A 3DVAR land data assimilation scheme: Part 2, test with ECMWF ERA-40

In the first part of this paper, a SDVar Land Data Assimilation Scheme (LDAS) is presented. With virtue of this land data assimilation system, this part of the paper demonstrates the results and error analysis of assimilating air temperature data observed at various meteorological stations in China into the output of ECMWF ERA-40. The air temperature distribution of sparse observation zones is obtained, which shows the validity of the assimilation procedure. The 3DVar LDAS can greatly improve the ECMWF background estimates with the high quality observations of air temperature from the Chinese meteorological stations. By comparing the assimilated air temperature field and the ECMWF background field to the observations, the assimilation outputs have better agreement with the air temperature variation trend than the ECMWF background. Another advantage of the assimilated result is that it can describe the extreme air temperature more accurately.