摘要
The Weather Research and Forecasting model(WRF)is configured for the region of(15°–41°N,105°–135°E),which covers the same area with the MASNUM(Key Lab.Marine Science&Numerical Modeling,State Oceanic Administration)wave-tide-circulation coupled operational forecast system.Three numerical experiments are implemented to investigate the effects of the real-time forecasted sea surface temperature(SST)and the nonbreaking wave-induced vertical mixing(Bv)on the track forecast of all 33 tropical cyclones(TC)in the model domain area during 2008 and 2011.The first experiment employs NCEP FNL(NCEP final analysis)SST as WRF’s bottom condition as the Control run,which is also the default setup of WRF.The second and third experiments use real-time forecasted SST from the MASNUM forecast system with and without Bv,respectively.The forecasted track results are compared with Japan Meteorological Agency’s best track data.For 24-h forecast,the averaged TC position error of Experiment with Bv is reduced by 9%compared to the Control experiment,while the forecasted track error of Experiment without Bv is reduced by only2%compared to the Control experiment.For the 48-h forecast,the averaged track errors are reduced by 10%and6%with Bv and without Bv compared to the Control experiment,respectively.These results suggest that the real-time forecasted SST can improve the performance of WRF in forecasting TC track,and the Bv plays an important role in reducing the forecast error of TC track.Comparatively,Bv can improve more on the track of stronger TC.
The Weather Research and Forecasting model （WRF） is configured for the region of （15°-41°N, 105°- 135°E）, which covers the same area with the MASNUM （Key Lab. Marine Science ＆ Numerical Modeling, State Oceanic Administration） wave-tide-circulation coupled operational forecast system. Three numerical experiments are implemented to investigate the effects of the real-time forecasted sea surface temperature （SST） and the nonbreaking wave-induced vertical mixing （By） on the track forecast of all 33 tropical cyclones （TC） in the model domain area during 2008 and 2011. The first experiment employs NCEP FNL （NCEP final analysis） SST as WRF＇s bottom condition as the Control run, which is also the default setup of WRF. The second and third experiments use real-time forecasted SST from the MASNUM forecast system with and without By, respectively. The forecasted track results are compared with Japan Meteorological Agency＇s best track data. For 24-h forecast, the averaged TC position error of Experiment with By is reduced by 9 % compared to the Control experiment, while the forecasted track error of Experiment without Bv is reduced by only 2 % compared to the Control experiment. For the 48-h forecast, the averaged track errors are reduced by 10 % and 6 % with Bv and without By compared to the Control experiment, respectively. These results suggest that the real-time forecasted SST can improve the performance of WRF in forecasting TC track, and the Bv plays an important role in reducing the forecast error of TC track. Comparatively, By can improve more on the track of stronger TC.
基金
supported by the Public Scienceand Technology Research Funds Projects of Ocean(201105019)
