Water budgeting, Rachio not adapting to ET/weather variation in San Diego


#1

Well I initially purchased the Rachio with hopes of the software adapting to weather variation in San Diego (I realize that may sound ridiculous). But it seems that the water budgeting feature is instead reacting only to average seasonal variation. Due to drought restrictions I can not really use the flex schedule as I can only water on Sun/Th.

Throughout the end of last summer and extending through October there were multiple episodes of Santa Ana conditions with temperatures approaching 95 to 100 degrees and very low humidity. The Rachio kept decreasing the watering times as if just reacting to average seasonal variation. When we had rain in September and then a rapid onset of colder weather in November, there was no significant adjustment the other way either.

In August and October I ended up with dead grass from underwatering. In November I ended up with a dead palm tree from overwatering.

I feel like I have to go in and manually override the water budgeting almost any time we get rain or a Santa Ana to account for weather changes.

Is ET not being measured and used to calculate watering times based on the actual weather? I have a Davis weather station in my backyard that measures UV and solar and reports to PWS weather.

What is “Water Budgeting” based on? Over the past several months, it does not seem to reliably adjust for temperature and precipitation variation in manner that is effective for San Diego with my current setup.

Should I have things set up in a different manner?

Thanks!


#2

Et is only used for flex. You can use a flex and set the 2 days you are allowed to water, were you having issues with that?

You are correct about the budgeting feature, that is how it behaves. If you want the robotic overlords to take over your watering you have to go flex.


#3

The current weekly water budgeting does actually use a rolling real-time trendline of ET data. We look backwards and forward to derive an average ET value for the upcoming schedule and do a percent change adjustment for the upcoming week.

The water budgeting adjustments should not be underwatering or severely overwatering. We only allow it to move in 10% increments (+/-) so I wouldn’t think that could account for the extreme conditions you are seeing. Just an observation.

Assuming you have chosen this as your PWS…

Here is some generic information you might already know about:

http://support.rachio.com/article/282-water-budgeting-and-evapotranspiration-watersmart

The equation just looks at the last time we adjusted versus ET values currently, and does a percent change. I’m guessing most adjustments are within the 5%-8% range.

If you’d like a deeper analysis [support@rachio.com] can do a quick review of how you have the system setup.

In our upcoming 2.5 release water budgeting on fixed schedules will be getting much simpler. We will make seasonal adjustments monthly and during the month continue to track real time ET to determine if we can skip schedules based on the last time you watered and next watering date.

Hope this helps.

:cheers:


#4

Thanks!!

That makes sense - I guess that was my problem. With Santa Ana conditions at 95+ degrees I would want to go straight to 100% to ensure that adequate water is provided. It must not have been increasing fast enough to compensate.

At the same time, with an inch of rain in November, I really do not need to water for at least a week, probably 2 weeks with the low temp in the 40s and high in the 50s and 60s. Instead, I was still watering at 65 or 70% twice weekly. And for December and January I just turned off watering completely for 2 months as Rachio still wanted to water at 50% twice weekly when it did not rain despite 6.25" the first week of January.

Is the low end capped at 50%? - seems like it never dropped below that

I guess I should go to flex schedules, but would that change the duration of watering as well or only the frequency?

It is a shame that the drought restrictions in Socal are such that a proper watering setup with flex schedules cannot be implemented properly. I really like the idea behind flex schedules and the continued development of other options as well

Can flex work effectively when set to use only 2 days per week?


#5

It will change both duration and frequency, but if your sticking to a 2 day a week schedule, frequency will still be pretty fixed. Unless there is a rain event.


#6

thanks, if frequency will decrease to less than twice weekly with flex that should hopefully fix the winter issues

for the summer, i think i would really need to have more days available and longer possible watering times to better adjust


#7

Yes, that’s the problem with these restricted watering windows. They make you do something stupid, like Th/Sun watering schedule. If you expect rain on Friday, you could skip Thu, but if it doesn’t rain your hosed.

