test_clim_prebaked.R

library(ggplot2)
library(raster)
library(rnaturalearth)
library(terra)
library(tidyterra)
library(patchwork)
library(sf)
library(targets)
theme_set(theme_void())


# Future
url_future <- "https://dd.weather.gc.ca/climate/cmip5/netcdf/scenarios/RCP2.6/monthly_ens/absolute/CMIP5_rcp2.6_monthly_abs_latlon1x1_TEMP_pctl50_P1M.nc" # nolint

# Historical
url_hist <- "https://dd.weather.gc.ca/climate/cmip5/netcdf/historical/monthly_ens/absolute/CMIP5_hist_monthly_abs_latlon1x1_TEMP_pctl50_P1M.nc" # nolint

download.file(url_future, dest = "future.ncdf", method = "wget")
download.file(url_hist, dest = "hist.ncdf", method = "wget")

# Make raster objects
tas_future <- rast("future.ncdf")
tas_hist <- rast("hist.ncdf")


# Load ON Census regions
on <- tar_read(raw_geom_data_on) %>%
  st_as_sf() %>%
  st_transform(crs = crs(tas_hist))

dundas <- on %>% filter(Region.Name == "South Dundas (MU)")

# Crop tas_future around dundas
tas_future_dundas <- crop(tas_future, dundas)

# Plot
ggplot() +
  geom_spatraster(data = tas_hist[[2]]) +
  scale_fill_whitebox_c(
    palette = "muted",
    labels = scales::label_number(suffix = "º")
  )