An accompanying code and data repository for the paper Pasquarella, V.J., Mickley, J.G., Barker Plotkin, A., MacLean, R. G., Anderson, R. M., Brown, L. M., Wagner, D. L., Singer, M. S., & Bagchi, R. (2021). Predicting defoliator abundance and defoliation measurements using Landsat-based condition scores. Remote Sensing in Ecology and Conservation. https://doi.org/10.1002/rse2.211
Here, we compare the relative predictive ability of various parameterizations of remote-sensed forest condition assessments on abundance and damage from a Lymantria dispar outbreak in southern New England.
- Data files should be stored inside the data directory.
- Scripts for Earth Engine workflow are stored in the earthengine_workflow directory and the repository can be accessed on Earth Engine here.
- Analyses should be stored in the analyses directory.
- Use the analysis template in analyses/ as a starting point.
- Defoliation Models - Comparison of defoliation models for forest condition assessment
- Using AIC weights to pick a winner - Can AIC model weights be used to pick a consensus best model parameterization?
- Results can be viewed interactively via an Earth Engine App
- Mapped results are available as a Zenodo dataset