HadEX3 climate extremes: India

library(varunayan)
library(dplyr)
library(ggplot2)
library(tidyr)

The HadEX3 dataset from the Met Office Hadley Centre provides global gridded climate extremes based on 27 ETCCDI indices, at 1.25 x 1.875 degree resolution over 1901-2018.

Available indices and their descriptions:

list_hadex3_indices() |>
  select(index, category, description, unit, annual, monthly) |>
  head(12)

Download

# Annual TXx (max of daily max temp) for India
india_txx <- hadex3_bbox(
  index = "TXx",
  start_year = 1951, end_year = 2018,
  north = 37, south = 6, east = 98, west = 68,
  frequency = "annual"
)

# Annual Rx1day (max 1-day rainfall) for India
india_rx1day <- hadex3_bbox(
  index = "Rx1day",
  start_year = 1951, end_year = 2018,
  north = 37, south = 6, east = 98, west = 68,
  frequency = "annual"
)

# Monthly TXx for seasonal analysis
india_txx_mon <- hadex3_bbox(
  index = "TXx",
  start_year = 1981, end_year = 2018,
  north = 37, south = 6, east = 98, west = 68,
  frequency = "monthly"
)

data_path <- system.file("extdata", "india_hadex3.rds", package = "varunayan")
if (!nzchar(data_path)) {
  knitr::opts_chunk$set(eval = FALSE)
} else {
  saved <- readRDS(data_path)
  india_txx <- saved$TXx_annual
  india_rx1day <- saved$Rx1day_annual
  india_txx_mon <- saved$TXx_monthly
}

Annual temperature extremes trend

We take the spatial mean of annual TXx (the hottest day) across India’s grid cells and add a loess trend.

txx_annual <- india_txx |>
  group_by(year) |>
  summarise(mean_txx = mean(value, na.rm = TRUE), .groups = "drop")

ggplot(txx_annual, aes(x = year, y = mean_txx)) +
  geom_line(color = "#BDBDBD", linewidth = 0.6) +
  geom_smooth(
    method = "loess", span = 0.3, se = TRUE,
    color = "#B2182B", fill = "#B2182B", alpha = 0.15, linewidth = 1.1
  ) +
  labs(
    x = NULL, y = "TXx (\u00b0C)",
    title = "India: Annual Maximum Temperature (TXx), 1951-2018",
    subtitle = "Spatial mean across grid cells; line = loess trend"
  ) +
  theme_minimal(base_size = 11) +
  theme(panel.grid.minor = element_blank())

Seasonal patterns from monthly data

Averaging monthly TXx over 1981-2018, we see the typical pre-monsoon peak in May-June and the cooler post-monsoon months.

seasonal <- india_txx_mon |>
  group_by(month) |>
  summarise(mean_txx = mean(value, na.rm = TRUE), .groups = "drop") |>
  mutate(month_label = factor(month.abb[month], levels = month.abb))

ggplot(seasonal, aes(x = month_label, y = mean_txx, group = 1)) +
  geom_line(color = "#B2182B", linewidth = 1.2) +
  geom_point(color = "#B2182B", size = 2.5) +
  labs(
    x = NULL, y = "Mean TXx (\u00b0C)",
    title = "India: Seasonal Cycle of TXx (1981-2018 mean)"
  ) +
  theme_minimal(base_size = 11) +
  theme(panel.grid.minor = element_blank())

Extreme rainfall trend (Rx1day)

Annual maximum 1-day precipitation (Rx1day) over India 1951-2018.

rx1_annual <- india_rx1day |>
  group_by(year) |>
  summarise(mean_rx1 = mean(value, na.rm = TRUE), .groups = "drop")

ggplot(rx1_annual, aes(x = year, y = mean_rx1)) +
  geom_line(color = "#BDBDBD", linewidth = 0.6) +
  geom_smooth(
    method = "loess", span = 0.35, se = TRUE,
    color = "#2166AC", fill = "#2166AC", alpha = 0.15, linewidth = 1.1
  ) +
  labs(
    x = NULL, y = "Rx1day (mm)",
    title = "India: Annual Maximum 1-Day Rainfall (Rx1day), 1951-2018",
    subtitle = "Spatial mean; line = loess trend"
  ) +
  theme_minimal(base_size = 11) +
  theme(panel.grid.minor = element_blank())

Decadal spatial distribution of TXx

We compare the spatial pattern of TXx across three decades.

decades <- india_txx |>
  mutate(decade = cut(year,
    breaks = c(1950, 1970, 1990, 2018),
    labels = c("1951-1970", "1971-1990", "1991-2018"),
    include.lowest = TRUE
  )) |>
  group_by(decade, latitude, longitude) |>
  summarise(mean_txx = mean(value, na.rm = TRUE), .groups = "drop")

ggplot(decades, aes(x = longitude, y = latitude, fill = mean_txx)) +
  geom_tile() +
  scale_fill_distiller(palette = "RdYlBu", direction = -1, name = "\u00b0C") +
  coord_fixed(ratio = 1) +
  facet_wrap(~decade, nrow = 1) +
  labs(
    x = "Longitude", y = "Latitude",
    title = "India: Mean Annual TXx by Decade"
  ) +
  theme_minimal(base_size = 10) +
  theme(panel.grid = element_blank())