Announcement

Collapse
No announcement yet.

Soil moisture sensors better than ET for irrigation

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    #31
    Originally posted by Michael McSharry View Post
    I If I installed it closer the surface then it fluctuated what looked like surface temperature.

    On the other hand, Irrometer claims their granular matrix Watermark sensor doesn't dissolve and require "no maintenance." Apparently it uses a water permeable membrane to help hold itself together. This paper gives a formula to compensate for temperature effects, like what you observed: http://www.google.com/url?sa=t&rct=j...634,bs.1,d.cWc I guess that means a soil temperature probe would be needed near the moisture sensor so as to calculate the correction.

    Comment


      #32
      Complicating matters is that it's helpful (maybe even essential) to know some basic facts about the type of soil one has and its salinity before picking a soil moisture sensor: http://www.bae.ncsu.edu/topic/go_irr...onitoring-.pdf

      Comment


        #33
        I can't know whether these were truly fair comparisons against ET controllers, but Irrometer has a page of graphs like this:


        (http://www.irrometer.com/landscape.html#wem ) which Irrometer says "show results that are typical of the savings users can achieve using our soil moisture based technology to make controllers 'Smart'." I'm skeptical, but is it possible the comparisons really are both fair and typical?

        By the way, how much soil moisture might come from dew formation? My morning grass often feels plenty wet, even with no rain or irrigation to explain it. Just speculating, but if nontrivial amounts make it to the soil, that might be an area where a properly setup soil moisture controller and a properly setup ET controller might differ in their assessments. If one rules out improperly configured controllers, I wonder what else might account for differences.
        Last edited by NeverDie; April 4, 2014, 10:35 AM.

        Comment


          #34
          I think I may have this figured out now. Most likely, almost no matter what sensor is picked, it's going to benefit from calibration using the exact soil it will be installed in. I'm inferring that from a recent master's thesis that examined the issue for four different soil moisture sensors (http://digitool.library.colostate.ed...8xMjA1MjM=.pdf), including the watermark one discussed earlier in this thread.

          So, in the interest of moving beyond the theoretical, I've ordered a few relatively inexpensive wireless soil moisture sensors whose mechanism I don't know but which mount flush at the surface so they should be easy to install, maintain, and relocate if necessary. I'm hoping they will have good transmission range, because fixing that would take extra effort. If I'm lucky, I'll be able to calibrate them purely based on my subjective impression of turf health (which is what matters). This way, if a hot spot develops that's not due to a failing in the irrigation heads, I'll just yank the sensor out and stick it back down in the hotspot. I'm supposing only one (maybe two?) driest spot can exist in a zone, and I'm hoping this approach will nail a sensor to the very center of it. July and August will likely be the biggest stress test on whether this low effort approach will work. If I notice the low effort approach is failing, then my fallback plan is to acquire an Irrometer soil tensiometer (because its results are accurate independent of soil type) to troubleshoot/debug and perhaps calibrate the wireless sensors more accurately (or, in the worst case, reveal that my wireless soil moisture sensors are inherently not good enough and that I should have purchased something else). Irrometer developed a tensiometer (the TGA model) specifically for turf, and its gauge is easily upgraded to support continuous data acquisition for troubleshooting, should that be needed or useful.

          I had thought I would use ET measurements as a check on whether any given soil moisture sensor (SMS) is operating within reason (based on irrigation amounts), so an alarm could be flagged. That may still be worthwhile, but perhaps I can also compare SMS's against one another to flag possible faults early and before they get out of hand. The goal, of course, is to eventually arrive at a truly efficient system that, as much as possible, is also set-and-forget.

          If there's interest, I'll post a follow-up afterward on how it went.
          Last edited by NeverDie; April 4, 2014, 11:29 AM.

          Comment


            #35
            Originally posted by NeverDie View Post
            If there's interest, I'll post a follow-up afterward on how it went.
            Hi ND,

            I'm very interested in your results. I've been thinking about adding water sensors as well, and your posts have made me realize it won't be as easy as I'd hoped it would be.

            Cheers
            Al
            HS 4.2.8.0: 2134 Devices 1252 Events
            Z-Wave 3.0.10.0: 133 Nodes on one Z-Net

            Comment


              #36
              Originally posted by Michael McSharry View Post
              I If I installed it closer the surface then it fluctuated what looked like surface temperature.
              By the way, for anyone interested in the Watermark sensors, Irrometer also makes a version, the model 200SS-V, which has the temperature compensation already built-in, and which allegedly provides a linear 0-3 volt output: https://www.intelesense.net/public/d...Vdatasheet.pdf

              The reason these sensors are made using gypsum is to buffer against possible salinity effects and salinity variations in the soil. Otherwise, for instance, potassium chloride (the K in some NPK fertilizer) would skew the sensor readings after a fertilization. Even in the Watermark sensors, I've read that the gypsum still does dissolve over time, which I would guess governs the lifespan of the sensor (allegedly 5+ years).

