Analysing the climate at spatial points for a given period
Source:vignettes/points-df-mat-sf.Rmd
points-df-mat-sf.Rmd## Warning: package 'knitr' was built under R version 4.3.2With {easyclimate} you can easily download daily, monthly and annual climate data for a given set of points or polygons within Europe. To download and install the latest version of {easyclimate} from GitHub follow the instructions in https://github.com/VeruGHub/easyclimate
In this tutorial we will work through the basics of using easyclimate
with coordinate points. You can enter coordinates as a
data.frame, matrix, sf or
SpatVector object. At the end we will have a data frame
with climate variables for each point.
Example 1: Introducing coordinates as a data frame
First, specify longitude and latitude coordinates in a data frame
with the column names lon and lat. Here we are
simulating coordinates for three random sites in southern Spain:
library(easyclimate)
library(ggplot2)
library(dplyr)
coords <- data.frame(
lon = rnorm(3, mean = -5.36, sd = 0.3),
lat = rnorm(3, mean = 37.40, sd = 0.3)
)
ggplot() +
borders(regions = c("Spain", "Portugal", "France")) +
geom_point(data = coords, aes(x = lon, y = lat)) +
coord_fixed(xlim = c(-10, 2), ylim = c(36, 44), ratio = 1.3) +
xlab("Longitude") +
ylab("Latitude") +
theme_bw()
Now, download the climatic data for the selected locations. To get
daily data, all you have to do is use the function
get_daily_climate, specifying the period
(e.g. 2008-05-25 for a single day or 2008:2010
for several years), and the variables to be downloaded (precipitation
Prcp, minimum temperature Tmin or maximum
temperature Tmax).You can also use the function
get_monthly_climateand get_annual_climateto
download monthly and annual data, specifying the period
(e.g. 2008-05for a single month, 2008 for a
single year, or 2008:2010 for several years):
Sys.time() # to know how much time it takes to download
## [1] "2026-07-24 16:12:12 CEST"
daily <- get_daily_climate( # daily data
coords = coords,
period = 2008:2010,
climatic_var = c("Prcp", "Tmin", "Tmax")
)
Sys.time()
## [1] "2026-07-24 16:15:23 CEST"
kable(head(daily))| ID_coords | lon | lat | date | Prcp | Tmin | Tmax |
|---|---|---|---|---|---|---|
| 1 | -5.265025 | 37.46851 | 2008-01-01 | 0.15 | 2.98 | 13.32 |
| 1 | -5.265025 | 37.46851 | 2008-01-02 | 19.97 | 6.45 | 14.29 |
| 1 | -5.265025 | 37.46851 | 2008-01-03 | 10.06 | 6.22 | 13.84 |
| 1 | -5.265025 | 37.46851 | 2008-01-04 | 0.00 | 5.09 | 12.18 |
| 1 | -5.265025 | 37.46851 | 2008-01-05 | 0.00 | 4.36 | 13.67 |
| 1 | -5.265025 | 37.46851 | 2008-01-06 | 0.00 | 9.85 | 14.58 |
Sys.time()
## [1] "2026-07-24 16:15:23 CEST"
monthly <- get_monthly_climate( # monthly data
coords = coords,
period = 2008:2010,
climatic_var = c("Prcp", "Tmin", "Tavg", "Tmax")
)
Sys.time()
## [1] "2026-07-24 16:15:38 CEST"
kable(head(monthly))| ID_coords | lon | lat | date | Prcp | Tmin | Tavg | Tmax |
|---|---|---|---|---|---|---|---|
| 1 | -5.265025 | 37.46851 | 2008-01 | 44.10 | 5.08 | 10.79 | 16.50 |
