French broom (Genista monspessulana) modeling in the Marin Municipal Water District

We will model the spatial distribution and likely spread of French broom on the Mt. Tamalpais watershed using multivariate geostatistical, ordinary least square and geographically weighted regression techniques. The general approach will assemble a stack of environmental predictor layers (e.g. climate and soils) and to overlay species occurrence spatial data on these data layers to tabulate the pattern of environmental associations with the occurrences. A second set of random points will be overlaid with the data layers to provide a background of pseudo-absence data for statistical discrimination. These data are then imported into a statistical software environment for modeling, creating a classifier to predict the likelihood of a species occurrence based on specific environmental factors. This classifier is then applied to the GIS layers to produce a map that shows the likelihood that the species occurs in a specific geographic location.