Rainfall Forecasting and its Applications FIJI CASE STUDY

Rainfall Forecasting and its Applications FIJI CASE STUDY Janita Pahalad National Climate Centre Aust. Bureau of Meteorology Simon Mc. Gree Climate Service Division Fiji Meteorological Service

Rainfall Forecasting Models – In Brief • FMS Rainfall Prediction Model (RPM) – ▪ ▪ ▪ Operational since July 1999 3 months forecast 25 stations ▪ ▪ ▪ Operational since August 1999 1 to 12 months forecast 21 stations • Australian Rainman Model (Aus. Rain) –

Statistical Models • Requires good quality, unbroken long historical • • data Simple, easily implemented, can run on desktop PC May require initial programming expertise Easily modified Implemented anywhere with local data

Applications • Drought Prediction Scheme • Flooding incidences • Frequency of tropical cyclones

User Applications • • • Agriculture – namely sugar industry Forestry Disaster managers Water resource managers Military Red Cross Hospitals Civil Aviation Construction Tourism

Information Dissemination • FMS Monthly Weather Summary • Website: http: //www. met. gov. fj/ • Media (Newspapers/radio) • Direct contact • Press Releases • DMET’s monthly meeting with DISMAC

Concerns and Future Needs • Users – sceptical about forecast – still fairly new concept in this region • Lack of understanding (or acceptance) of probabilistic forecasts • Need for public education on ENSO • Need to expand user network • Need for timely national climate forum • Need to further customise user products

Concerns and Future Needs • Educating the media is most needed • Limited interaction between FMS and water sector – should be enhanced • Examples like FSC and FPL should be applied in other sectors

Main Points • FMS example can be applied to other PIC. • Local knowledge is important and should be used • Regional/global models can be applied but there may be a need to downscale • Climate forecasting is still fairly new in the Region – public awareness is vital • Many prefer deterministic forecasts