Transcript
LAUREN CRAWFORD 0:06
Good morning everyone and thanks for joining us for our seminar titled Water Potential 401: Advances in Field Water Potential Measurements, presented by Dr. Doug Cobos. Doug is one of Decagon’s water potential experts, and was the lead researcher for the development of Decagon’s MPS-1, MPS-2, and MPS-6 water potential sensors. He’s been at Decagon now for 12 years, and has served as the Director of Research and Development Department for the past seven years. He’s also adjunct faculty at Washington State University. During his time at Decagon, he’s also led the development of the TECP sensor that was part of the Mars Phoenix lander, which was a really fun project for us, and also the new GS1 rugged soil moisture sensor, in addition to many, many other instruments. We’re also really proud of Doug because he won first in his age group this weekend in his first Olympic length triathlon. So during this seminar, please feel free to ask Doug questions by typing them into the Questions field in the GoToWebinar software. We’ll go through as many questions as we possibly can during the seminar. And if we don’t get to your questions, either during or after the seminar, we’ll email you with the answer. Next week, you’ll receive an email with a link to the archive of today’s seminar, a link to the slides of the PDF of the slides today and also your personalized certificate of completion. Now I’ll hand the mic over to Doug.
DOUG COBOS 1:38
Okay, thanks for that, Lauren. That’s going to be a little bit tough to live up to. But we’ll see what we can do here. So thanks for joining today. As the title shows, we’re going to talk about advances in field water potential measurement. Okay, so I want to show you the outline of what we’re going to talk about real quick. We’re going to start out with the previous— Okay, so not going to spend much time on the importance of water potential, spend just a couple of slides on it. Because if you guys are tuned into this presentation, then you already know why water potential might be important to you. But what I want to spend the time on is, first of all, talking about the characteristics that make good field water potential sensors. And those we’re going to focus primarily on the measurement range accuracy, and then we’re going to talk about things like ease of use, and lifetime and maintenance issues. And then we’re going to talk about each of the field water potential sensors that’s available out there in some detail and highlight these different characteristics that make— that are important for a good sensor. So we’re going to talk about measurement range and accuracy of each of these sensors. And then I want to spend a little bit of time at the end zeroing in on some recent advances that we have made here at Decagon in the sensing of soil water potential in the field and mostly because we’re little bit proud of it. So, as you all know the second law of thermodynamics tells us that water will always flow from high potential to low potential. So, the water potential gradient in the environment determines the direction and rate of flow of water in the soil plant atmosphere continuum. So water is always going to flow from high potential to low potential. So soil water flow and drainage are controlled by water potential. For those of you who come from an engineering background, the soil suction, which is the same as water potential, just without the negative sign, also controls the mechanical strength of soil. And so this is pretty important if you’re putting foundations or roads or buildings on soil, you need to know the mechanical strength and that is determined by the water potential, essentially. Also, water potential determines the availability of water for biological processes such as seed germination, but probably the reason that most of you are here is because the water potential determines if the water in the soil is available for plants to utilize. And this is one of the primary reasons that you might measure water potential in the field environment.
DOUG COBOS 4:16
So let’s talk a little bit now about about making those measurements in the field. And let me make the statement right now that unfortunately, there is no perfect water potential sensor, okay. All water potential sensors have their own limitations, they either have limited measurement range, and we’ll spend quite a bit of time today talking about the effective measurement range of the various sensors. There are limitations in accuracy. Even if it is a very wide measurement range, then often the accuracy is not the same over the entire measurement range. Some of the sensors are difficult to use, meaning that you have to do routine maintenance. And some of the sensors actually have a limited lifetime, in that they’re not stable in the soil environment and will dissolve over time. I also want to make the statement that there really are no true direct measurements of water potential. All your water potential sensors really rely on measuring the water potential of water that’s in equilibrium with the soil water. And we will talk about that as we get into the specific sensor types. So let’s talk a little bit about measurement range. Here’s a little chart that shows the range of water potentials that you might see in the field environment. Of course, pure free water is defined as having zero water potential, and so saturated soil would have a matric potential of zero. That water is not bound, it’s basically free. Field capacity we define, it’s something like negative 10 to negative 33 kilopascals. And notice I’ll talk about— in this presentation, I’m trying to talk strictly in terms of kilopascals. You can see the other common units, megapascals, bars, there many other units that are used, but these are some of the most common ones. Permanent wilting point, or the end of the plant available water range is about negative 1500 kilopascals and then air dry, if we’re talking about 50% relative humidity is about negative 100,000 kilopascals of water potential. And so ideally, we would like our water potential sensor to make accurate measurements across this entire measurement range all the way from zero to air dry, which is typically what you would see in the field environment. But unfortunately, there is not a sensor existing that can measure accurately across that full range. There are sensors that have sensitivity across that full range, but none of them are going to measure accurately. And so this is our job here at Decgon and other places where people are developing, you know, new and better water potential sensors to try and increase the range and increase the accuracy across that range. So one of the primary things that we’re going to talk about when we talk about each of the individual sensors is the effective measurement range of those sensors. And you can see here also, if you focus in on the right side of this chart, on the field instrument side, you can see kind of a graphical depiction of where the effective measurement range is of the various instruments. So the tensiometer on the far right, basically zero to negative 100 kilop ascals with good accuracy across that range. And then some of the indirect methods, the MPS-2 and MPS-6 and the heat dissipation and granular matrix sensors have their various measurement ranges listed there and some information also about the accuracy of the sensors in those various ranges. But we’re going to break this down in more detail as we go along.