It will lengthen the watering times based on root depth and depletion, but only to the limit that it thinks the ground can hold based on the parameters you have set. It does split the cycles so you can have a soak period.

Overall, I like it. I do feel like it isn’t cutting back enough for the winter months, my ET is showing 0.10" per day on average. I have a dormant bermuda lawn so it doesn’t need much water at all (1" per month), so just have to go manual. I haven’t had it for a summer yet, so I’m a bit afraid of how much it will water.


#8

Yes, we cap at +/- 50%.

If you want close to flex, but simpler, we will be releasing a new concept called water as needed with our 2.5 software coming soon.

Agreed, restrictions really limit what a smart controller can do.

Not really, it will more than likely water on those days since it knows it is bounded by restricted days.

With heavy restrictions, a fixed schedule with smart cycle, water budgeting, and weather intelligence (virtual rain sensor) might ultimately be the best bet.

:cheers:


#9

The only reason I was pushing to flex is that et has a reactive effect in flex instead of a progressive effect in fixed.

I’m making some big assumptions but I would assume that your are trending with a linear regression over another assumption of 4 weeks of data? I don’t think the duration really matters, the weather data is very sawtooth when you plot it by daily highs. Use this link…if rachio is not using a sinusoidal curve here, water budgeting is completely useless, this data will will flatten a slope, a lot.
https://www.wunderground.com/history/airport/KSAN/2015/8/30/WeeklyHistory.html?req_city=&req_state=&req_statename=&reqdb.zip=&reqdb.magic=&reqdb.wmo=

I’m also making a huge mistake by correlating temp with et, but I’m willing to bet they are trending fairly similar in this case, maybe not.


#10

@jmelan, thanks for reaching out. We hate hearing of dead plants. I’m curious if we need to calibrate your zone settings? In reviewing your account and zones, I’m not seeing any tree zones. Let me know which zone(s) the under/over watering occurred on and I’ll review the seasonal adjustments from last fall.

Flex schedule watering durations do not change (unless a zone attribute is adjusted), but the interval dynamically adjusts with ET.

@brkaus, it might be worth double checking the root zone depth on your zones. I assume you probably have warm season grass configured in your zone setup if you have bermuda grass which has a default root zone depth of 9 inches. The 2.5 release will also expose the coefficient for each zone, so this could be adjusted to help fine tune the water requirements for each zone.


#11

so this is not from my weather station as it was not up and running until january, but one close by that tracks reasonably well with mine

you can see the large spikes every few weeks but especially in early October - note these are also often accompanied by a significant drop in humidity. also the large dropoff near Nov 1.

airport data for San Diego is often used but really the airport is directly adjacent to the water and the temperature stability there (or rainfall totals) does not represent the other 90% of San Diego very well.

with Santa Ana conditions (eg October) the ground can dry up so quickly that wildfires start, so it seems like I would want to go to 100% watering instantly for that week (but maybe not?)


#12

Thanks emil - you guys are really fantastic with customer service. I appreciate any suggestions you may have.

So, I do not have any zones with one type of plant except for grass zones. The zones with trees also have shrubs, vines, rose bushes, etc. I have 15 zones, 4 are grass zones, but the rest really are mixed. Unfortunately, there are also areas of some zones with rather variable sun exposure, so I just tried to average each zone into a category the best I could.

Zones 1-4 are grass, mostly fescue but with other random grass in there as well. All 4 areas ended up with large dead patches in October that are starting to recover now. The grass areas have about 4-6 inches of loam over “Redding cobbly loam” which is really mostly clay and large pebbles.

Zones 9 and 10 were the areas with the biggest overwatering issues in November, I lost the palm tree in Zone 10 and have a very unhappy palm in zone 9. Again, the soil is a layer of loam over the native “Redding cobbly loam”

Really there are multiple issues contributing to the palm tree loss, rapid temp drop with rain in a palm tree planted only 4 months earlier. The first winter is the most stressful time for a new palm and the palm was a Ravenea glauca that really does not like “wet feet”. The year before I lost shrubs in a different zone due to failing to adjust my old manual watering timer quickly enough in November.