keeping it local…

[Mark]

I have a Davis VP+ and an OS with two extension modules. I’m in a very remote location and get my internet connection over a metered satellite connection. my irrigation and weather station system eats up an enormous portion of my monthly data allocation. is there any way for tye os to directly address the weather station without having to gomto the web and get it off wunderground?

[Ray]

I am not sure how to do so — what kind of connection interface your weather stations provides? If it provides an HTTP interface, it should be possible, at least by modifying the firmware, to have OS directly talk to the weather station. Also, by default the OS firmware queries the weather server once every hour, and the frequency can be easily reduced by changing one number in the firmware. You should also check your weather station’s update frequency to see if there is a way to reduce the number of times it sends data to wunderground over a day.

[Mark]

Well, it emits a stream that Wunderground can parse and Davis publishes the schema of that stream. Also you could use Cumulus as an intermediary like Wunderground. I use Cumulus i host on an RPi on my network to feed Wunderground. I would love to be able to just point my OS to my little Cumulus server and keep the data transfer within mymlan…

[Mark]

I was thinking that, since Cumulus is open source, maybe it would be possible to tap into it to receive the weather station data rather than the Davis station direct. Cumulus in my case aggregates both the weather station and a separate Davis sensor array containing soil temp and moisture and leave transpiration. So it woul be optimal to have the OS take astream off my Cumulus server running locally on an RPi. The alternative would be for OS to understand the stream that comes off the weather station itself and just point the OS to that IP address and make the connection that way. I know I am over simplifying this but it feels like most of the work is done — the stream from the weather station is documented, the stream from WU is documented, the stream types from Cumulus are documented. If we could just link these then remote farmers like me could really take best advantage of the data without hitting a data transfer wall. And we could observe the weather in a high frequency way which, in the microclimate environment in which I live, is very valuable in terms of saving precious water.

[Author]

Posts