Forest-fire spread is controlled by mountain- and valley-scale winds in BC, but these winds are not resolved in the present-day coarse-resolution weather analyses. When these analyses are used as a starting point for numerical weather prediction (NWP) models, the resulting forecasts are inaccurate. To remedy this problem, we created a technique to utilize the 150-200 surface weather stations deployed by the BC Ministry of Forests (MoF), along with 400 additional stations deployed by other agencies, to make new daily analyses. These can be used to initialize the high-resolution NWP models that are run daily at UBC. The skill of this approach will be measured by verifying the resulting forecasts against the corresponding observations, and comparing against forecasts that did not incorporate this data. Also, the daily forecasts were disseminated on the web for real-time use and evaluation by MoF fire weather forecasters.
Roland B. Stull, Xingxiu Deng.
Deng, Xingxiu, Stull, Roland B.. 2004. Improved numerical prediction of fire weather by assimilation of MoF surface weather observations. Forest Investment Account (FIA) - Forest Science Program. Forest Investment Account Report. FIA2004MR250