U.S. Joint Fire Science Program


Date of this Version


Document Type



JFSP Project Number: 10-1-07-16


US government work.


Fire weather forecasts rely on numerical weather simulations where the grid size is 4 km x 4 km or larger. In areas of complex terrain, this model resolution will not capture the details of wind flows associated with complicated topography. Wind channeling in valleys, wind speed-up over mountains and ridges, and enhanced turbulence associated with rough terrain and tall forest canopies are poorly represented in current weather model applications. A number of numerical wind flow models have been developed for simulating winds at high resolution; however, there are limited observational data available at the spatial scales appropriate for evaluating these types of models. In response to this need for high resolution validation data, we collected wind measurements at very high spatial resolution over a range of meteorological conditions from three different types of terrain/landcover features: an isolated mountain covered predominantly by grass and sagebrush, a steep river canyon covered predominantly by grass, and a dissected montane drainage with a tall forest canopy. We used data from the isolated mountain and the steep river canyon to evaluate surface wind predictions from routine weather forecasts and a high resolution wind simulation model, WindNinja, developed specifically for fire behavior applications. Data from the third field site will be used for future model evaluations planned to investigate the effect of tall forest canopies on surface wind predictions. Analyses of observations from the isolated mountain and steep river canyon sites indicate that operational weather model (i.e., with numerical grid resolutions of around 4 km or larger) wind predictions are not likely to be good predictors of local near-surface winds (i.e., at sub-grid scales) in complex terrain. Under periods of weak synoptic forcing, surface winds tended to be decoupled from large-scale flows, and under periods of strong synoptic forcing, variability in surface winds was sufficiently large due to terrain-induced mechanical effects that a large-scale mean flow would not be representative of surface winds at most locations on or within the terrain feature. These findings are reported in a manuscript titled “High Resolution Observations of the Near-Surface Wind Field over an Isolated Mountain and in a Steep River Canyon” submitted for publication in Atmospheric Chemistry and Physics. Links to the observed data from this effort as well as an online interface to query, visualize, summarize, and download subsets of the data are available at: http://www.firemodels.org/index.php/windninja-introduction/windninja-publications. Findings from the model evaluations work indicate that using WindNinja to downscale from numerical weather prediction (NWP) model winds can, in some cases, improve the accuracy of surface wind forecasts in complex terrain. Predictions of surface wind speeds and directions improved with downscaling via WindNinja when flow features induced by large scale effects were adequately captured by the NWP model used to initialize WindNinja. This suggests that WindNinja could be incorporated into current fire forecast methods to provide better short-term forecasts for fire management operations. These findings are reported in a manuscript titled “Downscaling Surface Wind Predictions from Numerical Weather Prediction Models in Complex Terrain with a Mass-consistent Wind Model” that will be submitted to the Journal of Applied Meteorology and Climatology later this spring.