An impact study of five remotely sensed and five in situ data types in the Eta Data Assimilation System

The impact of 10 data types used in the Eta Data Assimilation/Forecast System (EDAS) is studied for extended-length time periods during three seasons. Five of the data types are remotely sensed satellite data, and the other five are in situ. The satellite data types include three-layer and vertically integrated precipitable water, temperature data down to cloud top, infrared cloud-drift winds, and water vapor cloud-top winds. The five in situ data types consist of two rawinsonde and two aircraft observation types along with surface land observations. The work described in this paper is relevant for Eta Model users trying to identify the impact of remotely sensed, largely maritime data types and in situ, largely land-based data types. The case studies chosen consist of 11-day periods during December 1998, April 1999, and July 1999. During these periods, 11 EDAS runs were executed twice daily. The 11 runs include a control run, which utilizes all data types used in the EDAS, and 10 experimental runs in which one of the data types is denied. Differences between the experimental and control runs are then accumulated and analyzed to demonstrate the 0-h sensitivity and 24-h forecast impact of these data types in the EDAS. Conventional meteorological terms evaluated include temperature, u component of the wind, and relative humidity on five pressure levels. These diagnostics are computed over the entire model domain and within a subsection centered on the continental United States (CONUS). The entire domain results show that a modest positive forecast impact is achieved from all 10 data types during all three time periods. Rawinsonde temperature and moisture observations and infrared cloud-drift wind observations have the largest positive impact season to season; however, both precipitable water data types provide significant positive forecast impact during the summer and transition seasons. Rawinsonde temperature and moisture, rawinsonde winds, aircraft winds, and infrared cloud-drift winds have the largest positive impact season to season over CONUS. The three-layer precipitable water data type produces large positive forecast impact over CONUS during July. In general, the forecast impacts are smaller for nearly all data types over CONUS than over the entire model domain. There are also more negative forecast impacts for both the in situ and remotely sensed data types over CONUS than over the entire domain.