DERIVING DYNAMIC INFORMATION ON FIRE FUEL DISTRIBUTIONS IN SOUTHERN CALIFORNIAN CHAPARRAL FROM REMOTELY-SENSED DATA

The major objective of the research reported in this paper was to perf orm a pilot study for mapping fire fuel properties of southern Califor nian chaparral shrubs and associated forest ecotones using multi-date (seasonal) Landsat Thematic Mapper (TM) image data. Information on pas t, current and future (predicted) distributions of vegetation properti es pertaining to fire fuels are necessary for understanding, monitorin g, managing and modeling fire in changing Mediterranean landscapes. Se asonal and interannual variability of some vegetation properties can b e extracted from multi-date remotely sensed data that may provide usef ul information pertaining to fire fuel distributions. Brightness (BR), Greenness (GR) and Wetness (WT) transform images were derived from tw o geometrically and radiometrically registered TM images acquired in S an Diego County, California, USA, near the beginning and end of the 19 86-1987 growing season. Seasonal difference images of these transforms were also generated. Transform and difference values were stratified by vegetation community type and stand age or slope aspect, using an e xisting raster geographic information system database. The BR and GR i mages provided unique spatial information pertaining to vegetation con ditions. Seasonal change in BR was highest for burn and meadow areas, while high values in seasonal change in GR corresponded to deciduous o ak woodlands, grasslands and vegetated north facing slopes. End-of-sea son GR for Mixed Chaparral varied with stand age in a manner which was similar to field measurements of total and live standing biomass. Dif ferential illumination effects seem to dominate spatial variations in BR and GR values and their seasonal change as a function of slope aspe ct.