D. Noone et H. Stern, VERIFICATION OF RAINFALL FORECASTS FROM THE AUSTRALIAN BUREAU OF METEOROLOGYS GLOBAL ASSIMILATION AND PROGNOSIS (GASP) SYSTEM, Australian meteorological magazine, 44(4), 1995, pp. 275-286
The Australian Bureau of Meteorology (BOM) implemented a high resoluti
on version of its Global Assimilation and Prognosis System (GASP) on 2
2 March 1994. A dataset comprising rainfall forecasts for three Austra
lian cities from the first available 100 days of GASP quantitative pre
cipitation forecasts was constructed and verified against observed rai
nfall. Several measures of skill were used to evaluate GASP prediction
s. These were comparisons with persistence, climatology, 'model output
statistics', 'perfect prog' and BOM Regional Forecast Centre (RFC) pr
edictions. GASP generally provided useful guidance up to four days in
advance. However, it poorly forecast rain events associated with strea
m weather and post-frontal cold air, while precipitation accompanying
fronts was indicated with considerable reliability.