Time-lapse camera network reveals intra- and inter-species variation in green-up timing across the range of Sierra Nevada bighorn sheep
Code and data to produce results presented in John et al. 2023, Remote Sensing in Ecology and Conservation. Please cite our article (will update throughout production process):
- John, C., Kerby, J., Stephenson, T.R., and Post, E. In press. Fine-scale landscape phenology revealed through time-lapse imagery: implications for conservation and management of an endangered migratory herbivore. Remote Sensing in Ecology and Conservation.
Basic folder structure should match this repository. Using RStudio Project, working directory and relative filepaths should be automatically assigned.
Follow the contents of the scripts/ folder in numerical order. Descriptions follow:
- 000_modisFunctions.R: Curve fitting functions for MODIS data
- 001_cameraFunctions.R: Curve fitting functions for time-lapse GCC and snow cover data
- 101_GEE_tidy.R: Convert wide-format Earth Engine NDVI output to long-format data.frame
- 102_modPhenology.R: Derive phenology indices for MODIS data
- 201_organizeGCC.R: Compile and clean raw greenness data
- 202_camPhenology.R: Derive phenology indices for time-lapse camera data
- 301_networkSummary.R: Plot mean camera-level time series (Fig 2)
- 302_stanModels.R: Fit brms models and plot elevational lapse rates (Fig 3-6)
- 303_aerialCoverage.R: Measure temporal variation in availability of space at peak-greenup (Fig 7)
Be sure to uncompress the exif.zip and extract.zip subdirectories in the data/ folder. Note that NDVI (MODIS), elevation (DEP) and imagery (NAIP) were accessed via Google Earth Engine; code for reconstructing these data are available in their corresponding data/ folder but are required only for generating maps; wrangled data to reproduce analyses are in data/.