Dew point temperature prediction using artificial neural networks

Dew point temperature is the temperature at which water vapor condenses. It is an important weather variable used to estimate frost, fog, rain, snow, dew, evapotranspiration, near-surface humidity, and other meteorological variables. Dew point temperature directly or indirectly contributes to productivity of plants, crop damage during freezes, human comfort levels, and the loss of human life during heat waves. Although several studies have focused on the estimation of dew point temperature, little attention has been given to short term prediction. An artificial neural network (ANN) is a robust computational tool useful for prediction. The goal of this research was to develop ANNs that predict hourly dew point temperatures for up to twelve hours. This system of ANNs was trained on historical weather data from stations located throughout the state of Georgia. These ANNs will be implemented as part of a web-based decision support system.