DOUG COBOS 7:47
The second limitation that we want to talk a little bit about is the accuracy. So often the first principles, or I guess you could call them more direct sensors, like the thermocouple psychrometers. And the tensiometers often don’t need a lot of calibration and should be accurate just from using first principles. But most of the sensors that are available for field deployment are indirect sensors that are that are calibrated. And their accuracy really is determined by a few things, first of all the quality of the calibration or if they’ve been calibrated at all. And then also the hysteresis in response time. And finally, the stability. If the sensor isn’t stable in soil, then you can’t really expect it to read the same and read accurately for long periods of time. And we will talk some about the quality of calibration and a little bit about the stability. We’re not going to discuss hysteresis and response time, just because of time constraints on the presentation. There are also limitations that I’m just calling general limitations, and some of these might be the ease of use. So first of all, is your temperature, or is your sensor temperature stable? Well, if it’s not temperature stable, do you have to correct for that? And if you have to correct for that, is that a complex temperature correction? And if so, then that makes it a pretty difficult sensor to use in the field. Also, we have to worry a little bit about longevity of the sensor. If the sensor you know dissolves in low pH soil solution and you only get a year of maintenance free or a year of lifetime out of it, then that’s not terribly useful for a long term study. Finally, you do have to worry a little bit about routine maintenance of sensors. Some sensors will have problems, for instance, tensiometers, if the soil dries down too much, if it dries past about 100, negative 100 kilopascals, then you cavitate and you have to go back out and refill the sensor once the soil wets back up. And so those maintenance considerations are also pretty large for sensors in the field environment.
DOUG COBOS 9:55
So here’s the outline again and let’s start talking now about the individual field water potential sensors. And I’m going to go in and talk first about the liquid equilibration sensors then the vapor equilibration sensors and then the porous matrix equilibration sensors where we’ll spend most of our time because most of the field sensors fall into that category. So our liquid equilibration instrument is the tensiometer. And I know most of you are familiar with tensiometers. Tensiometers have been around since 1960s, and are a great measurement of water potential and by far the most accurate and precise measurement in the wet range of soil water potentials. So with this instrument, there’s a small chamber that has some water in it, and that water is allowed to come into contact with the soil water through the ceramic cup that you’ll see at the end of the tensiometer. As the matric forces pull water from the tensiometer, it exerts a tension on the water inside the tensiometer. And when that tension inside the tensiometer is the same as the matric potential outside the tensiometer in the soil, then you’ve reached equilibrium and you can measure the tension on that water with a pressure transducer, and that gives you the matric potential of the soil. Biggest problem with tensiometers is the limited measurement range. So you’ll see on the slide, 0 to negative 90 kPa is a good rule of thumb. And that really depends on your elevation, because it depends on atmospheric pressure. But if you exceed that range and go into the range, maybe where plants are starting to exhibit signs of water stress, you know, drier than negative 90 or 100 kilopascals, then the tensiometer will cavitate or boil at room temperature and you’ll get a bubble in the water column, which gas now in the water column is expansive, so you lose your signal and basically get no reliable information. And in fact, once the soil even wets back up, you’ll generally have a bubble left in the tensiometer. So then you have to go out and refill the tensiometer. And so this is the really the biggest limitation of tensiometers is that you can’t use them if your soil’s going to dry past about negative 90 kPa. So they’re excellent in the wet range, but can’t use them in the dry range. And people of course, have been working on strategies to try and extend the range of tensiometers. And I wanted to talk about a couple of instruments, which I would say are more or less experimental at this stage. People have been putting a lot of effort into them, but I don’t think either are really proven for field use yet. So the extended range tensiometer is really similar to a regular tensiometer in that it’s just a liquid water system with a ceramic, but a lot of material science has gone into these to try and minimize the water volume, and also make the surfaces extremely smooth. And then if you use some specialized equipment to fill these tensiometers and pressurize the water, which forces all the gas into solution, then you can impede the cavitation. So if there aren’t any micro bubbles, there aren’t any gas phase, if there isn’t any gas phase in the tensiometer, then you can retard that cavitation and get to very negative water potentials before you actually experience the bubble or the gas phase. And so the range on these, I mean, if you go out and dig through the literature, you’ll see ranges that go way, way down. But I think a range of maybe zero to negative 2000 kilopascals, or even drier in some instances are possible with this technique. But it’s really kind