiment with larger domain without margin (EXP2) shows a better skill scores in evaluation of precipitation forecasting. We also improved the performance of precipitation nowcasting by optimizing the cost functions of Variational Echo Tracking (VET) in MAPLE. These weighting factors were determined to the optimum combination. The spatial variation of motion vectors was reduced by optimization of weighting factor. The relatively smaller variation results in better forecast skills and lower false alarm ratio.
Acknowledgment
This research is supported by “Development and application of cross governmental dual-polarization radar harmonization (WRC-2013-A-1)” project of the Weather Radar Center, Korea Meteorological Administration.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner