Skip to main content
Author(s):
Jennifer L. Rechel, Seth H. Peterson, Dar A. Roberts, Jan W. van Wagtendonk
Year Published:

Cataloging Information

Topic(s):
Fuels
Fuel Descriptions
Ecosystem(s):
Sagebrush steppe

NRFSN number: 10998
FRAMES RCS number: 2903
Record updated:

Fuel moisture is one of the major components of fire risk assessment in the western United States. Regional and landscape fuel moisture estimates are currently derived from coarse resolution remotely sensed imagery without ground measurements to validate the estimates. Additionally, these estimates are determined using the Normalized Difference Vegetation Index (NDVI) which typically results in low R2 values indicating a poor relationship between NDVI and the actual live biomass fuel moisture. We collected above ground standing live biomass fuel moisture for coniferous and deciduous forested and mixed forest/shrublands at seven sites in the western United States during early season (leaf-out) and late season (senescence) to detect changes in fuel moisture over a fire season. Spectral mixture analysis (SMA) expresses pixel reflectance as endmember (EM) fractions; typically green vegetation (GV), nonphotosynthetic vegetation (NPV), soil, and shade. We linked the ground % fuel moisture values with the following MODIS image products: NDVI, Normalized Difference Water Index (NDWI), Normailzed Difference Infrared Index (NDII), Normalized Difference Green Red Index (NDRGI), Visible Atmospherically Resident Index (VARI) and GV, NPV, soil, and shade endmember fractions. R2 values for live fuel moisture and the image products during the high risk (hot and dry) fire season for a particular study site ranged from 0.4 to 0.7, during wetter periods R2 values were lower, ranging from 0.2 to 0.4. NDVI, NDII, and GV endmember fractions performed best, with NDWI having lower values. NDVI and GV are sensitive to vegetation cover, whereas NDWI is sensitive to vegetation condition, and works best when a pixel has complete canopy cover, which may not be the case for a 500m pixel. Soil fraction was important only for the most sparsely vegetated site. Imagery-based live fuel moisture predictions improve regional fire severity modeling by increasing the temporal resolution and spatial coverage of ground-based % live fuel moisture values.

Citation

Rechel, Jennifer L.; Peterson, S. H.; Roberts, Dar A.; Van Wagtendonk, Jan W. 2005. Predicting seasonal fuel moisture in the western United States using endmember fractions at multiple spatial and spectral resolutions. In: Proceedings of the 5th international workshop on remote sensing and GIS applications to forest fire management; 2005 June 16-18; Universidad de Zaragoza; Zaragoza, Spain. p. 245-248.

Access this Document