MeteoSwiss-Forecast

Script to fetch the MeteoSwiss Weather Forecast data and generate a graph out of it.

The graph contains the rain and temperature forecast. later one is enhanced with the uncertainty band. Additionally, a blue marker indicates the last forecast model simulation.

The graph is highly configurable, how ever in the default configuration it tries to adapt the style of the MeteoSwiss App.

Default

python3 meteoswissForecast.py -z 8001 -f myForecast.png -m meta.json

Enhancements

Additionally, you can switch to a dark mode, change the time divisions, select the number of days and even mark the min/max temperature per day.

Example 1

python3 meteoswissForecast.py -z 6986 -f myForecast.png -m meta.json --days-to-show 3 --width 800 --symbol-division 2 --min-max-temperature

Example 2

python3 meteoswissForecast.py -z 8001 -f myForecast.png -m meta.json --days-to-show 3 --time-division 2 --width 1200 --height 300 --min-max-temperature --dark-mode --locale de_DE.utf8 --date-format="%A, %-d. %b" --time-format="%H"

Marking of Current time

The repo contains an extra script to add a mark of the current time (red bar). One might want to update this every minute or so. Since the generation of the forecast is rater slow, one might want to only update the current time mark at a high frequency but only generate the forecast once every hour.

python3 markGraphic.py -i myForecast.png -o myForecast-marked.png -x 52 -y 50 -w 295 -H 161

Preparations

Requirements

Run python3 -m pip install -r requirements.txt to install all needed Python packages.

Get Symbols

MeteoSwiss provides the symbols as SVG files. How ever we need them as PNG files, preferably with a transparent background. The folder symbols provides the already converted symbols. Alternatively you can generate them yourself:

BASE_URL="https://www.meteoschweiz.admin.ch/etc.clientlibs/internet/clientlibs/meteoswiss/resources/assets/images/icons/meteo/weather-symbols"
mkdir symbols
cd symbols
for i in {1..50}; do wget "$BASE_URL/$i.svg"; done
for i in {101..150}; do wget "$BASE_URL/$i.svg"; done
for f in *.svg; do convert -background transparent -resize 256x256 -density 500 $f ${f%.svg}.png; done
rm *.svg

Notes

The scripts use the timezone of Europe/Zurich and take the daylight saving into consideration. If this does not work for you, use the utc-offset parameter for an override.

Docker

Get it from Dockerhub (outdated)

~~https://hub.docker.com/r/caco3x/meteoswiss-forecast~~

Build it manually

docker build -t meteoswiss-forecast .

Run it in a Docker Container

Adjust the parameters in Dockerfile
Build the docker image with docker build -t meteoswiss-forecast .
Run it with docker run -it --rm -v /volume1/web/MeteoSwiss-Forecast:/data --name my-meteoswiss-forecast -p 12080:80 meteoswiss-forecast
Call it in a webbrowser: http://localhost:12080

replace existing Docker Container (Update)

docker build -t meteoswiss-forecast .
docker stop meteoswiss-forecast
docker rm meteoswiss-forecast
docker run -d -P --name meteoswiss-forecast -p 12080:80 -v /volume1/web/MeteoSwiss-Forecast:/data meteoswiss-forecast

Legal

The scripts only use publicly available data provided by the website of MeteoSwiss.

The scripts are provided under the terms of the GPL V3.

TODO

Update Images in readme
Refactor code, split data extractor and plot generator code

Internals

Data Fetching Flow

As of Nov. 2022, the workflow and URLs changed:

Get Overview Version

https://www.meteoschweiz.admin.ch/product/output/weather-pill/versions.json

{"currentVersionDirectory":"version__20221116_0707"}

Get Overview Data based on Zip code (contains city name)

https://www.meteoschweiz.admin.ch/product/output/weather-pill/version__20221116_0707/de/800100.json

{"path":"/lokalprognose/zuerich/8001.html","temp_high":"13","name":"Zürich","temp_low":"8","weather_symbol_id":"3"}

Fetch Forecast Data

Get Forecast version

https://www.meteoschweiz.admin.ch/product/output/forecast-chart/versions.json

{"currentVersionDirectory":"version__20221116_0709"}

Get Forecast Data

https://www.meteoschweiz.admin.ch/product/output/forecast-chart/version__20221116_0709/de/800100.json

[{"max_date":1668636000000,"rainfall":[[1668553200000,0.2],[1668556800000,0.2],[1668560400000,1],[1668564000000,0.7],[1668567600000,0.2],[1668571200000,0],[1668574800000,0],[1668578400000,0],[1668582000000,0],[1668585600000,0],[1668589200000,0],[1668592800000,0],[1668596400000,0],[1668600000000,0],[1668603600000,0],[1668607200000,0],[1668610800000,0],[1668614400000,0],[1668618000000,0],[1668621600000,0],[1668625200000,0],[1668628800000,0],[1668632400000,0],[1668636000000,0]],"sunrise":1668580293000,"day_string":"Heute","sunshine":[[1668553200000,0],[1668556800000,0],[1668560400000,0],[1668564000000,0],[1668567600000,0],[1668571200000,0],[1668574800000,0],[1668578400000,0],[1668582000000,0],[1668585600000,0],[1668589200000,21],[1668592800000,36],[1668596400000,36],[1668600000000,37],[1668603600000,33],[1668607200000,2],[1668610800000,0],[1668614400000,0],
[..]

caco3/MeteoSwiss-Forecast