Document: RAN-3-68-47

Modeling the spatial patterns of forest fuels in the Los Alamos region.

BALICE, R.G.* ¹, S.W.KOCH ¹ and S.R.YOOL ²

Los Alamos National Lab, Los Alamos, NM 87545 NM USA ¹
University of Arizona, Tucson, AZ 85721 USA ²

Abstract:
To develop more robust GIS databases for wildfire behavior modeling and for wildfire hazard assessment, we examined the spatial and topographic relationships of fuels in the forest and non-forest communities in the region surrounding Los Alamos, New Mexico. To accomplish this we sampled 76 plant communities over a two-year period. The field data include topographic information, soils characteristics, understory fuels and vegetation and overstory canopy structures. The data were summarized and analyzed using linear modeling techniques. From these analyses, it was found that both overstory and understory fuels are related to linear and quadratic representations of topographic and remotely sensed variables. As an example, the final regression model for percent canopy cover was significant (F = 16.23, P - value < 0.0001) with an R - square of 0.571. This and other regression models are being used as inputs of conditional simulation to predict fuel levels at unsampled pixels within the study area.

Keywords:

This abstract is being presented at: 2:45 PM in session:
Oral Session #64: Remote Sensing.