Subregion-scale hindcasting of contrail outbreaks, utilizing their synoptic climatology

Andrew Mark Carleton, Armand D. Silva, Jase Bernhardt, Justin VanderBerg, David J. Travis

Research output: Contribution to journalArticle

Abstract

Contrail statistical prediction methods are often location specific. To take advantage of the fact that the upper-tropospheric (UT) meteorological conditions that favor "clear-sky outbreaks" of persisting contrails, or contrail favored areas (CFAs), tend to be synoptic in scale, a visual UT-map technique to hindcast CFAs has been developed and tested for subregions of the contiguous United States (CONUS) that have high outbreak frequencies in midseason months (January, April, July, and October) of 2000-02. The method compares daily maps with the composite fields for outbreak days (CON) versus nonoutbreak days (NON), and those assessments are evaluated using standard skill measures. Binary logistic regression determines which UT variables are significant predictors, individually and in combination. The reproducibility of the outbreak hindcast results is tested on the same subregions for the corresponding months of 2008-09. The results confirm the importance of UT relative humidity and vertical-motion (omega) map patterns in regional clear-sky outbreaks. Although the hindcast skill is modest, sensitivity tests suggest that the method will be substantially improved when a longer-term climatological dataset of outbreaks becomes available (to increase sample sizes) and with explicit consideration of the synoptic types on CON days. The latter is demonstrated specifically for the southern CONUS in January, where to improve hindcast success one should also consider the vertical wind shear in the upper troposphere, given the importance of the subtropical jet stream in contrail outbreaks there. Further development of the method to improve its skill ultimately should permit its use in combination with existing objective (statistical and physical models) methods of contrail prediction.

Original languageEnglish (US)
Pages (from-to)1733-1755
Number of pages23
JournalJournal of Applied Meteorology and Climatology
Volume54
Issue number8
DOIs
StatePublished - Jan 1 2015

    Fingerprint

All Science Journal Classification (ASJC) codes

  • Atmospheric Science

Cite this