| 1 | -5.265025 | 37.46851 | 2008-02 | 42.97 | 7.65 | 13.07 | 18.49 |
| 1 | -5.265025 | 37.46851 | 2008-03 | 8.98 | 5.63 | 13.16 | 20.70 |
| 1 | -5.265025 | 37.46851 | 2008-04 | 166.02 | 9.14 | 16.06 | 22.98 |
| 1 | -5.265025 | 37.46851 | 2008-05 | 52.93 | 12.04 | 18.21 | 24.38 |
| 1 | -5.265025 | 37.46851 | 2008-06 | 0.00 | 16.29 | 24.80 | 33.32 |
annual <- get_annual_climate( # annual data
coords = coords,
period = 2008:2010,
climatic_var = c("Prcp", "Tmin", "Tavg", "Tmax")
)
Sys.time()
## [1] "2026-07-24 16:15:46 CEST"
kable(head(annual))| ID_coords | lon | lat | date | Prcp | Tmin | Tavg | Tmax |
|---|---|---|---|---|---|---|---|
| 1 | -5.265025 | 37.46851 | 2008 | 513.43 | 11.09 | 17.73 | 24.36 |
| 1 | -5.265025 | 37.46851 | 2009 | 639.27 | 12.35 | 18.82 | 25.28 |
| 1 | -5.265025 | 37.46851 | 2010 | 1010.98 | 12.59 | 18.31 | 24.02 |
| 2 | -5.602953 | 36.94713 | 2008 | 705.04 | 11.40 | 17.62 | 23.84 |
| 2 | -5.602953 | 36.94713 | 2009 | 739.14 | 12.13 | 18.53 | 24.92 |
| 2 | -5.602953 | 36.94713 | 2010 | 1021.30 | 12.68 | 18.42 | 24.15 |
Here we extract different components of the date:
library(lubridate)
daily <- daily |>
mutate(
date = as.Date(date),
month = months(date),
year = format(date, format = "%y")
)
monthly <- monthly |>
mutate(date = ym(date))Finally, you can visualize the daily, monthly and annual climate results. For example, let’s plot the precipitation for one of the sites:
clim_daily_site1 <- daily |>
filter(ID_coords == 1)
ggplot(clim_daily_site1) +
geom_line(aes(x = date, y = Prcp), colour = "steelblue") +
labs(x = "Date", y = "Daily precipitation (mm)") +
theme_bw()
clim_monthly_site1 <- monthly |>
filter(ID_coords == 1)
ggplot(clim_monthly_site1) +
geom_line(aes(x = date, y = Prcp), colour = "steelblue") +
labs(x = "Date", y = "Monthly precipitation (mm)") +
theme_bw()
clim_annual_site1 <- annual |>
filter(ID_coords == 1)
ggplot(clim_annual_site1) +
geom_line(aes(x = date, y = Prcp), colour = "steelblue") +
labs(x = "Date", y = "Annual precipitation (mm)") +
scale_x_continuous(breaks = scales::breaks_width(1)) +
theme_bw()
Or calculate the daily mean temperature and plot it against
tmin and tmax:
library(tidyr)
temp_long <- daily |>
mutate(Tmean = (Tmin + Tmax) / 2) |>
pivot_longer(
cols = c("Tmin", "Tmax", "Tmean"),
names_to = "temp_vars",
values_to = "temp_values")
ggplot(temp_long, aes(x = factor(ID_coords), y = temp_values,
fill = temp_vars, color = temp_vars)) +
geom_violin(size = 1, alpha = .7) +
scale_fill_manual(values = c("#C54A52", "#FAFBC5", "#4B8AB8")) +
scale_color_manual(values = c("#C54A52", "#FAFBC5", "#4B8AB8")) +
ylab("Temperature (ºC)") + xlab("") +
theme_bw()
ggplot(temp_long, aes(x = date, y = temp_values, color = temp_vars)) +
geom_point(alpha = .3) +
scale_color_manual(name = "Variables",
values = c("#C54A52", "#FAFBC5", "#4B8AB8"),
guide = guide_legend(override.aes = list(alpha = 1))) +
ylab("Temperature (ºC)") + xlab("Date") +
theme_bw()
Example 2: Introducing coordinates as a matrix
{easyclimate} handles different input data, try now with matrices!
