Accession Number:

ADA422988

Title:

Developing a Forecast Tool for Cloud-to-Ground Lightning in the North Central and Northeastern United States

Descriptive Note:

Master's thesis Jun 2003-Mar 2004

Corporate Author:

AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF ENGINEERING AND MANAGEMENT

Personal Author(s):

Report Date:

2004-03-01

Pagination or Media Count:

134.0

Abstract:

Cloud-to-ground CG lightning is a hazard to the Air Force for both air and ground operations Forecasting CG lightning is a necessary and extremely important requirement for Air Force meteorologists and forecasters. The 15th Operational Weather Squadron requested a forecast tool capable of predicting CG lightning within a 25 and 10 nautical mile radius of the 13 military locations in their area of responsibility. To fulfill their request, forecast decision tools were created using a Classification and Regression Tree CART data analysis program. Four decision trees were produced for each location using the period of record from March through September, 1993 to 2002, CART compared the upper air stability indices and surface data at 12-hour intervals with CG lightning data occurring within the next 12 hours to determine prediction rules. Data from 2003 were used as independent verification of the decision trees. The CART decisions trees were examined using contingency tables and verification tests to determine the value of the products created. The straight forward forecast rules and verification test results confined that the decision trees would be a valuable prediction tool. Combined with forecaster knowledge, forecast models and other tools, the decision trees would provide an excellent forecast method for determining the occurrence of CG lightning. Therefore, the results are recommended to the 15th Operational Weather Squadron for use as a CG lightning forecast tool.

Subject Categories:

  • Meteorology

Distribution Statement:

APPROVED FOR PUBLIC RELEASE