PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
February 5, 2026Australian meteorological magazine0 citations

The use of linear regression to improve official temperature forecasts

View Full Paper
FWF. WoodcockBSB. Southern

Key Points

  • The aim is to improve the accuracy of daily maximum and minimum temperature forecasts in Australia.
  • Combined MOS forecasts with RFC forecasts and local observations.
  • Developed Local Forecast Equations (LFE) for various Australian cities.
  • Tested the LFE using independent data to evaluate improvements.
  • Reduced errors exceeding 3°C in maximum temperature forecasts by about 30%.
  • Reduced errors in minimum temperature forecasts by about 50%.
  • Significant improvements due to incorporating MOS forecasts in LFE.

Abstract

Model output statistics (MOS) forecasts of daily maximum and minimum temperature are combined with Regional Forecast Centre (RFC) forecasts and 3 pm local observations to produce Local Forecast Equations (LFE) for daily maximum and minimum temperatures for Adelaide, Brisbane, Canberra, Hobart, Melbourne, Perth and Sydney. Testing of the LFE, using independent data, indicates that a substantial improvement on current operational accuracy can be achieved. Specifically, the number of errors exceeding 3°C in the operational forecasts of maximum temperatures can be reduced by about 30 per cent and of minimum temperatures by about 50 per cent. The improvement in the forecasts is mostly due to the inclusion of MOS forecasts in the LFE and partly due to the statistical correction of RFC forecasts. A change to current operational forecast strategy is suggested.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Woodcock et al. (1983) studied this question.

synapsesocial.com/papers/698435c9f1d9ada3c1fb5069https://doi.org/10.1071/es83007
Ask AI
Helpful
Bookmark
Share
View Full Paper