TerrAdapt:Cascadia Documentation
TerrAdapt:Cascadia
  • QUICK START GUIDE
  • HOW TO USE TERRADAPT:CASCADIA
    • Public License and Citation
    • Appropriate Uses and Key Limitations
    • TerrAdapt:Cascadia Map Portal
    • TerrAdapt:Cascadia Dashboard
  • DATA INPUTS
    • Data Inputs Overview
    • Remote Sensing Data
    • Climate
    • Energy and Transportation Infrastructure
    • Topography, Hydrology, & Soils
  • METHODS AND VALIDATION
    • Methods Overview
    • Landcover
      • Taxonomy
      • Training Data
      • Covariates
      • Model Development
      • Model Validation
    • Forest Structure
      • Training Data
      • Covariates
      • Model Development
      • Model Validation
    • Rangeland Fractional Cover
      • Training Data
      • Covariates
      • Model Development
      • Model Validation
    • Change Detection and Ecological Disturbance Modeling
      • Taxonomy
      • Covariates
      • Training Data
      • Model Development
    • Human Footprint
    • Habitat
      • Species Distribution Modeling
      • Ecosystem-based Models
      • Core Habitat
      • Habitat Centrality
    • Connectivity
      • Mapping Connectivity Networks
      • Corridors
      • Corridor Centrality
      • Mapping Barriers
Powered by GitBook
On this page

Was this helpful?

  1. METHODS AND VALIDATION
  2. Landcover

Model Development

PreviousCovariatesNextModel Validation

Last updated 1 year ago

Was this helpful?

We used the random forest classifier () implemented in Google Earth Engine (GEE) to predict each landcover class in our from the at all locations. We trained 10 replicate models per landcover class in a k-fold approach () with a unique 'fold' of 1/10th of the data withheld for as described above in the section. The model predicted the probability that the observation belonged to the class of interest rather than any other class.

In addition to assessing the probability that a pixel belonged to each landcover class in our taxonomy, we also created a discrete landcover classification based on the landcover class with the highest probability in each pixel. We modified this discrete classification by assigning all pixels that coincided with the road classes of 'freeway' or 'primary road' a class value of 19 to indicate the presence of a major road.

Breiman 2001
taxonomy
covariates
training data
Fushiki 2011
validation
training data