--- title: "climateStability Vignette: the Basics" author: - Hannah Owens date: "`r Sys.Date()`" output: rmarkdown::html_vignette: fig_caption: yes toc: true toc_depth: 3 vignette: > %\VignetteIndexEntry{How to use climateStability} %\VignetteEngine{knitr::rmarkdown} %\VignetteEncoding{UTF-8} --- ```{r setup, include=FALSE} library(climateStability) library(terra) knitr::opts_chunk$set(error = TRUE) ``` # Introduction `climateStability` is a small set of `R`-based tools for generating climate stability estimates based on time slices of climate data. # Calculating Stability First, you give the `deviationThroughTime` function the location of a folder containing time slice data (as ASCII rasters) for a single variable, such as that produced by PaleoView (Fordham, *et al*. 2017, Ecography), and the time elapsed *between slices*. It will calculate the average standard deviation over time for each cell. Note that `deviationThroughTime` will use *all* the raster files in the given folder in order, so be careful that it only contains the files from which you want to calculate stability. You may want to also verify that `R` loads the files in the correct order. That is, verify that your files are named in such a way that they are in order when read in. The function accommodates datasets with even temporal intervals, as well as uneven intervals. First, here is an example with a dataset in which 1,000 years has elapsed between each raster layer. ```{r example of deviation through time with even time slices, eval=FALSE} # deviationThroughTime using even time slices precipDeviation <- deviationThroughTime(variableDirectory = "inst/extdata/precipfiles/", timeSlicePeriod = 1000) temperatureDeviation <- deviationThroughTime(variableDirectory = "inst/extdata/tempfiles/", timeSlicePeriod = 1000) ``` Here is an example with a dataset for which there are uneven time intervals between rasters (every 1,000 years between 1kybp to 5 kybp, and then rasters for 10kybp, 15kybp, and 21kybp). In this case, `timeSlicePeriod` can take a vector specifying the time elapsed between slices. NOTE: make sure the time interval vector is in the *same order* as your time slices. There should be one fewer `timeSlicePeriod` items than the number of climate slice files. ```{r example of deviation through time with uneven time slices, eval=FALSE} # deviationThroughTime using uneven time slices unevenPrecipDeviation <- deviationThroughTime(variableDirectory = "inst/extdata/precipfilesUneven/", timeSlicePeriod = c(1000, 1000, 1000, 1000, 5000, 5000, 6000)) unevenTemperatureDeviation <- deviationThroughTime(variableDirectory = "inst/extdata/tempfilesUneven/", timeSlicePeriod = c(1000, 1000, 1000, 1000, 5000, 5000, 6000)) ``` You have been provided with two pre-calculated deviation-through-time estimates from Owens and Guralnick. Submitted, Biodiversity Informatics. These were calculated from 100-year climate means for 20 1,000-year time slices from 21 kbp to 1 kbp using data from the TRaCE 21k climate experiment, run in the CCSM 3.0 climate model. These layers will be used to illustrate the remaining `climateStability` package functions. Let's load them into the `R` environment. ```{r load the results of deviation through time from the package} precipDeviation <- terra::rast(system.file("inst/extdata/precipDeviation.asc", package = "climateStability")) temperatureDeviation <- terra::rast(system.file("inst/extdata/temperatureDeviation.asc", package = "climateStability")) ``` The next step is scaling the deviation through time estimate from 0 to 1, and then take the inverse of the result. This how you estimate stability for a given climate variable, per Owens and Guralnick, 2019. Biodiversity Informatics. ```{r calculate stability from deviation} precipStability <- stabilityCalc(precipDeviation) tempStability <- stabilityCalc(temperatureDeviation) ``` Finally, to estimate climatic stability, multiply the individual climate variable stability estimates together. ```{r calculate climate stability from individual climate variable stability estimates} climateStability <- rescale0to1(precipStability * tempStability) ``` # Plots showing stability These data can be visualized either as rasters or as line graphs showing the relationship between latitude and stability. ## Stability rasters This is what climate the stability rasters look like. ```{r plot stability rasters, fig.height=4, fig.width=8} plot(precipStability, main = "Relative Precipitation Stability") plot(rangeBuilder::gshhs, add = T) plot(tempStability, main = "Relative Temperature Stability") plot(rangeBuilder::gshhs, add = T) plot(climateStability, main = "Overall Relative Climate Stability") plot(rangeBuilder::gshhs, add = T) ``` ## Stability line graphs The `climateStability` package comes with a useful tool for calculating mean stability per line of latitude, `latitudinalMean`. This can then be used to plot climate stability for a given variable as a response to latitude. ```{r plot latitudinal mean in stability, fig.height=4, fig.width=8} #Calculate mean values at rasters plot(latitudinalMean(precipStability), main = "Precipitation Stability by Latitude", ylab = "Relative Stability", type = "l") plot(latitudinalMean(tempStability), main = "Temperature Stability by Latitude", ylab = "Relative Stability", type = "l") plot(latitudinalMean(climateStability), main = "Climate Stability by Latitude", ylab = "Relative Stability", type = "l") ``` Additionally, `climateStability` can calculate mean stability versus the absolute value of the latitudinal mean using the function `absLatitudinalMean`. ```{r plot absolute latitudinal mean in stability, fig.height=4, fig.width=8} #Calculate mean values at rasters plot(absLatitudinalMean(precipStability), main = "Precipitation Stability by Absolute Latitude", ylab = "Relative Stability", type = "l") plot(absLatitudinalMean(tempStability), main = "Temperature Stability by Absolute Latitude", ylab = "Relative Stability", type = "l") plot(absLatitudinalMean(climateStability), main = "Climate Stability by Absolute Latitude", ylab = "Relative Stability", type = "l") ```