Predicting slow‐drying fire weather index fuel moisture codes with NOAA‐AVHRR images in Canada's northern boreal forests

Abstract

Fire danger predicted by the Canadian Fire Weather Index, a system based on point‐source weather records, is limited spatially. NOAA‐AVHRR images were used to model two slow‐drying fuel moisture codes, the duff moisture code and the drought code of the fire weather index, in boreal forests of a 250,000 km² portion of northern Alberta and the southern Northwest Territories, Canada. Temporal and spatial factors affecting both codes and spectral variables (normalized difference vegetation index, surface temperature, relative greenness, and the ratio between normalized difference vegetation index and surface temperature) were identified. Models were developed on a yearly and seasonal basis. They were strongest in spring, but had a tendency to saturate. Drought code was best modelled (R ² = 0.34–0.75) in the spring of 1995 when data were categorized spatially by broad forest cover types. These models showed improved spatial resolution by mapping drought code at the pixel level compared to broadly interpolated weather station‐based estimates. Limitations and possible improvements of the study are also discussed.

Acknowledgements

This study was funded by the National Science and Engineering Research Council of Canada and the Canadian Forest Service. Satellite imagery was provided by Rob Fraser and Rasim Latifovic of the Canada Center for Remote Sensing. Weather data were provided by the Canadian Forest Service, Alberta Department of Energy and Environment, Northwest Territories Forest Management Division, and Parks Canada. We are thankful to Meg Krawchuk for her help in acquiring weather data. We express our appreciation to Dr Rolf Turner of the University of New Brunswick Statistics Department and Dr John Kershaw and Hannelie Botha of the University of New Brunswick Faculty of Forestry for their assistance in statistics and analysis during this study.