Objective Forecasts of Surface Temperature from One to Three Days in Advance

Abstract

An objective method of forecasting surface temperature at 39 cities in the United States for 1, 2 and 3 days in advance has been developed and tested under operational conditions. This method utilizes daily input to multiple regression equations derived previously for predicting 5-day mean temperature from fields of 700-mb height and surface temperature. Comparative verification shows that the objective method produces forecasts which approximate the skill of subjective predictions made by experienced meteorologists. Recent experiments are described which indicate that improved forecasts result when 700?1000 mb thickness is included as a predictor. The beneficial synoptic climatology which accrues from these studies is also illustrated.