Create heatmap to visualize the daily temperature in England over 100 years

heatmap_daily_temp_england.R

library(tidyverse)
library(lubridate)

# Dataset URL
url_eng_daily_max <- "https://www.metoffice.gov.uk/hadobs/hadcet/cetmaxdly1878on_urbadj4.dat"

# Read in data from URL
eng_daily_max <- data.table::fread(
  url_eng_daily_max,
  # Encode -999 as NA
  na.strings = "-999",
  fill = TRUE,
  # Column names
  col.names = c("year", "day", 1:12),
  data.table = FALSE)


# Reshape data to tidy format
eng_daily_max_tidy <- as_tibble(eng_daily_max) %>%
  gather(key = "month", value = "t_max_c", -year, -day) %>%
  mutate(
    # Original data was in tenths of a degree C. Transform to degree C
    t_max_c = t_max_c / 10,  
    # Get date of measurement
    date = str_c(year, month, day, sep = "-"),
    date = ymd(date),
    # Get date as day_in_year (1-366)
    year_day = yday(date)
    ) %>%
  # Remove invalid date caused by `gather`, e.g. day 31 in some months
  filter(!is.na(date))


# Create labels for use in graph
labels_x <- c("Jan-01","Feb-01","Mar-01","Apr-01","May-01","Jun-01",
              "Jul-01","Aug-01","Sep-01","Oct-01","Nov-01","Dec-01")
breaks_x <- c(1, 32, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335)

celcius <- scales::degree_format(unit = "C")
breaks_fill <- c(seq(-5, 35, by = 5))
breaks_fill_label <- celcius(breaks_fill)

# Heatmap
eng_daily_max_tidy %>% 
  ggplot(aes(x = year_day, y = year, fill = t_max_c)) +
  geom_tile() + 
  scale_fill_distiller(palette = "RdYlBu",
                       na.value = "transparent",
                       breaks = breaks_fill,
                       label = breaks_fill_label) +
  guides(fill = guide_colorbar(title = NULL,
                               reverse = FALSE)) +
  scale_y_reverse(breaks = c(seq(1880, 2010, by = 10), 2019),
                  expand = c(0, 0)) +
  scale_x_continuous(breaks = breaks_x,
                     labels = labels_x, 
                     expand = c(0, 0),
                     limits = c(1, 366)) +
  labs(title = "Maximum Daily Temperature in Central England",
       subtitle = "From 1878-01-01 to 2019-08-31",
       caption = glue::glue("
          Source: Hadley Centre Central England Temperature (HadCET) Dataset 
          Dataset used: Daily maximum HadCET, 1878 to date
          Retrieved on 2019-09-03
          "),
       x = "Day", y = "Year") +
  hrbrthemes::theme_ipsum_rc() +
  theme(legend.key.width = unit(10, "pt"),
        legend.key.height = unit(30, "pt"),
        legend.title = element_text(size = 10),
        axis.text.x = element_text(size = 10),
        axis.text.y = element_text(size = 10)) 

# ggsave("england_daily_max.png", width = 8, height = 6)