OpenSprinkler › Forums › OpenSprinkler Unified Firmware › I'm thinking Mr. Zimmerman didn't live where I live › Reply To: I'm thinking Mr. Zimmerman didn't live where I live
I’ve been reading a lot on the theories regarding weather related to irrigation requirements. I’ve come to this conclusion: It is complicated 🙂
So complicated, in fact, that I’m not sure a meaningful equation can really be determined that will work for everyone, particularly since available data isn’t relevant or universal enough (ie. not everyone has rainfall data and those that do aren’t necessarily applicable to their actual property). Temperature is going to be the best indicator because even if a sensor at your local airport isn’t absolutely accurate to your property, it is likely to be relatively accurate to your property (ie. weather station reports X degrees while yours is really Y degrees but when theirs drops by 10% so does yours, etc). Rainfall, however, is highly localized so unless you’re collecting info at your property it isn’t likely to be relevant info. In fact, I feel the best two indicators are the temperature and amount of daylight, both of which are available without on-site sensors. Actual rainfall should be subtracted from actual irrigation (though not necessarily linearly) but only using on-site data.
I previously stated forecast data should be incorporated, but I’ve changed my mind on that – again due to quality of data. I did some observations locally over the past 7 days and 100% of the time Weather Underground predicted > 0mm precipitation there was none, and 100% of the time I actually had rainfall, it was not predicted in the previous day’s forecast. That makes for pretty useless irrigation adjustment. This is also why I suggest only using on-site data for actual rainfall because during the times it actually rained, the WU data said it didn’t and every time WU said it did… it didn’t. Zero % accuracy. Granted, that’s over a statistically insignificant data size of 7 days but it was enough for me to lose confidence in anything but actual on-site rain data.
Even with actual onsite rain data… to make a meaningful adjustment, we need to know not only how much hit the ground, but also how much our irrigation puts onto the ground! We can know that, but it takes a) measurement by the user and b) ability to input data into OS. I can, for example, know by my meter that my Zone 1 is putting out 10L/min to 100 square meters of lawn for 10 minutes (@ 100%) which means each square meter received 1L (assuming even distribution). If our rain sensor says we got 1mm of rain, then 1 square meter received 1L. In that perfect case, we know we need to run our irrigation for 0% since 100% was satisfied by rain. Was that enough water for our plants? Who knows, but it was the same amount it would have received via irrigation so if it is wrong, it would have been wrong anyway. If the rain gauge says 0.5mm then our irrigation should be reduced to 50%. But we can only know that by being able to input actual irrigation flow, time and land size irrigated.
Therefore, for rain data to mean anything, it should be locally gathered and we must be able to input our per-zone irrigation flow and area (time is known by the controller). Also per-zone enabling (to account for zones within a greenhouse) and/or weighting (for sub-surface drip) of rain adjustment should be available.
I said I would try to come up with something concrete, but at the same time I’ve learned this is going to be more difficult than I thought… a different project also fell into my lap. So I’m not going to come up with the “Morehouse Method” in the immediate future.
Earlier I suggested Humidity was over-emphasized. An obvious reason (that I didn’t think of initially) is that rain and humidity are related, so adjusting based on both double-hits the equation. Since rain = 100% humidity (by definition) one might suggest there’s only reason to account for one or the other but not both. It isn’t really that simple though. Humidity is important, because it directly influences evapotranspiration which, arguably, is the most complete analysis we can use to estimate watering adjustments… but it depends on where you are and what types of plants you’re irrigating. Ultimately, if you’re sticking with plants natural to your environment, then you don’t need irrigation at all (since those plants had millions of years to adapt to the level of natural moisture in the area). Since lawns don’t belong anywhere, and people like exotic plants, the affect of humidity on those plants is going to depend on how far out of “normal” they are. Like I said… complicated.
But maybe I’m over-thinking it. Consider how we use “dumb” controllers. We pick some amount of time to water – based on experience or just a guess – and we increase or decrease that depending on how the plants are doing. Once we have that more or less dialed in we know, by common sense and observation, that we have to increase that in the summer (or is that not universal?). Generally speaking, temperature increases in the summer so perhaps temperature is the primary factor we should consider, along with a linear adjustment for actual rainfall. But what temperature? Does nighttime temperature even matter? Maybe it does, I’m not sure (I kind of think it doesn’t). The amount of daylight also increases in the summer – the extent of which depends on how far you are from the equator. So maybe something like mean hourly daytime temperature times number of daylight hours to give some kind of “power” indication? Do solar radiation figures give us this? Unfortunately, solar data isn’t commonly available.
Further complicating things is type of irrigation. I’m fully sub-surface, for example, so I don’t believe wind has any effect on my watering needs while those with sprinklers shouldn’t be watering during windy times (unless they want to water the roads and neighbor’s houses). Drip irrigation for a garden in a greenhouse has vastly different adjustment requirements vs. sprinklers over turfgrass. Turfgrass heavily shaded by trees has considerable differences vs. full sun.
A lot of different consideration are in play. I’m not sure, at this point, that I can follow through with my hope to provide a new method that is meaningful to others. I’m going to try to work on one specific to my needs and, perhaps, it might be useful for others to adjust. I personally feel I want to focus on daytime hours, daytime temperatures, and actual rainfall past 24 hours… the latter requiring I buy more equipment.
Soil moisture sensors might be a way to go too. However, in reality you don’t want the root soil to be wet all the time so that’ll need some thought.
Sorry for adding more questions than answers.