Here we are retrieving daily precipitation data for a single year (2008):
coords_mat <- as.matrix(coords)
Sys.time()
## [1] "2026-07-24 16:15:49 CEST"
mat_prcp <- get_daily_climate(
coords = coords_mat,
period = 2008, # single year
climatic_var = "Prcp"
)
Sys.time()
## [1] "2026-07-24 16:15:56 CEST"
kable(head(mat_prcp))| ID_coords | lon | lat | date | Prcp |
|---|---|---|---|---|
| 1 | -5.265025 | 37.46851 | 2008-01-01 | 0.15 |
| 1 | -5.265025 | 37.46851 | 2008-01-02 | 19.97 |
| 1 | -5.265025 | 37.46851 | 2008-01-03 | 10.06 |
| 1 | -5.265025 | 37.46851 | 2008-01-04 | 0.00 |
| 1 | -5.265025 | 37.46851 | 2008-01-05 | 0.00 |
| 1 | -5.265025 | 37.46851 | 2008-01-06 | 0.00 |
mat_prcp <- mat_prcp |>
mutate(
date = as.Date(date),
month = months(date),
year = format(date, format = "%y")
) |>
relocate(lon, lat, date, year, month, Prcp)
ggplot(mat_prcp, aes(x = date, y = Prcp, color = Prcp)) +
geom_point() +
scale_color_continuous(name = "Precipitation (mm)") +
scale_y_continuous(expand = c(0, 0)) +
coord_cartesian(ylim = c(0, 60)) +
ylab("Daily precipitation (mm)") +
xlab("Date") +
theme_bw()
month_name <- format(ISOdate(2021, 1:12, 1), "%B")
mat_prcp |>
mutate(month = factor(month, rev(month_name))) |>
ggplot(aes(x = month, y = Prcp, color = month)) +
geom_jitter(size = 2, alpha = .3, width = .3, show.legend = FALSE) +
scale_y_continuous(expand = c(0, 0)) +
coord_flip(ylim = c(0, 60)) +
ylab("Daily precipitation (mm)") + xlab("Month") +
theme_bw()
We can also retrieved monthly data using the same matrix of coordinates and add the average values to the previous plot:
Sys.time()
## [1] "2026-07-24 16:15:56 CEST"
mat_monthly_prcp <- get_monthly_climate(
coords = coords_mat,
period = 2008, # single year
climatic_var = "Prcp"
)
Sys.time()
## [1] "2026-07-24 16:15:57 CEST"
kable(head(mat_monthly_prcp))| ID_coords | lon | lat | date | Prcp |
|---|---|---|---|---|
| 1 | -5.265025 | 37.46851 | 2008-01 | 44.10 |
| 1 | -5.265025 | 37.46851 | 2008-02 | 42.97 |
| 1 | -5.265025 | 37.46851 | 2008-03 | 8.98 |
| 1 | -5.265025 | 37.46851 | 2008-04 | 166.02 |
| 1 | -5.265025 | 37.46851 | 2008-05 | 52.93 |
| 1 | -5.265025 | 37.46851 | 2008-06 | 0.00 |
mat_monthly_prcp <- mat_monthly_prcp |>
mutate(
date = ym(date),
month = month(date, label = TRUE),
year = year(date)
) |>
relocate(lon, lat, year, month, Prcp)
mat_monthly_prcp |>
ggplot(aes(x = month, y = Prcp, color = month)) +
geom_jitter(size = 2, alpha = .3, width = .3, show.legend = FALSE) +
geom_point(
aes(x = month, y = Prcp),
size = 2
) +
scale_y_continuous(expand = c(0, 0)) +
coord_flip(ylim = c(0, 60)) +
ylab("Monthly precipitation (mm)") + xlab("Month") +
theme_bw() +
theme(legend.position = "none")
Example 3: Introducing coordinates as simple feature objects
Here we introduce coordinates as a sf object, and
retrieve minimum temperature for a single day (1 January 2001) and mean
temperature of a single year (2001):
library(sf)
coords_sf <- st_as_sf(
coords,
coords = c("lon", "lat")
)
sf_tmin <- get_daily_climate(
coords = coords_sf,
period = "2001-01-01", # single day
climatic_var = "Tmin"
)
ggplot() +
borders(regions = c("Spain", "Portugal", "France")) +
geom_point(data = sf_tmin, aes(x = lon, y = lat, color = Tmin), size = 2) +
coord_fixed(xlim = c(-10, 2), ylim = c(36, 44), ratio = 1.3) +
scale_color_gradient2(name = "Minimum\ntemperature (ºC)",
low = "#4B8AB8", mid = "#FAFBC5", high = "#C54A52",
midpoint = mean(sf_tmin$Tmin)) +
ylab("Latitude") +
xlab("Longitude") +
theme_bw()
sf_tavg <- get_annual_climate(
coords = coords_sf,
period = 2001, # single year
climatic_var = "Tavg"
)
ggplot() +
borders(regions = c("Spain", "Portugal", "France")) +
geom_point(data = sf_tavg, aes(x = lon, y = lat, color = Tavg), size = 2) +
coord_fixed(xlim = c(-10, 2), ylim = c(36, 44), ratio = 1.3) +
scale_color_gradient2(name = "Mean\ntemperature (ºC)",
low = "#4B8AB8", mid = "#FAFBC5", high = "#C54A52",
midpoint = mean(sf_tavg$Tavg)) +
ylab("Latitude") +
xlab("Longitude") +
theme_bw()
Learn more
Now you know how to obtain a data frame with different climatic variables with {easyclimate}, using point coordinates in different formats and downloading data for multiple locations and periods. Check out this other vignette if you need to extract the data of a complete area.