Rainbird for example, show it dipping below the 50% depletion. So even though they assumed 1/2" irrigiation, they still accounted for the dip below the depletion level when calculating the current level.
Similarly with CSU - they show an increasing MAD (which is an interesting concept in the case of annual overseeding for lawn ie increasing MAD based on plant maturity)
ironically, the other end (capacity) is also different. Once saturation is reached, there is no ability to keep adding. Runoff is assumed. In Rachio, I have a zone predicted to be at nearly 200% capacity, so would be a long time before irrigation will be called for when it is in fact needed sooner.
The permanent wilting point (a level beyond the allowed depletion) should be considered. This is emergency level - if approached, irrigation should be added, regardless of predicted precipitation or not. Because if the rain doesn’t come, it’s too late.
The CSU paper is especially interesting if you are quantitatively minded.
Thanks @DigitalBoy. We’ve read all of these and more
Still trying to take the most common sense approach that works across the broadest base, still figuring out what that is.
Different papers have different concepts of how much saturation should be allowed. I agree 200% is too much, going to look across soil types and modify our algorithm to make that closer to reality.
Yes, we’ve actually worked with Jose Chavez, one of the authors of that paper. CSU is in our backyard and my Alma mater
If the scheduling algorithm detects that you’ll bottom out on a day that you can’t water, it’ll will try to “top off” the zone on the first day you’re allowed to water prior to the day you would have otherwise bottomed out.
I was sure you had , was providing for the community if people are interested in the math. Fun stuff.
[quote=“tburch, post:23, topic:1747”]
If the scheduling algorithm detects that you’ll bottom out on a day that you can’t water, it’ll will try to “top off” the zone on the first day you’re allowed to water prior to the day you would have otherwise bottomed out.
[/quote]Not if rain was forecast. It would not irrigate, you would fall below if the rain never came, but Rachio would assume you are still right at the allowed depletion. The additional loss would not be counted.
ok, i have read through this quickly coupled with the fact that im half a tard, but, at quick glance, i dont know if a negative value will work in some of these equations, ill have to look at that a little more.
but would raising the mad to 1 (or 99 depending on your vantage point) solve this original issue. i mean, it looks to me like we are arguing whether the algorithm(s) should be inclusive or exclusive of the mad. it sounds inclusive and thus causing this issue? or am i cutting up too much coke here with occaum’s razor?
by setting it to non zero, will allow the failed percip to cause moisture to drop below mad thus spawning a watering event? that or mad becomes exclusive.
raising or lowering the limit wouldn’t help at all. The issue is that Rachio does not account for losses below the limit, regardless of where you set the limit. Hence, your current moisture levels are over-inflated, even when over the allowed limit (assuming you did actually hit the limit at some time in the past).
So if you raise it from 50% to 25% depletion (ie. retain 75%, up from 50%), if you dip below 25% depletion, that amount you dip below will be ignored. Raising it is likely to cause more issues as you are more likely to hit it, more often.
how did you come to this conclusion? looking at your graph mad is at 100 so how could it go any lower/higher or is that the display issue @franz was getting at?
Just also noticed that I again have rain in the forecast the day after I am allowed to water. Instead of “topping off” to 100% on the day I’m allowed to water it would be nice if the system saw the forecasted amount of precipitation and took that into account. So if I needed 1" to top off and there was .25" in the forecast for the following day, the Rachio would only run long enough to put .75" into the ground and then let the rain top off the next day. Otherwise the rain water will put me above 100% and likely just run off the property.This would be a cool “advanced” feature to toggle on/off as an end user.
[quote=“plainsane, post:29, topic:1747”]
how did you come to this conclusion? looking at your graph mad is at 100 so how could it go any lower/higher or is that the display issue @franz was getting at?
[/quote]MAD is not set to 100%. It is the default 50%.
Look at the below (real) example. Take 7/22. Starting balance was 0 (meaning I’m at allowed depletion). I lost 0.09 in. And I ended at…0 in. Rachio did not count the 0.09 in loss as it is floored at the depletion level. It is not a display only issue (that was a separate problem and has been fixed already). Remember that the 0 level (MAD) is not true zero, it is relative to MAD. Moisture level can, and does, drop below MAD (assuming anything less than 100% MAD) and needs to be computed as such.
