Dowdy, AJ; Bates, BC; Chandler, RE; (2018) Lightning prediction for Australia using multivariate analyses of large-scale atmospheric variables. Journal of Applied Meteorology and Climatology pp. 525-534. 10.1175/JAMC-D-17-0214.1.

Preview

Text Bates_Lightning.pdf - Published Version Download (502kB) | Preview

Abstract

Lightning is a natural hazard that can lead to the ignition of wildfires, disruption and damage to power and telecommunication infrastructures, human and livestock injuries and fatalities, and disruption to airport activities. This paper examines the ability of six statistical and machine-learning classification techniques to distinguish between non-lightning and lightning days at the coarse spatial and temporal scales of current general circulation models and reanalyses. The classification techniques considered were: a combination of principal component analysis and logistic regression; classification and regression trees; random forests; linear discriminant analysis; quadratic discriminant analysis; and logistic regression. Lightning flash count observations at six locations across Australia for the period 2004 to 2013 were used, together with atmospheric variables from the ERA-Interim reanalysis. Ten-fold cross validation was used to evaluate classification performance. It was found that logistic regression was superior to the other classifiers considered, and that its prediction skill is much better than climatology. The sets of atmospheric variables included in the final logistic regression models were primarily composed of spatial mean measures of instability and lifting potential, and atmospheric water content. However, the memberships of these sets varied between climatic zones.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item