An Adaptive Data Processing System for Weather Forecasting

View More View Less
  • a Stanford University, Stanford, Calif.
  • | b U.S. Weather Bureau, San Francisco, Calif.
© Get Permissions Rent on DeepDyve
Restricted access

Abstract

An adaptive pattern classification system termed Adaline is described. This system, when trained on 200 winter sea level pressure and 24-hr pressure change patterns covering the area from 25N to 65N and 110W to 170W, was able to make “rain”-“no rain” forecasts for the San Francisco Bay Area on 100 independent cases that compared favorably with the official U.S. Weather Bureau forecasts for the same periods. The pattern of weights developed by Adaline on the training data were meteorologically significant, showing that areas of low pressure and 24-pressure falls centered to the northwest of San Francisco were favorable for “rain” today with these centers shifting progressively westward to be favorable for “rain” tonight and tomorrow.

Abstract

An adaptive pattern classification system termed Adaline is described. This system, when trained on 200 winter sea level pressure and 24-hr pressure change patterns covering the area from 25N to 65N and 110W to 170W, was able to make “rain”-“no rain” forecasts for the San Francisco Bay Area on 100 independent cases that compared favorably with the official U.S. Weather Bureau forecasts for the same periods. The pattern of weights developed by Adaline on the training data were meteorologically significant, showing that areas of low pressure and 24-pressure falls centered to the northwest of San Francisco were favorable for “rain” today with these centers shifting progressively westward to be favorable for “rain” tonight and tomorrow.

Save