rOpenSci package or resource used*
Given the recent heatwave, I was curious about plotting up a timeseries of the local Environment Canada weather stations. For this, I used the weathercan package (GitHub - ropensci/weathercan: R package for downloading weather data from Environment and Climate Change Canada)
I’ll walk through the few lines of code that it takes to search, download, and plot the data:
First, load a couple libraries:
library(dplyr)
library(ggplot2)
library(weathercan)
library(ggrepel)
Then, search the station IDs that have data for this calendar year and match my city name.
stn <- stations_search("Prince George", interval = "hour", ends_earliest = "2021")
ids <- stn$station_id
Download the data using the station IDs.
df <- weather_dl(station_ids = ids, start ="2021-01-01", end = "2021-07-01")
Optional: Add elevation to the station name.
df <- df %>%
mutate(station_name = paste0(station_name, " (", elev, " m)"))
Per station, calculate the min/max
df_minmax <- df %>%
group_by(station_name) %>%
filter(temp == min(temp, na.rm = T) | temp == max(temp, na.rm = T))
Plot it up!
df %>%
ggplot(aes(time, temp)) +
geom_line(, color = "grey50") +
geom_point(data = df_minmax) +
geom_text_repel(data = df_minmax, aes(label = paste(temp, time))) +
labs(x = "", y = "°C") +
theme_bw() +
theme(aspect.ratio = 0.6) +
facet_wrap(~station_name, ncol = 3)
