Looking for Ideas on Weather Rain Hold Irrigation Event

[klafollette]

I'm looking for any ideas/direction you all have found works well to tie irrigation run/hold decisions into weather forecasts.

I have HS3 with a stand-alone Rain-Bird irrigation system, and just built an ESP8266 relay that I can control from Homeseer via a HS device and MQTT. I have relay wired into the Rain-Bird 'Rain Hold' contacts, so I can either remotely/manually put the Rain-Bird System on a Rain Hold by turning on/off the HS device. Or, wanting HS event logic that could automatically suspend irrigation.

Currently I using WeatherXML plugin to get forecast info from nearby NOAA station into HS. I can get 'Rain Today' On/Off, and a % precipitation forecasted values for today/coming days. I'm thinking something like an event that says if 'Rain Today' = ON AND % Precipitation = 100%, then put on rain hold.

We're in Colorado mountains, and have a lot of localized micro-climate weather, so even if we have 75%+ rain in forecast, it is very frequent that we end up with nothing. So pretty much looking for near certainty before having event trigger rain hold. I might even want to take today and tomorrow at 100% before considering putting on rain hold.

Of course the next level would be my own actual rain sensor or moisture sensors, but not quite ready to go there yet.

[Pete]

Initially had a 10-12 zone Rain-Bird installation done in the early 2000's. Rain-Bird also installed a rain sensor sponge at the time.

Went to using a Rain8Net configuration with two modules. Initially ran a serial catXX cable from the old Rain Bird box to Homeseer 2 (one side of the house garage wall to the basement to the other side of the house). Worked well for a couple of years with first generation mcsSprinkler.

Afterwards tied the Davis weather station / WeatherXML to mcsSprinklers. Added an inline water meter and water valve later on. Started with the Davis Rain bucket then added the Dallas 1-Wire rain bucket and a digital rain sensor (well two) for helping with the watering. Best was using E.T. calculations being done on mcsSprinklers. One digital sensor was used as just a rain sensor (or fog) and the other used as a measurement sensor. The measurements were OK but the tipping bucket was always the best mechanism.

Over time installed mcsSprinklers on a Seagate Dockstar and ran mcsSprinklers on it with Mono. It was the first (before HS3) plugin to run in Linux Mono. (years before HS3). Switched the catXX from serial to ethernet POE. Used first generation USB SSD drives for the device. Worked well.

Fit it all in the Rainbird box in the garage. I also adjusted (mostly to tinker with) the Rainbird rain sensor sponge and connected it to a wireless transmitter. It did not work as well as the digital rain sensors.

Over time and using E.T. my lawn looked nice.

Today I would probably go same as you to using MQTT. It is easy and works here 100% of the time.

[klafollette]

Thanks Pete for your experiences. One thing I'm not grasping, what is "E.T."? Started reading through mcsSprinklers manual, but didn't find "E.T.". Would you see mcsSprinklers plugin worth it to just pull weather data and make determination to put existing Rainbird into hold mode?

[AllHailJ]

I had the Rain8Net and quite frankly I thought it sucked. The GoControl WI15VZ-1 Smart Irrigation Controller. I have set up multiple scripts and have shared them on the forum. I am also in CO and have gone to 3 shorter watering cycles per zone and this has saved considerable water and cost; Lawn is in better shape as well. One of the devices on the GoControl is an enable/disable device. I control watering from HSTouch. I also think you would need to add how much rain is in the forecast. A piddly shower still requires watering especially when under watering restrictions. I have a Rain8Net that I will give away for the price of shipping if you are interested.

Hope this helps.

[mterry63]

Adding a weather station and plugin that tracks accumulation and average temperature was my only solution to actively modifying irrigation behavior programmatically.

As you mentioned, depending on your location and climate, online forecasts may be dramatically different than your real world experience.

I would say just buckle down and do it.


Sent from my Pixel 2 using Tapatalk

[Pete]

E.T. = evaportranspiration

[klafollette]

Well, now this week hasn't been a total waste. I learned a new word..."evapotranspiration". Thanks for that

[Michael McSharry]

Anticipation of rainfall or solar radiation are different animals. mcsSprinklers has provisions for user tweaking of forecast days, percentages and amount of rainfall to be used to hold off irrigation. I do not personally use a future rainfall restriction on irrigation for the very reasons stated at the start of the thread. There is just too much local variability.

If it does rain then the computed moisture content remaining in the soil will have increased. This delays the next irrigation cycle and the amount of delay is proportional to the amount of rain that has fallen.

ET is very often used in commercial irrigation to optimize water use and plant health. It's effectiveness is only as good as the quality of the sensors providing the algorithm's inputs. Local weather stations, when properly maintained, provide the best source. The two most important measurements are the amount of solar radiation and the amount of rainfall. Temperature and Wind are also important when they are off-nominal.

The valve controller (Rain8 vs. GoControl vs. others) has no bearing on the rainfall hold. I do agree that Rain8 is problematic, but if one was to wind the clock back 15 years it was the best consumer product for automated irrigation that was available. Much like comparing X10 lights with the lighting options that are available today. Technology moves on.