Due to the availability of observations and the effectiveness of bias correction, it is still a challenge to assimilate data from the polar orbit satellites into a limited-area and frequently updated model.This study assessed the initial application of satellite radiance data from multiple platforms in the Rapid-refresh Multi-scale Analysis and Prediction System (RMAPS).Satellite radiance data from the advanced microwave sounding unit-A (AMSU-A) and microwave humidity sounding (MHS) were used.
Two 12-day retrospective runs were hubbell 5362w conducted to evaluate the impact of assimilating satellite radiance data on 0–24 h forecasts using RMAPS.The forecasts, initialized from analyses with and without satellite radiance data, were verified against observations.The results showed that satellite radiance data from AMSU-A and MHS had a positive impact on the initial conditions and the forecasts of RMAPS, even over the relatively data-rich area of North China.
Compared to the control run that only assimilated conventional observations, an improvement of about 36.8% can be obtained for the temperature bias between 300 hPa and 850 hPa and 0.65% for the average RMSE.
Satellite radiance observations from 1200 UTC contribute relatively significantly (77.8%) to the bias improvement of the initial temperature field.For the wind at 10 m, the bias and root-mean-square error (RMSE) pure energy jeans both had a reduction for the 0–12 h forecast range.
An improvement can be also found for the skill score of the 3-h accumulated rainfall below 10.0 mm in the first 12 h of the forecast range.There was a slight improvement in the skill score of the 6-h accumulated rainfall above 50 mm over North China, with a 20.
7% improvement for the first 12 h of the forecast.The inclusion of satellite radiance observations was found to be beneficial for the initial temperature, which consequently improved the forecast skill of the 0–12 h range in the RMAPS.