I personally use wetter.com quite often as the single source of truth, to checkout how it outperforms on my location and how reliable it especially in terms of rain predictions, i'd started to capute the data for some time.
docker-compose up -dand start scraping by adding something like this to you prometheus config.
- job_name: wettercom_exporter
scrape_interval: 1m
static_configs:
- targets: [DEXXX0241] # extracted from the URL
labels: { location: "Regensburg" }
- targets: [DE0006194] # https://www.wetter.com/deutschland/leipzig/DE0006194.html
labels: { location: "Leipzig" }
relabel_configs:
- source_labels: [__address__]
target_label: __param_target
- source_labels: [__param_target]
target_label: instance
- target_label: __address__
replacement: internal.prom.klaut.io:9294