Spatial Interpolation Map Temperature using ggplot in R

Question

I would like to create 3 maps of spatial interpolation of a country (like belgium) using the GADM geopkg data that can be found there : https://gadm.org/download_country.html.

The idea would be to create 3 interpolated maps :

Voronoi interpolation
IDW interpolation
Ordinary Kriging interpolation

I also would like to use the library ggplot2. My aim is to carry out a spatial interpolation with a few observations, but which are interplaced throughout the country.

Does anyone have any ideas for debugging my code and helping me to implement with IDW /kriging?

I am also attaching an example found on the Internet of what I am trying to reproduce as much as possible. I'm also trying to ensure that, as in the photo, the temperature scale is categorised by degree or half degree.

What i want to do

Here's my starting point of code:

This code tries to do the same thing with Voronoi, but for some reason the code can't access the temperature field in the geom_sf

library(tidyverse)
library(sf)
library(sp)
library(spatstat)
library(maptools)


belgium_gadm <- st_read("gadm41_BEL.gpkg", layer="ADM_ADM_0")

temps_data <- data.frame(
  city = c("Brussels", "Antwerp", "Ghent"), 
  lat = c(50.8503, 51.2194, 51.0543), 
  lon = c(4.3517, 4.4024, 3.7174), 
  temperature = c(15, 16, 17)
)

coordinates(temps_data) <- ~lon+lat

temp_ppp <- as.ppp(
  matrix(c(temps_data$lon, temps_data$lat), ncol = 2),
  W = as.owin(st_bbox(belgium_gadm))
)

voronoi_result <- dirichlet(temp_ppp)

voronoi_sf <- (as(voronoi_result, "SpatialPolygons"))


voronoi_sf$temperature <- over(voronoi_sf,temps_data[,2], fn = mean)$temperature


ggplot() +
  geom_sf(data = belgium_gadm, fill = NA, color = 'black') +
  geom_sf(data = voronoi_sf, aes(fill = temperature)) +
  coord_sf() +
  theme_minimal()

dieghernan · Accepted Answer · 2023-06-01 12:07:46Z

Issue here is that voronoi_sf <- (as(voronoi_result, "SpatialPolygons")) is not a sf object but a SpatialPolygon object. Convert it to sf with the proper CRS (EPSG:4326 I pressume since you are working with lon/lat coordinates):

library(tidyverse)
library(sf)
library(sp)
library(spatstat)
library(maptools)

belgium_gadm <- giscoR::gisco_get_countries(country = "BEL")

temps_data <- data.frame(
  city = c("Brussels", "Antwerp", "Ghent"), 
  lat = c(50.8503, 51.2194, 51.0543), 
  lon = c(4.3517, 4.4024, 3.7174), 
  temperature = c(15, 16, 17)
)

coordinates(temps_data) <- ~lon+lat

temp_ppp <- as.ppp(
  matrix(c(temps_data$lon, temps_data$lat), ncol = 2),
  W = as.owin(st_bbox(belgium_gadm))
)

voronoi_result <- dirichlet(temp_ppp)

voronoi_sf <- (as(voronoi_result, "SpatialPolygons"))


voronoi_sf$temperature <- over(voronoi_sf,temps_data[,2], fn = mean)$temperature

Add these lines

#######################
#### NEW CODE HERE ####
####################### 
# See here how to convert to sf
voronoi_sf <- st_as_sf(voronoi_sf)
# Need to assig crs
st_crs(voronoi_sf) <- st_crs(4326)

# Same for cities
temps_data_sf <- st_as_sf(temps_data)
st_crs(temps_data_sf)<- st_crs(4326)

# Final plot
ggplot() +
  geom_sf(data = belgium_gadm, fill = NA, color = 'black') +
  geom_sf(data = voronoi_sf, aes(fill = temperature), alpha = 0.5) +
  geom_sf(data = temps_data_sf) +
  theme_minimal()

^{Created on 2023-06-01 with reprex v2.0.2}

Collectives™ on Stack Overflow

Spatial Interpolation Map Temperature using ggplot in R

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related