Automatic Forecasts Algorithms

Abstract

SUMMAR Y The need for forecasts at an increasingly higher number of locations and time frequency requires a fully automated forecast system. Due to model biases, statistical post-processing is required to improve the quality of the forecasts in variables such as temperature, humidity and wind speed. The automatic forecasting algorithms developed at IPMA provide forecasts up to 10 days and use ECMWF and AROME data.