Weather Forecast lying means dead grass

We had 1.55” of rain here in the last 75 days. Forecast constantly promises rain tomorrow, or next hour, only for that to keep being pushed back as time elapses. The result is I am battling Rachio to water my lawn. It won’t water it for the next two weeks. Using the Empty feature makes no difference. I get the “bad imput => bad results”, but I cannot control THAT input.

What I am suggesting here is that Rachio uses a weighted “trust” on the forecast, where if it already refrained from watering once and that rain forecast has not materialized, it should go ahead and water on next occasion. This is a choice to make between never watering vs overwatering a little. I’ll take the overwatering any day. During this weird conditions in South Florida, my grass would be dead if I just trusted Rachio. Again, not necessarily its fault, but don’t blindly trust forecast, it is, after all, a probability, not a certainty.


This is the forecast:

It looked exactly the same a week ago, too. No rain, despite the forecast.

That looks like the Tempest unit’s forecast. Not Rachio’s forecast. What is the forecast in the Rachio app?

Correct. The Rachio view matches that, the problem is it never rains, it just says it will. And consequently, Rachio keeps skipping watering and I use quick runs in the morning to compensate. This is some past “forecast” still shown in Rachio.

The Rachio screenshot says Today is sun/cloud without rain. Today being 6/6. You don’t show any days past today from the Rachio app. The Tempest screenshot has Today (again 6/6) as a sun/cloud/rain with a 20% chance. Views don’t match.

Are you using Flex daily? From my understanding, Flex Daily uses current weather predictions whereas, Flex Monthly uses historical weather data. Neither would water if rain above the rain threshold was determined by the weather station.

the forecast changes several times a day. The “will rain” keeps being pushed back. The problem is I have watering scheduled before sunrise, so the decision to skip happens around midnight. I thought you wanted the “past”. See how Rachio still shows rain each of past days, but only a drizzle of 0.01” one day. No rain on all the other days, but Rachio happily skipped watering. Here’s current status:

Again, this is last week all over again. A week ago, forecast looked identical, rain every day. It did not rain and I had to babysit waterings, otherwise my new sod would be dead by now.

Rain drops on past days isn’t what happened, it was what had been predicted. I edited my post above while you were typing. Please review. And what is your rain threshold set at?

Flex Daily as the post was created in that category. I disagree that Flex Daily uses rain skip threshold (set at 1/4” in my case). Fixed uses rain skip, Flex Daily runs a balance sheet with the actual rain observed/expected. The Flex Daily never “skipped” per se, it won’t show in weather intelligence skips either, the fixed schedules I had to create would do that and I even disabled rain skip to make them work - they do take forecast into account.

Correct. Rachio still shows the “past forecast” is what I said. That was the forecast, rain. Today was “rain” until the evening when it turned to partly cloudy. It’s been like this every single day. Starts with a rain forecast, ends up not raining at all. It’s like the perfect storm of chain events. Not necessarily Rachio’s fault, and I cannot blame Tempest either, Apple Weather does the same thing. Keeps saying it’ll rain at 6pm until 4pm when that becomes 9pm and so on. Rain is running away from me :sob:

Sorry, I didn’t look at the category your post was started in - Flex Daily.

My Flex Daily uses the “bucket”. My zones are set to 50% depletion. When I look at the zones page, each zone has a soil moisture %. My 5 yard zones are currently showing 4% (which isn’t actually 4%, it is 4% before getting to the 50% depletion). Rachio uses 0 - 110% to indicate the amount from 50% to saturation. Confusing at first.

Anyway, the app has scheduled a run tomorrow morning which will occur unless we do get rain. And I think we’d have to get 1/4" or more before Rachio wouldn’t water as that is my rain threshold setting. But, it might simply add the lower amount of rain to the % and perhaps move the watering to Sunday.

What are your advanced settings on a zone?

Ok, now we’re speaking the same language :smiley:

Yes, the actual precipitation + remaining predicted precipitation for today is added to the “bucket” in the precipitation row. If the result at the end of the day (balance + irrigation + precipitation - evapotranspiration) is high enough, watering is pushed back. It’s exactly what’s been happening for a week now, each morning, due to the high precipitation prediction, Rachio decides to push that watering back, only to receive no rain for the day. Then the story repeats the next day. This causes the balance to be a constant 0 (0%).

