Climate Stripes

Climate Stripes#

Climate stripes (also known as “warming stripes”) are a minimalistic visualization popularized by Ed Hawkins that encodes the long-term trend of a climate variable as a sequence of coloured bars — one bar per year, coloured from blue (cooler / drier) to red (warmer / wetter).

Wetterdienst can generate climate stripes from DWD daily observations for two kinds:

  • temperature — based on the annual mean air temperature

  • precipitation — based on the annual precipitation sum

The feature is exposed via the command line interface and the REST API. The hosted web frontend also offers an interactive Climate Stripes view.

Command Line Interface#

The stripes command group has two subcommands: stations and values.

Stations#

List the stations for which climate stripes can be generated:

# Stations available for temperature stripes.
wetterdienst stripes stations --kind temperature

# Only currently active stations, as CSV.
wetterdienst stripes stations --kind precipitation --active true --format csv

Options:

option

description

default

--kind

temperature or precipitation (required)

—

--active

only include stations that are still reporting

true

--format

output format, one of json, geojson, csv

json

--pretty

pretty-print the output

false

Values#

Render the stripes image for a single station, selected either by --station (station id) or by --name (fuzzy name match):

# Temperature stripes for station 1048 (Dresden-Klotzsche) written to a PNG file.
wetterdienst stripes values --kind temperature --station 1048 --target dresden.png

# Precipitation stripes selected by name, restricted to a year range, as SVG to stdout.
wetterdienst stripes values --kind precipitation --name Hohenpeissenberg \
  --start_year 1900 --end_year 2020 --format svg

Options:

option

description

default

--kind

temperature or precipitation (required)

—

--station

station id to plot

—

--name

station name to plot (fuzzy match)

—

--start_year

first year to include

—

--end_year

last year to include

—

--name_threshold

similarity threshold (0–1) for --name matching

0.80

--show_title

draw the station name as a title

true

--show_years

draw the first and last year as axis labels

true

--show_data_availability

draw a marker for the data availability

true

--format

image format, one of png, jpg, svg, pdf

png

--dpi

image resolution in dots per inch

300

--target

write the image to this file instead of stdout

—

REST API#

When the REST API is running, the same functionality is available via three endpoints (the examples below use httpie):

# List stations available for temperature stripes.
http localhost:7890/api/stripes/stations kind==temperature

# Get the underlying yearly values for a station.
http localhost:7890/api/stripes/values kind==temperature station==1048

# Render the stripes image for a station (write the binary body to a file).
http --download localhost:7890/api/stripes/image kind==temperature station==1048

The image endpoint accepts the same rendering options as the CLI (start_year, end_year, show_title, show_years, show_data_availability, format, dpi).