phenomap

This takes in a series of multi-layer raster files and returns a phenology projection raster. phenomap is capable of analyzing satellite-derived NDVI and snowmelt time series on a regional or global scale. Currently 'phenomap' is optimized for data that derives from the MOD 13 and MOD 10 data products, but it is expected that this project will eventually expand to also support GIMMS AVHRR and perhaps Landsat datasets.

Motivation

Though landscape phenology is increasingly becoming a focal point of investigations of migration timing, hitherto no R packages existed that were able to reconstruct satellite-derived phenological metrics in space. Thus, it was my motivation to develop this package en passant with John 2016 (M.S. thesis in ecology) in order to enable a broader group of researchers to project landscape timing measures in space.

Prerequisistes

phenomap requires at least R version 4.1.0, and has been tested on Mac OSX Monterey v12.3.1 and Windows 7.

This package uses methods from the plyr and stringr, and most importantly phenex and terra packages. I recommend loading the terra package so that the product of mapPheno() can then be inspected and further processed as needed.

phenomap does not support .hdf files, and therefore it is recommended that users download and mosaic data. I find that pyModis is a reliable tool for such pursuits.

Warning

phenomap requires a considerable amount of computing power and even for small datasets it may take a while to process the data. I therefore recommend cropping datasets to the minimum possible extent and projecting to the coarsest possible resolution. At this point phenomap does not support parallel processing, but it is expected to in future versions.

Installation

Use devtools to install this package...

devtools::install_github("JepsonNomad/phenomap")

... until it is available on CRAN, at which point install phenomap using

install.packages(phenomap)

Code example

library(phenomap)
File_List.VI <- list.files(pattern="TeenyCrop") # these files can be found in the /inst/extdata folder
PhenoFactor = "VI"
threshold = 0.5
year = 2016
NDVI = 1
VIQ = 3
DOY = 4
PR = 5
verbose = TRUE

To project the timing of 50% GWI, use mapPheno with phase = "greenup", and threshold = 0.5.

Sample.Greenup <- mapPheno(File_List = File_List.VI, PhenoFactor = PhenoFactor,
                           phase = "greenup", threshold = threshold, year = year,
                           NDVI = NDVI, VIQ = VIQ, DOY = DOY, PR = PR,
                           verbose = verbose)
plot(Sample.Greenup)

To project the timing of senescence, use phase = "senescence":

Sample.Senescence <- mapPheno(File_List = File_List.VI, PhenoFactor = PhenoFactor,
                           phase = "senescence", threshold = threshold, year = year,
                           NDVI = NDVI, VIQ = VIQ, DOY = DOY, PR = PR,
                           verbose = verbose)
plot(Sample.Senescence)

To inspect broad measures of timing, compare histograms:

breaks<-seq(from=0,to=365,by=5)
hist(as.matrix(Sample.Senescence), breaks=breaks, ylim=c(0,150),
     col="ORANGE", main = "Snowmelt, Greenup, and Senescence",
     xlab="Day of Year")
hist(as.matrix(Sample.Greenup), breaks=breaks,
     col="palegreen", add=T)

To project the duration of the growing season, compare greenup and senescence rasters:

Growing.Season <- as.matrix(Sample.Senescence) - as.matrix(Sample.Greenup)
GS.raster <- rast(Growing.Season)
plot(GS.raster)

License

This project is licensed under the GPL-3 license.

JepsonNomad / phenomap