Dry thunderstorms are a major source of wildfires, and are disproportionately responsible for igniting major wildfires, yet currently dry thunderstorm predictions at high spatial resolutions are unavailable for most of the country. We propose development of improved dry lightning algorithms to expand their coverage, and improve accuracy and usability of such predictions in fire management. In JFSP project #01-1-6-08, a discriminant algorithm was developed for the Pacific Northwest. Additional work by NOAA's Storm Prediction Center (SPC) has produced a statistical scheme to predict the location and intensity of lightning outbreaks. We propose here to build on this work in 3 ways: -Incorporate additional data to expand these predictions to other geographic areas as appropriate, including the interior of Alaska -Incorporate predictions of large lightning outbreaks (storms that generate hundreds to thousands of lightning strikes over relatively small areas) which ignite multiple fires, overwhelm suppression resources, and lead to a disproportionate number of uncontrolled wildfires -Determine the feasibility of using an alternative, physically based algorithm that includes temperature and moisture from multiple vertical atmospheric layers Specifically, we will obtain lightning strike data for Alaska and Canada (which was not available at the time we started our original JFSP project) to generate risk predictions for these domains. Additionally, we will incorporate a methodology developed by SPC to predict the probability of high numbers of cloud-to-ground flashes. By integrating these two approaches we will be able to create models for large lightning outbreaks occurring without significant rainfall and put into place forecasts available through our website. To address a weakness in the existing algorithm (it considers meteorological variables at only two vertical levels in the atmosphere), we will also develop and test a new geographically independent algorithm for identifying dry thunderstorm days, using moisture and temperature variables throughout a deeper atmospheric layer.