Tuesday, 8 January 2013: 4:45 PM
Room 4ABC (Austin Convention Center)
Thomas J. Galarneau Jr., NCAR, Boulder, CO; and T. M. Hamill, M. Fiorino, and J. S. Whitaker
Ensemble reforecasts are a long time series of ensemble forecasts from a stable data assimilation and forecast system, which can be used to statistically adjust real-time forecasts. Reforecasts provide the large sample size that may be necessary to make statistically meaningful corrections to real-time forecasts of rare events such as tropical cyclones (TCs). Statistical adjustment of temperature and precipitation forecasts in midlatitudes using NOAA's first-generation Global Forecast System (GFS) ensemble reforecast dataset provided dramatic forecast improvements. A new second-generation GFS ensemble reforecast dataset has been produced at NOAA/ESRL/PSD. This new reforecast dataset was generated using the current 2012 version of the operational NCEP GFS ensemble at T254L42 resolution out to eight days, and T190L42 thereafter out to 16 days. The dataset includes 11 members run daily at 0000 UTC from mid-1984 to present initialized with the Climate Forecast System reanalysis. Perturbed initial conditions were generated with the operational ensemble transform with rescaling technique. We applied an advanced tracker and diagnostic process that not only tracks existing TCs but locates cyclones that form during the forecast.
The aim of this presentation is to exploit the new second-generation ensemble reforecast dataset across a variety of research and forecasting problems for TCs, including track, intensity, and genesis forecasts, long-lead (week 2) products, and post-landfall precipitation and track forecasts.
The talk will describe the data set in more detail, and then describe some simple applications for TC forecasting, such as the use of the reforecasts to statistically adjust TC track and intensity forecasts over the North Atlantic basin. The utility of producing a high-resolution regional reforecast for selected cases using the Weather Research and Forecasting model, with the global reforecast data serving as the boundary and initial conditions, will also be discussed.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner