One of the most unexpected and wonderful things about my Ambient Weather station is that within its settings app, you can configure an additional URL for it to send periodic pings to. In my case, I whipped up ambient2mqtt[0] as a way to capture the output of the weather station locally and do what I want with it - in this case, post the data to MQTT and InfluxDB.
A few years after that I also purchased a Purple Air PM2.5 monitor. I was pleasantly surprised that this device exposes a local REST interface for polling the data. Naturally, I wrote purpleair2mqtt[1] to pull the data from that and post to MQTT and InfluxDB.
Together these two applications and pieces of hardware give me a great view of the weather conditions. And like the author of the article, I found that other weather information tended to be off significantly. Doubly true as I live next to a lake in a rural-ish part of New England which is about 500ft higher than the city and airport where our weather data comes from. Over the last couple of years I've been pleased to see that my data has become the defacto source for weather data in my town (I publish it to numerous sources) and that the PM2.5 information gets brought into lots of reports, including the EPA's AirNow system - which is great, because it doesn't look like there's another PM2.5 monitor within 20 miles of me.
That's awesome! I just kinda write stuff that's interesting for me and put it up there on GitHub. Super happy that it helped someone else. Also, super happy that it still works with Home Assistant. I built the MQTT support for Home Assistant, but then work and life got in the way and I ended up not really using Home Assistant.
I cycle close to 100 miles per week and a having a good Weather app makes a huge difference for planning my rides. I developed https://apolloweather.com/ after Apple announced it was shutting down Dark Sky.
For me the most important thing is not the accuracy but having easier access to certain metrics to make better decisions. Feels Like and wind are a priority for me. Also UV Index during summer (I live in Florida). I want to know the best days to ride. I also want to know if I should use a thermal base layer or leg warmers. I’m willing to accept some discomfort during the first hour if it will get warmer later. Some rain is acceptable when is warmer. No rain is acceptable when its cold.
Will I have a headwind on the way back? Then I should manage my energy better during the first half.
As long as I see the big picture quickly for the whole week, I’m Ok with accuracy being less than perfect.
I went with WeatherKit because it was easy to use. TBH I never paid too much attention to the accuracy since I care more about the big picture. I’ve used many weather apps since I started cycling and learned to live with it. I treat the one that I use at the moment as the source of truth.
Love all the details here! For me, I skipped the crazy expensive Davis hardware and have built a network around the EcoWitt GW1100 station. It collects data at 900mhz from other EcoWitt sensors and sends it over a wifi network to WeeWX. In my case WeeWX is running in a FreeBSD jail, but it could be on any 24-7 powered computer on your network.
WeeWX is one of those hackers projects that I just love. It's not pretty, and a lot of the edges are real rough, but damn, it just works, all day, everyday, and stores data in sqlite with a funky, but beautiful HTML output as well.
According to many threads at wxforum.net, Ecowitt displays (and hubs like the GW1100/GW2000) will pick up Ambient Weather sensors but AW displays won't pick up Ecowitt sensors (AW did some kind of firmware tweak in addition to slapping their logo on the front).
Also, Ambient Weather apparently also started selling rebadged sensors from someone besides Ecowitt.
I'm a Meteorologist and after trying all kinds of commercially available and home-brew systems over the years I settled on a Weatherflow unit, by Tempest, it JUST WORKS and does everything I need, minus air quality which I have a Purple Air sensor for.
I threw a CD drive motor on a stick with plastic spoons and an ADC. Been logging data every second for a couple years. I think I'm near 20 million rows of unused data now.
Monitoring weather data at home is a cool project! I have a Davis station installed on a pole above my roof and it is interesting to see the difference between feelings and measures (esp. wind gust and rain rate).
It is very simple to use and works really well - it is also supported by WeeWX through an out-of-tree driver, but is easy to setup. It also allows you to keep your console in your living room while you collect data in your basement.
I would love this for my David VP2 station. It looks like it’s not available in the United States but they say it’s resold by ambient weather. Does anyone know of an alternative source? If this is it that’s more than twice the EU price. https://ambientweather.com/amweatherbridgepro.html
How useful are wind vanes and anemometers on typical home weather stations?
When I was putting together my DIY weather station I got this weather meter kit [1] from Sparkfun that includes a rain gauge, a wind vane, and an anemometer.
Later I realized that because of all the tall trees all around my neighborhood (Western Washington) and the less tall trees at my house and the house itself being in the way the wind direction and speed a couple meters above the ground near my house would not really be all that interesting.
In retrospect I wish I had just bought the rain gauge. Sparkfun doesn't sell that separately but Argent Data Systems has a product that seems to be identical to the Sparkfun one [2], but they also sell the individual sensors separately. Here's the rain gauge [3].
My backyard is surrounded by a six foot privacy fence. The back of the yard has a six foot hill and one of the sides is all cypress trees. So much of the wind into my backyard is blocked but I still find the anemometer to be useful as a relative measure. I completely ignore the wind direction but if I see that there's been a 15 mph wind in my backyard, I know the wind in the neighborhood is a good bit higher.
One of my favorite definitions from the Devil's Dictionary is: BAROMETER, n. An ingenious instrument which indicates what kind of weather we are having.
Looks like you can skip the logger and get a cheap rtlsdr to log data wirelessly (using e.g. https://github.com/bemasher/rtldavis). A cheaper still route would be to grab the outdoor sensors for a weather station supported by rtl_433 (or receive signals from a neighbor!)
If you don't mind a DIY approach you can use ESPHome to quickly integrate some low cost sensor modules (e.g. BME280s).
This is cool. I think setting up a weather station will be my next home automation project. Real-time wind / rain data might be useful for some things. Has anyone integrated a weather station with Home Assistant, and do you use the data for anything?
Not OP, but price and availability. Before about a year or two ago, it was a cheap-and-cheerful way to put some smarts into any project. Now it's impossible to get at twice the price, and the alternatives aren't quite there yet in terms of software support.
I still like reading about the projects you can do with it though. I find it inspiring for my own side things (even though fewer of them involve the RPi itself -- you can push microcontrollers really far these days!)
To get things clear, I love RPi and use rp2040 for lot of my projects. Use to go with RPi for small projects and now I have lot of ideas and unable to get a RPi for the last 1.5 yrs
Yeah it's been crazy out there. I've been trying to adapt by pushing into the ESP32, which has the WiFi connectivity I need and is cheap and available. But if I had built a business around RPi (as my old employer almost did -- they dodged a bullet there) I'd be in big trouble.
Fair enough. Sometime last year I just gave up on ever getting one and switched from reading these posts as “here’s an interesting project to try” to “here’s an interesting thing someone did”.
But, yea, it’s not really feasible to do these at home anymore, which is frustrating
A few years after that I also purchased a Purple Air PM2.5 monitor. I was pleasantly surprised that this device exposes a local REST interface for polling the data. Naturally, I wrote purpleair2mqtt[1] to pull the data from that and post to MQTT and InfluxDB.
Together these two applications and pieces of hardware give me a great view of the weather conditions. And like the author of the article, I found that other weather information tended to be off significantly. Doubly true as I live next to a lake in a rural-ish part of New England which is about 500ft higher than the city and airport where our weather data comes from. Over the last couple of years I've been pleased to see that my data has become the defacto source for weather data in my town (I publish it to numerous sources) and that the PM2.5 information gets brought into lots of reports, including the EPA's AirNow system - which is great, because it doesn't look like there's another PM2.5 monitor within 20 miles of me.
[0]: https://github.com/pridkett/ambient2mqtt [1]: https://github.com/pridkett/purpleair2mqtt