I have water availability of 0.12 in/in, with root depth set to 6in and allowed depletion set to 50%. This means that the range 0-100% on the chart represents 0.12 * 6 * 0.5 = 0.36". The top 6" of soil can hold 0.72" of water, with the top 50% of that (3") holding 0.36". I understand how the allowed depletion works, it’s a weird choice, but you get used to it. Nozzle precipitation is about 0.52in/h, with an 80% efficiency. This means 0.36" at 80% efficiency and 0.52in/h precipitation rate takes about 50 minutes (0.36 / 0.8 / 0.52 * 60 min). Back to my problem, here are the observed facts:

  1. Forecast says precipitation is a-coming - albeit at typically 20% chance, for a total of 0.5" or more. 0.36" will make the balance sheet exceed 100% regardless of starting point.
  2. Rachio trusts that (despite the 20% chance) and assumes rain will compensate a watering skip
  3. Rain forgets to arrive, no precipitation, leaving the end balance back at 0%
  4. new day, repeat from step 1.

the lawn is therefore never watered. Even today, Rachio says, based on future forecast, that the next watering is June 21st. Forecast is rain every day, but so it was last week too. Still no rain, except for a 0.01" last night. 1.56" of rain in the last 77 days. Last real rain was 3" on Mar 23rd.

Possible solutions:

  1. take precipitation chance into account - dunno, maybe apply the chance % to the expected amount? More thought needs to be given here. A 20% chance of rain almost ALWAYS ends in no rain, or 0.01", or TRACE :frowning:
  2. once future precipitation prediction was taken into account but not materialized, for the next watering chance, ignore the prediction once, allowing a watering to go through. It seems better to over water than not water at all?

Another example: last Monday, I had watered manually and the 5 zones with new sod were somewhat mid range of moisture levels, with no planned watering for Tuesday (I can only water Tue, Thu, Sun). I had hit Empty on all 5 of them bringing them to 0%, this caused Rachio to show watering the next morning, Tuesday. Went to sleep happy that this is going to work, woke up to only 2 of the 5 actually having watered. What happened was that the precipitation amount expected has changed throughout the morning and each schedule had checked it an hour before it ran. As the predicted amount dropped, the last two schedules in the list of 5 were allowed to go ahead and water. I am using independent Flex Daily schedules for each zone to avoid the problem with end before sunrise, where regardless of how many zones will water, they start at a time as if all would water). I’ve been observing this for a while now, and am pretty sure I understand the workings. Again, Rachio works with the data it has, bad data leads to bad decisions. Let’s try and account for bad data, is what I’m proposing here.

Thank you,

If you can only water 3 days a week, that isn’t ideal for using Flex Daily. Flex Daily works best when it can water whenever it calculates that it needs to.

Here is where I’d start.

  1. Ensure your Rachio is looking at your weather station.
  2. Delete all of your schedules and create 1 Flex Daily with all zones AND all days ending before sunrise.

Run this for a few days ( or week or two) and see if it works correctly.

If it does work, check with whoever is restricting watering days and see if you can get a waiver for using a smart controller. You may have to send the info about the smart controller and how it actually saves water by watering only when needed. From what I’ve read, some counties do provide a waiver.


  1. It is using my Tempest. Same happens with national weather (it also suggests rain is coming :D)
  2. I can try that, sure - what’s different this week is that chance of precipitation is in the 50-70%, as opposed to 20% for the last week, so I hope I’m not out of time. Actually, I hope it damn rains :smiley: I don’t want to keep paying the water bill :imp:

The Palm Beach Water Department has that rule of 3 days a week. We’re lucky, Broward next door has 2 days a week. I am aware of the issue with Flex Daily and less than 4 days, it has worked fine until the fake forecasts started. Again, I think this is the perfect storm situation, when all events happen in a certain way. I have set all schedules to allow all days, and it still says June 21st as next, I’ll disable all schedules and create one with the 5 zones and see where that goes.

Not looking promising so far, but I have to let it run for a few days. Still says next watering day is June 20th. Forecast is rain every day, which is why June 20th.

the forecast as seen by Rachio:

How about a screenshot of your zone page showing all the zones & info?

I’m up in central Florida. My schedule has tomorrow & Sunday as watering days and then next is the 19th. Chance of rain between 10% and 70% on each day in-between. Documented here to see how it changes over the next week.

They are showing 110% because I manually ran them, the grass is wilting…
Let me go ahead and Empty the five zones

If I empty the zones, it shows as if it will water tomorrow morning - however, come morning, it won’t. At 1am when it checks for forecast and sees “rain coming” it simply pushes back the watering, happened two nights already.

I have a theory. The problem may come from the fact that the moisture graph is capped at 0%-110%. In all reality, 0% really means 50% with my allowed depletion of 50%. The moisture level can actually go below 0%. If it did, the sum of precipitation+forecast+irrigation-evapotranspiration would provide a more accurate look at the state of the soil. It would allow waterings even if rain is predicted, because it would take more water to reach a level in the top 50% allowed depletion of mine. The fact that 0% is the bottom and no further water depletion is possible, allows small amounts of rain to skip the watering, even when the soil is in dire need of water.