8C.6 Next-Generation OK-FIRE Modeling System

Wednesday, 15 January 2020: 11:45 AM
151A (Boston Convention and Exhibition Center)
Michael D. Klatt, University of Oklahoma, Norman, OK

The OK-FIRE program supports firefighting activities and prescribed burn planning for the state of Oklahoma. The components of OK-FIRE are 1) wildfire danger and smoke dispersion modeling; 2) web-based decision-support tools; and 3) training seminars for using these tools. OK-FIRE is a joint project of Oklahoma State University and the University of Oklahoma and is operated by the Oklahoma Mesonet.

The core of OK-FIRE is a wildfire danger model. This model uses the National Fire Danger Rating System (NFDRS) calculations with some local modifications. Weather conditions are ingested from the Oklahoma Mesonet for real-time analyses and from the North American Mesoscale Forecast System (NAM) for forecasts. Live fuel moisture is modeled using satellite-based Normalized Difference Vegetation Index (NDVI) data. Dead fuel moisture is calculated using the Nelson model with local adaptations. The model output includes the Spread Component, Energy Release Component, Ignition Component, and Burning Index fire. Currently, the model updates hourly with real-time analyses and forecasts out to 84 hours.

In 2016, planning began for the next major version of OK-FIRE. There were two main goals for the wildfire danger model upgrade, 1) switching to Moderate Resolution Imaging Spectroradiometer (MODIS) data for NDVI and 2) increasing the model resolution from 1000 meters to 500 meters. Secondary goals were to 1) increase the real-time update frequency from 60 minutes to 15 minutes and 2) simplify the model code to take advantage of hardware advances. As work progressed, it was decided that a more extensive rewrite of the model and its supporting applications was necessary. As well as satisfying the primary and secondary update goals, this rewrite will 1) allow the model to be deployed to a wider range of platforms (e.g. cloud computing); 2) make the system more robust and easier to run; 3) make the model more suitable for research purposes; and 4) introduce the latest best practices for software development.

This presentation will describe the latest version of the OK-FIRE wildfire danger model, including comparisons to the previous version of the model and national products from the US Forestry Service and the National Weather Service.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner