Numerical Weather Prediction Model Performance over High Southern Latitudes

Abstract

Increasingly, output from numerical weather prediction (NWP) models is being used for real-time weather forecasts for the Antarctic and for Antarctic-related climate diagnostics studies. Evidence is presented that indicates that in broad terms, the NWP output from the major global models is providing useful representations of synoptic-scale systems over high southern latitude areas. For example, root-mean-square (rms) errors in the European Centre for Medium-Range Weather Forecasts (ECMWF) model predictions of the 500-hPa height field indicate a day's gain in predictability since the mid-1990s: average rms errors in ECMWF +72 h 500-hPa height field prognoses for the calendar year 2000 were close to 50 m, compared to similar errors in the +48 h prognoses in 1995. Similar relative improvements may be noted for all time steps out to +144 h. Moreover, it is determined that, of the models considered here, the ECMWF model is clearly the most successful model at 500-hPa-height prediction for high southern latitudes, with the United Kingdom Met Office (UKMO) and National Centers for Environmental Prediction Aviation (AVN) models the next most accurate, and with the Australian Bureau of Meteorology's Global Assimilation Prediction (GASP) and Japanese Meteorological Agency (JMA) models lagging in accuracy. However, improvements in the temporal and spatial resolution of observational data that are available to the analysis and assimilation cycles of the NWP models, and improvements in the horizontal resolutions of the models, are required before the use of NWP output at high southern latitudes is as effective as in more northern areas of the world. Limited area modeling is seen as having potential for complementing the global models by resolving the finer-scale orography and topography of the Antarctic.