This post is a follow-on to “Modeling Activities at WFO Mobile” regarding the Weather Research and Forecasting (WRF) model local predictions of Tropical Storm Claudette from 17 August, using MODIS SSTs in version 3 of the Environmental Modeling System (EMSv3). This past week, SPoRT ran a similar configuration of the WRF-NMM model within EMSv3 using MODIS SSTs for one run, and NCEP’s Real-Time Global (RTG) SSTs in another simulation (the current standard for operational SSTs), both initialized at 0300 UTC 17 August. That way, we can exactly compare the impacts of the high-resolution SPoRT MODIS SSTs to the coarser resolution RTG SSTs. All other model configurations were identical: GFS model initial and boundary conditions, physics parameterizations, etc.
The results from these runs were rather intriguing. While we were unable to duplicate the identical result that WFO Mobile captured, the trends were similar. Lower mean sea level pressures tended to develop further east in the WRF run using MODIS SSTs, closer to the actual landfall of the system. Figure 1 depicts the difference in SSTs that went into the model’s lower boundary condition at water points. Notice the locally warm pocket of SSTs offshore of western Florida, with locally cooler SSTs closer to shore.
The SST differences had a direct impact on the short-term evolution of the mean sea level pressure (MSLP) and 10-m wind fields, as shown in the 2-h to 6-h forecasts in Figures 2 through 4 (same times as in the WFO Mobile blog post). In the 2-h forecast (Fig. 2), the forecast wind field is already responding to the improved lower boundary forcing with the MODIS SSTs as indicated by a slight change in the MSLP field and more notably the 10-m wind field. By the 4-h forecast, a distinct enhancement in cyclonic flow is evident in the 10-m wind difference field offshore of western Florida, associated with slightly lower MSLP (Fig. 3). Finally, at 6 hours, an elongated region of enhanced convergence/cyclonic flow is evident, coincident with a high/low couplet of MSLP differences on the order of a few tenths of a millibar. This is coincident with the location of the mesolow depicted in the WFO Mobile blog post. The maximum change in the individual wind components was on the order of 3 m s-1 at 6 hours (not shown).