              This may be a dumb question, but is it possible to provide necessary amounts of K to plants without affecting soil salinity? e.g. maybe the K is supplied in some kind of compound that doesn't skew soil moisture sensors but that the plant can absorb and readily convert into the type of K it needs? If anybody happens to know, please post. If you're a gardening guru, here's a chance to shine.
              Last edited by NeverDie; April 4, 2014, 11:03 AM.

              Comment


                #37
                Also, the same as with ET, it may be worthwhile not to water (or to water less) if the probability of rain is high enough, or if it's too windy, etc. This is where getting accurate weather info (whether via HomeSeer or by other means) and weather forecasts would still be useful. Accuweather used to say not just whether it would rain, but also how much rain (in inches) to expect. Not sure whether Accuweather still does that. Do any of the other weather sites offer inches of rainfall predictions?

                Of course, one could argue that if you go to the bother of getting a bunch of weather data, why not just do ET? The solar radiation data isn't always easy to come by, but it can be approximated using latitude, longitude, time of year, and (maybe) further refined if you gather info regarding cloud cover. I don't view it as an either/or choice (since I'm hopeful ET can help cross-check against SMS, and visa versa), but someone who does view it either/or, or who is budgeted for either/or, might find this yet another strong argument in favor of irrigation control using ET derived from high quality weather data. To the degree the weather data isn't high quality, though, then I would think there's serious risk of garbage-input becoming garbage-output, with overwatering as the most common hedging strategy. If there's any truth to the Irrometer charts (above) comparing ET with SMS, I'm guessing it's that type of hedging strategy that's the biggest source of the difference in water consumption.
                Last edited by NeverDie; April 4, 2014, 12:40 PM.

                Comment


                  #38
                  mcsSprinklers has option to pull forecast from different sites and some are % and some are inches. I think WU is in inches. It really does not make that much difference because it is used to delay an irrigation decision. If your irrigation interval is fixed and the run time is being modulated then you also need to account for actual vs. forecast rain and account for rain received in your microclimate vs. rain that was received at your reporting station.

                  A chart that shows one method is generally always better than the other just means that the ET threshold was just set too high. If your crop is one that cannot tolerate any wilting then you will need to bias the ET to account for this. Grass is pretty tolerant.

                  Comment


                    #39
                    Originally posted by NeverDie View Post
                    The solar radiation data isn't always easy to come by, but it can be approximated using latitude, longitude, time of year, and (maybe) further refined if you gather info regarding cloud cover. I don't view it as an either/or choice (since I'm hopeful ET can help cross-check against SMS, and visa versa), but someone who does view it either/or, or who is budgeted for either/or, might find this yet another strong argument in favor of irrigation control using ET derived from high quality weather data. To the degree the weather data isn't high quality, though, then I would think there's serious risk of garbage-input becoming garbage-output, with overwatering as the most common hedging strategy.
                    The Davis Vantage Pro2 weather station can provide pretty decent weather measurements including wind, rainfall and solar radiation. It will also do an ET calculation. mcsSprinklers can import that data directly. Weather forecasts, on the other hand, are what they are.
                    Mike____________________________________________________________ __________________
                    HS3 Pro Edition 3.0.0.548, NUC i3

                    HW: Stargate | NX8e | CAV6.6 | Squeezebox | PCS | WGL 800RF | RFXCOM | Vantage Pro | Green-Eye | Edgeport/8 | Way2Call | Ecobee3 | EtherRain | Ubiquiti

                    Comment


                      #40
                      I guess the term that describes what I meant is "Quantitative Precipitation Forecast." I just checked, and accuweather still has them. However, even more helpful might be the type of forecast which shows the probabilities of different rainfall amounts:




                      Where can I find those?

                      Ordinary rain forecasts probabilities (i.e. probability of rainfall >= 0.01") aren't very useful.

                      Comment


                        #41
                        Originally posted by NeverDie View Post
                        "Quantitative Precipitation Forecast." I just checked, and accuweather still has them. However, even more helpful might be the type of forecast which shows the probabilities of different rainfall amounts Where can I find those?
                        Maybe the weather in Texas is more predictable, but in upstate NY, the forecast is of very limited use for a single property. At best, it describes an average over a county or two. At worst, it's no better than a guess. I do not inhibit watering unless the probability of rain is 80% or greater. Probabilities less than 50% are indistinguishable from zero.
                        Mike____________________________________________________________ __________________
                        HS3 Pro Edition 3.0.0.548, NUC i3

                        HW: Stargate | NX8e | CAV6.6 | Squeezebox | PCS | WGL 800RF | RFXCOM | Vantage Pro | Green-Eye | Edgeport/8 | Way2Call | Ecobee3 | EtherRain | Ubiquiti

                        Comment


                          #42
                          Originally posted by Uncle Michael View Post
                          The Davis Vantage Pro2 weather station can provide pretty decent weather measurements including wind, rainfall and solar radiation. It will also do an ET calculation. mcsSprinklers can import that data directly. Weather forecasts, on the other hand, are what they are.
                          It's puzzling that most of the weather station software doesn't offer to calculate ET, though maybe it's because most don't offer a solar radiation sensor and the Davis does.

