Visualising site locations using GeoJSON data with RStudio

GeoJSON data can be parsed using R to create a plot that visualises the geographic range of your TERN EcoPlots dataset. The GeoJSON data that is included in TERN data package downloads can be parsed immediately to create a plot that visualises the geographic range of your TERN EcoPlots dataset. Below is an example of such a basic plot.

Below is the script used to created the above plot. It contains code comments to give a basic explanation of each element of the plot creation process.

Setup

Before running the script in R or RStudio, there are a number of setup steps needed.

Download some TERN GeoJSON data. This example uses a dataset acquired from data search results in https://ternaus.atlassian.net/wiki/spaces/TERNSup/pages/2677768320/Finding+Data#Example-Two---Plants-in-Western-Australia.
Set your R working directory to an appropriate location on your device.
Copy your GeoJSON file to your R working directory and rename if desired.
Install the R packages that are called in the first code chunk.

# R package libraries required
library(geojsonio)
library(tidyverse)
library(stringr)
library(ggplot2)
library(sf)
library(rnaturalearth)
library(rnaturalearthdata)

# read in data from the GeoJSON file
plants <- geojson_read("geojson_PlantsInWA.json")

# some general exploration of our data 
View(plants)

summary(plants)
str(plants$features)


# Visualisation preparation tasks
# Transform GoeJSON data into a dataframe named plant_polys  
plant_polys <- 
  lapply(1:length(plants$features), 
         function(i){
           tmpdata <- plants$features[[i]]$geometry$coordinates[[1]][1] %>%
             data.frame()  %>% 
             rename('longitude'='.')  %>% 
             mutate(latitude= plants$features[[i]]$geometry$coordinates[[2]][1])  %>% 
             tibble %>%
             mutate(siteName = plants$features[[i]]$properties$siteName)
         }) %>%  bind_rows() 

# Have a look at how the dataframe is structured
str(plant_polys)

# Specifically every row
print(plant_polys, n=40) 

# Some site names contain a comma then further information.
# This will make the plot site names look a bit messy.
# OPTIONAL - truncate title to look neater on the plot
shortSiteName <- sapply(str_split(plant_polys$siteName,",",),'[',1)


# import world country polygons from the rnaturalearth package
world <- ne_countries(scale = "large", returnclass = "sf")

# Create the visualisation
# Draw a plot from the dataframe 
# using the world country polygons as the background
ggplot(data = world) +
  geom_sf() +
  geom_point(data = plant_polys, aes(x = longitude, y = latitude), size = 1, 
             shape = 23, fill = "darkgreen") +
  coord_sf(xlim=c(114,124), ylim=c(-36,-29), expand = FALSE) +
  labs(x = 'Longitude', y = 'Latitude', 
       title = "TERN Plots with identified plant species data in Western Australia") +
  annotate("label", 
           x = plant_polys$longitude + 0.7, 
           y = plant_polys$latitude + 0.25, 
           label = shortSiteName, 
           size = 3.5,
           color = "darkgreen"
  )

The final step is to export your plot image using the “Export” option from the plot viewer or the Plots menu.

This is a basic example of one of the many ways that GeoJSON data can be visualised using R. We recommend visiting sites such as The R Graph Gallery for ideas on ways to customise the aesthetics of your plot.

Setup

Aggregating and visualising biomass data example

0 Comments