When I bought an ambient weather WS-2902, I knew it had the ability to write its data to a custom server. Unfortunately the documentation on how the weather station sends data was lacking, and most of my searching on the internet for a solution didn't result in anything useful. I was able to eventally create a FastAPI endpoint that the weather station can write to. Once I had the weather observations on my server (ahem, raspberry pi) I decided to build a Plotly Dash dashboard to display the data on an old iPad. Once the dahsboard was situated I decided to incorporate forecast data from The National Weather Service as well as a GIF of the most recent radar loop (full disclosure, as a GIS lover, using this GIF seems like cheating, but it works really well).
A very simple FastAPI powered endpoint to take weather observation data from an Ambient weather WS2902 weather station compute the dewpoint and write it to a CSV.
Roughly based on some of the information from ambient weather here: https://ambientweather.com/faqs/question/view/id/1857/
The NWS has a great free API for accessing forecast data all over the United States. I have cron tasks set to get the forecast data hourly, and dump it to a set of CSV files since it doesn't change often and the radar loop is updated every 15 min.
The Dashboard was built using Plotly's Dash framework. All the plots in the dashboard are automatically updated every 5 minutes, and the current weather conditions (temp, wind, etc) are updated every 15 seconds. All the data is wrangled using polars the sunrise/sunset times used for shading the background of the plots are all calculated by the astral package.
When you mouse over any of the plots tooltips appear to show details about the data, and applying a filter to the temp/dewpoint figure will subsuquently filter the other figures to the same time period (which is a lot of fun to use when a big cold rain front comes in and the temp drops real fast, the wind picks up and the daily rain skyrockets!)