                          Or, maybe also because the calculations assume your weather station is mounted in the middle of a large expanse of "well irrigated" cool season grass which has 100% soil coverage, that's kept mowed to a 12cm height, that has a certain albeido, that your instrument readings are at prescribed heights, that no large obstructions block the wind, and so forth. Not sure what the sensitivity of the results are to those assumptions, but I'd be interested to learn.
                          Attached Files
                          Last edited by NeverDie; April 6, 2014, 06:59 AM.

                          Comment


                            #43
                            A high probability of rain doesn't imply a high forecasted amount, but a high forecasted amount does (?) seem to imply a high probability of rain. That's still a bit murky, which is why I'd prefer the type of breakdown the above chart is showing, which is the tell all. I'd guess its implicit in all the weather models. I wish I knew what the term for it is. Anyone know?
                            Your best weather predictions will come from a weather station you have on premise. Historically have always used the rain tipping bucket on premise for amounts of rain. That said my older Dallas Instruments rain tipping bucket really does work well.

                            You can utilize humidity sensors placed in numerous locations outside and even look at "fog" these days with the digital rain guage.

                            Yup here wife bugged me on why I spent so much time tweaking out our sprinkler system.

                            I also regraded thin topsoil areas plus added berms to redirect the water flow plus redirected water drainage.

                            All of the costs relating to the labor and materials for the "fixing" of the drainage and topology of my lawn cost probably 8X more than all of the measurement instrumentation I have included today in my sprinkler system stuff.

                            Those pieces were probably more effective than the measure pieces; but at a much higher cost. (drainage tubing, rock, dirt, custom cut limestone borders and very time consuming).

                            IE: One (1) drainage endeavor (of many to fix my lawn) was almost two weeks of digging at around 2-3K of a mixture of labor and drainage tubing.

                            Personally; and this is my opinion: I would start rebuilding the base infrastructure such that when it rains your yard makes full use of the rain water. When it doesn't rain and you sprinkle you utilize your sprinkling system to be most effective and efficient.
                            Last edited by Pete; April 5, 2014, 10:57 AM.
                            - Pete

                            Auto mator
                            Homeseer 3 Pro - 3.0.0.548 (Linux) - Ubuntu 18.04/W7e 64 bit Intel Haswell CPU 16Gb
                            Homeseer Zee2 (Lite) - 3.0.0.548 (Linux) - Ubuntu 18.04/W7e - CherryTrail x5-Z8350 BeeLink 4Gb BT3 Pro
                            HS4 Lite - Ubuntu 22.04 / Lenovo Tiny M900 / 32Gb Ram

                            HS4 Pro - V4.1.18.1 - Ubuntu 22.04 / Lenova Tiny M900 / 32Gb Ram
                            HSTouch on Intel tabletop tablets (Jogglers) - Asus AIO - Windows 11

                            X10, UPB, Zigbee, ZWave and Wifi MQTT automation-Tasmota-Espurna. OmniPro 2, Russound zoned audio, Alexa, Cheaper RFID, W800 and Home Assistant

                            Comment


                              #44
                              Originally posted by NeverDie View Post
                              Or, maybe also because the calculations assume your weather station is mounted in the middle of a large expanse of "well irrigated" cool season grass which has 100% soil coverage and that's clipped to 12cm height and has a certain albeido, that your instrument readings are at 2m height, that no large obstructions block the wind, and so forth. Not sure what the sensitivity of the results are to those assumptions, but I'd be interested to learn.
                              I don't worry too much about the absolute accuracy of the ET calculation. I'm more interested in the correlation between whatever calculated result I get and the apparent rate of water loss from each zone. Michael's software provides adjustable parameters to help tune that relationship, and by adjusting those values based on experience I find, especially given the variability within a zone, it is more than good enough for my needs.
                              Mike____________________________________________________________ __________________
                              HS3 Pro Edition 3.0.0.548, NUC i3

                              HW: Stargate | NX8e | CAV6.6 | Squeezebox | PCS | WGL 800RF | RFXCOM | Vantage Pro | Green-Eye | Edgeport/8 | Way2Call | Ecobee3 | EtherRain | Ubiquiti

                              Comment


                                #45
                                Originally posted by Pete View Post

                                Personally; and this is my opinion: I would start rebuilding the base infrastructure such that when it rains your yard makes full use of the rain water.
                                Other than terracing, what would that be?

                                Comment

                                Working...
                                X