Date of Award

3-8-2002

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Engineering Physics

First Advisor

Timothy M. Jacobs, PhD

Abstract

The Air Force Combat Climatology Center (AFCCC) is tasked to provide long-range seasonal forecasts for worldwide locations. Currently, the best long-range temperature forecasts the weather community has are the climatological standard normals. This study creates a stepping-stone into the solution of long-range forecasting by finding a process to predict temperatures better than those using climatological standard normals or simple frequency distributions of occurrences. Northern Hemispheric teleconnection indices and the standardized Southern Oscillation index are statistically compared to three-month summed Heating Degree Days (HDDs) and Cooling Degree Days (CDDs) at 14 U.S. locations. First, linear regression was accomplished. The results showed numerous valid models, however, the percent of variance resolved by the models was rarely over 30%. The HDDs and CDDs were then analyzed with Data-mining classification tree statistics, however, the results proved difficult to extract any predictive quantitative information. Finally a Data-mining regression tree analysis was performed. At each conditional outcome, a range of HDDs/CDDs is produced using the predicted standard deviations about the mean. Verification of independent teleconnection indices was used as predictors in the conditional model; 90% of the resulting HDDs/CDDs fell into the calculated range. An overall average reduction in the forecast range was 35.7% over climatology

AFIT Designator

AFIT-GM-ENP-02M-09

DTIC Accession Number

ADA404202

Download

Share

COinS