# Model Development

We used the random forest classifier ([Breiman 2001](https://link.springer.com/article/10.1023/a:1010933404324)) implemented in Google Earth Engine (GEE) to predict tree height (m), tree diameter (cm), and canopy cover (%) from the [covariates](https://terradapt.gitbook.io/terradapt-cascadia-documentation/methods-and-validation/forest-structure/covariates) at all [training data](https://terradapt.gitbook.io/terradapt-cascadia-documentation/methods-and-validation/forest-structure/training-data) locations. We trained 10 replicate models in regression mode per variable in a k-fold approach ([Fushiki 2011](https://link.springer.com/article/10.1007/s11222-009-9153-8)) with a unique 'fold' of 1/10th of the data withheld for [validation](https://terradapt.gitbook.io/terradapt-cascadia-documentation/methods-and-validation/forest-structure/model-validation). The final model was calculated as the average of the 10 replicates.