Quantification of forecast impact of ASCAT surface winds assimilated into NCEP's GFS model

A two-season Observing System Experiment (OSE) was used to quantify the impacts of assimilating the Advanced SCATterometer (ASCAT) surface winds product distributed by the European Organization for the Exploitation of Meteorological Satellites (EUMETSAT). The National Centers for Environmental Prediction (NCEP) Global Data Assimilation/Global Forecast System (GDAS/GFS) was used as the assimilation system and forecast model.

The impacts of assimilating the ASCAT surface wind products were assessed for two seasons by comparing the output through 168 hours of control simulations utilizing all the data types assimilated into the operational GDAS with experimental simulations using these new surface wind products. Quality control procedures required to assimilate the surface winds are discussed. Anomaly correlations (AC) of geopotential height at 1000 and 500 hPa were evaluated for the control and experiment during both seasons. The geographical distribution of forecast impact on the 10m wind, 500 hPa wind, 1000hPa temperature, and 500 hPa temperature fields are also presented.

In this presentation, the results of two studies are given. The results of a study using December 2007 version Gridpoint Statistical Interpolation (GSI) and 25 km high resolution ASCAT data indicated that the assimilation of ASCAT surface wind retrievals improves forecast skill in terms of traditional anomaly correlation. A comparison of the attributes of forecasts using the ASCAT data to a set of control forecast experiments by computing the geographic distribution of Forecast Impact (FI) was performed. Positive forecast impacts are noted for wind and temperature particularly in the tropics for 500hPa wind speed field starting from 6 hour forecast and beyond. The results of the earlier study are compared with those from a study using more recent version of the GSI and low resolution ASCAT data (50km) thinned to 100km.