A VAS-Numerical Model Impact Study Using the Gal-Chen Variational Approach

Abstract

Numerical experiments are conducted to assess the impact of incorporating temperature data from the VISSR Atmospheric Sounder (VAS) into a regional-scale numerical model using an assimilation technique developed by Gal-Chen. The technique uses a three-dimensional variational approach to combine the VAS observations with model temperature fields during the numerical integration. A nudging technique is also tested, whereby the model temperature field is constrained toward the VAS ?updated? values during the assimilation cycle. Results of the experiments indicate that the Gal-Chen assimilation technique successfully combines actual VAS temperature observations with the dynamically balanced model fields without destabilizing the model during the assimilation cycle. Furthermore, increasing the temporal frequency of VAS temperature insertions during the assimilation cycle enhances the impact on the model forecast through successively longer forecast periods. These results support the conclusions of earlier experiments with simulated geostationary satellite data that show an increasing positive impact on numerical simulations as the insertion rate of the satellite-based temperature information is increased. Incorporating the nudging technique further enhances the impact of the VAS temperature data, permitting the model wind field to adjust to the updated temperature fields and increasing the impact on the VAS data through a longer portion of the model simulation.