of unknown whether this extended range capability is going to last very long in the field, okay. So if you have water exchange coming into and out of the tensiometer, then eventually you’re going to get, you know, water perhaps with some dissolved gas or some air bubbles in it coming into the tensiometer. And then you might get cavitation at water potentials more toward zero. So I would say that these are not really well proven for the field use yet. But there’s a lot of work being done on these. So keep your eyes out for that. Another extended range tensiometer is the polymer tensiometer. And there’s quite a nice body of literature, body of work out in the literature about the polymer tensiometers also, because these have been in development for several decades. So with these instruments, a swelling polymer is put inside the tensiometer that creates a positive pressure in the tensiometer when it’s wet up, so you could think about this as being something like the expansive water retention material in a baby diaper that swells up and holds water really tightly. So, when this material is wet up it exerts a positive pressure, so it basically gives you a pressure offset. So maybe if you put one of these instruments together, put it in pure water, and that material swells up and gives you a positive 2000 kilopascals of pressure then as that dries out, it dries down to zero kilopascals of pressure when it’s dry, and so you don’t have the problems with cavitation. The biggest problems with the polymer tensiometer, from looking through the literature, are that those polymer materials are quite temperature sensitive. So the way that they retain water, their moisture characteristic is quite temperature sensitive. And also there is what they call a zero offset drift problem in wet soil. So remember, in wet soil, these tensiometers have a positive pressure of maybe positive 2000 kilopascals which is basically at the end of the measurement range of whatever pressure transducers use. Well, if anything changes there with the polymer, then you get maybe a few 100 kilopascal offset in that, which makes these not as useful at the wet end as a regular tensiometer would be. Now some of these issues are being worked on. And various papers have done a better job with that. So I would say again, that these sensors are still in the experimental stage and not well proven for field use. But there is the possibility that these could end up being a good instrument in the future.
DOUG COBOS 16:45
So those are the liquid equilibration techniques. I want to talk a little bit now about the vapor equilibration techniques. So the equation that you see here, on the left side is called the Kelvin equation. And what the Kelvin equation allows you to do is know the water potential if you know the temperature and relative humidity of the atmosphere in your soil environment. And you can see that in the little chart, we’ve done the math there, that pure free water with a zero water potential is a relative humidity of one or 100%. But notice that even at permanent wilting point, okay, negative 1500 kilopascals, well, that translates into a relative humidity of still very close to 99%. And so as exciting as this Kelvin equation is that says, Boy, all we need to do is measure the relative humidity and temperature of the soil and we know the water potential. Well yeah, that’s true, but even at what we would consider pretty dry soil, at permanent wilting point, the relative humidity is still 99%. And if you think about the accuracy of a typical capacitance type hygrometer, the accuracy typically is maybe plus or minus 3% relative humidity, between 95 and 100% relative humidity. So you can see that plus or minus 3% is a huge error bar in terms of water potential. So it takes a really specialized instrument to be able to measure relative humidity accurately enough to actually determine a meaningful water potential. And the one field instrument that is capable of doing that is the thermocouple psychrometer. If you look at the top right, you’ll see that a thermocouple psychcrometer is basically a little thermocouple junction that’s inside some porous membrane that allows the atmosphere around that thermocouple to equilibrate with the atmosphere of the soil. And then if you can measure, as with any psychrometer, if you can measure the dry bulb temperature and the wet bulb temperature of that thermocouple junction, then the relative humidity is simply proportional to the wet bulb cooling of that thermocouple junction. And so if you know the humidity, then you know the water potential. And so these instruments are really elegant in that this is a first principles measurement. But implementing this in the soil environment is quite difficult. So you need to be able to resolve temperature differences of about a 1,000th of a degree C. And that may sound pretty simple, but it is not simple at all. And there are only a few measurement devices that are able to make that type of measurement. That’s a nano volt type— nano volt level signal. And it’s really difficult to make those measurements. So one of the big drawbacks of the thermalcouple psychrometer is that even if the psychrometer is pretty cheap, the readout unit either from Westcore like you see down here or if you’re using a Campbell Scientific logger, those are pretty expensive and also pretty complex. You have to have some skill to be able to use these. The other big problem with thermocouple psychrometers is that any ambient temperature drift is— it will create a pretty big error in the measurement. So if you think about trying to measure a dry bulb temperature, and then