First forecast of a sudden stratospheric warming with a coupled whole-atmosphere/ionosphere model

We present the first medium-range “weather forecast” with a coupled whole-atmosphere/ionosphere model, built under the Integrated Dynamics in Earth's Atmosphere (IDEA) project, for the period of the January 2009 Sudden Stratospheric Warming (SSW). The IDEA model consists of the Whole Atmosphere Model (WAM) and Global Ionosphere-Plasmasphere (GIP) model running concurrently and exchanging necessary fields in the overlapping layer (about 90-600 km) at each time step. WAM was built on the NCEP operational Global Forecast System (GFS) model by extension from 60 km to about 600 km top altitude. It has been incorporated into the operational data assimilation system. In addition to unconstrained “free” runs, this provides a means to run it in an analysis or forecast cycle mode as GFS. In previous studies the analysis cycle output has been used retrospectively to drive standalone ionospheric models, including GIP, to study the electrodynamics and plasma density response to particular events such as the major January 2009 SSW. This one-way offline coupling has shown good agreement with ionospheric observations. Here we investigate a 30-day forecast by the IDEA model initialized from an analysis at 00Z on January 13, ten days prior to the peak of the SSW. The IDEA model successfully predicts both the time and amplitude of the peak warming in the polar cap. This is about 2 days before the operational GFS forecast. The forecast of the major tides SW2 and TW3 also shows an increase in amplitudes and phase changes in the dynamo region during and after the peak. The forecast of the ionosphere response is also analyzed for changes in daytime plasma drifts and total electron content (TEC) in the equatorial American sector. The changes in the tidal winds drive a clear shift to earlier hours of daytime upward drifts during and after the SSW peak. Correspondingly, TEC also exhibits a clear shift and increase to earlier hours. These changes compare well with available observations.