Impact of Atmospheric River Reconnaissance Dropsonde Data on NCEP GFS Forecast: A Case Study

Abstract:

Atmospheric rivers (ARs) are long narrow corridors of water vapor transport that serve as the primary mechanism to advect moisture into mid-latitude continental regions, including the U.S. West Coast. They are responsible for most of the horizontal water vapor flux outside of the tropics and a source of precipitation. Although the advancements in satellite data assimilation has greatly improved global model forecast skill, including the NCEP global forecast system (GFS), forecasting the AR features remains a challenge due in part to their formation and propagation over the ocean, where in-situ and ground-based observations are extremely limited. The AR Reconnaissance (AR Recon) Campaigns that took place during winter 2016, and 2018-2020 provides additional data by supplementing conventional data assimilation with dropsonde observations of the full atmospheric profile of water vapor, temperature, and winds within ARs.

In this study we used NCEP GFS version 15 (GFSv15) to examine the impact of the AR supplemental observations dropsonde data on GFS forecast. GFSv15 was implemented in operations in June 2019; it has been developed with the finite volume cubed-sphere dynamical core (FV3) and microphysics from GFDL, and 4D-Hybrid En-Var data assimilation (DA). The dropsonde data used were from the AR Recon 2020 campaigns, including 17 intensive observation periods (IOPs). Global control and denial experiments were conducted by using or denying the dropsonde data in the GFS from January 24 to March 18 for both DA and model forecast.

Preliminary analysis indicates that there is systematic improvement for the precipitation prediction over the U.S. West Coast when the dropsonde data are used. This is associated with improvement of the water vapor transport (IVT) forecast. The AR supplemental observations have helped to fill the data gap that is needed for the data assimilation to provide better GFS model initial condition. The overall GFSv15 performance and the associated dropsonde data impact will be examined in detail and be presented at the conference.