later in time measure a wet bulb temperature, then if the soil ambient temperature drifts a few thousandths of a degree in that time, then that’s going to give you a pretty substantial error. So you have to be really careful to use these in a very temperature stable environment, which if you’re talking about field soil, that means deep in the soil profile. If you try and use these shallow in the soil profile, where you get big diurnal temperature swings, then you’re going to have a whole bunch of problems. So you need to use these deep in the soil profile. One of the beautiful aspects of the thermocouple psychrometer is that the measurement ranges is really huge, okay, from zero to about negative 6000 kilopascals is about as far as you can go and still create the wet bulb from condensation on that thermocouple junction. But another big drawback is that the accuracy is plus or minus 100 kilopascals, which if you’re in really dry soil, okay, if you’re at permanent wilting point, plus or minus 100 kilopascals is great. Okay, that gives you great accuracy. But if you’re trying to measure in wet soil, so for instance, if you’re trying to measure in a soil that’s at negative 100 kilopascals, which is drier than field capacity, but still relatively wet, so if you’re trying to make a measurement there at negative 100, and your error bars are plus or minus 100, then your error is basically 100%. And so the bottom line is these are not really good for wet soil. They’re really good for dry soil, but a little bit complex to use.
DOUG COBOS 21:52
Oh, another— I put this slide in to remind myself. So another aspect of thermal— well, let’s back up and we didn’t talk about this early on, because most of you guys are familiar with this and I want to try and keep the seminar as short as possible. But the total water potential is simply the sum of the four components of water potential, which are the matric potential, the osmotic potential, the pressure potential, and the gravitational potential. When we’re measuring soil water potential, we really only care about two of those. We typically define the total water potential as matric plus osmotic potential. The thermocouple psychrometer does measure matric potential plus osmotic potential, whereas all of the rest of the devices that we’re talking about today, the tensiometers, and all of the granular matrix or porous matrix sensors, those only measure the matric component of water potential. So, that’s an important consideration to take into account when making measurements with the psychrometer versus making measurements with the other instruments.
DOUG COBOS 22:57
Okay, so now let’s switch gears to the last class of instrument that I want to talk about. And these are instruments that make use of porous matrix equilibration. And so you’ll see on the screen some moist soil, okay, some really wet soil pulled off the internet. Okay, so thought exercise: if you have a really wet soil, okay, this we could say it’s saturation, I guess it’s glistening, okay. So if we have a soil at zero water potential, and we insert a ceramic, a porous ceramic material at air dry or negative 100,000 kilopascals, then what’s going to happen to the water in the system? Well, the second law of thermodynamics tells us that water will always flow from high potential to low potential, so it’s going to flow from the soil into that ceramic disc, okay. And it’s going to do that until when? Well, it’s going to do that until the water potentials equilibrate. Okay, it’s always going to flow from high potential to low potential. And it doesn’t stop until the water potentials equalize. And so if you insert this porous material, its water potential will, after some amount of time, equilibrate and become the same as the water potential in the surrounding soil. And the converse is also true if you have a dry soil, as you can see here by a cracked soil, and you insert a moist ceramic disc, then what happens? Well, water is going to flow from high potential to low potential and eventually that ceramic will come to the water potential— the matric potential of the surrounding soil. So, if you can measure the water content of your ceramic disc or your porous material, and you know its moisture characteristic curve, then if you can measure, for instance, say we measure water content of point three, and then we can extrapolate and use our moisture characteristic curve and say hey, we have a water potential here of about negative 20 kilopascals. And that is the way that all of the other sensors that we’re going to talk about work.
DOUG COBOS 25:05
And this is not at all a new concept. In fact, I think it was 1908, the Livingston cones were developed and information was published on those. With the Livingston cones, there were just some ceramic cones that were dried out and weighed and then inserted into the soil and allowed to equilibrate. And then they pulled out of the soil, soil cleaned off them and weighed again, and from their mass, you could calculate the water content and from knowing this moisture characteristic curve, then you could predict the water potential of the soil because those had equilibrated with the water potential in the soil. And so really, all of the other techniques that we’re going to talk about, for the rest of the presentation, are just derivations of that original technique from over a century ago.
DOUG COBOS 25:55
So the first type I’d like to talk about are the gypsum block sensors, which use electrical conductivity to measure the water state of, in this case, the gypsum block and infer water potential. So basically, there’s a couple of electrodes in there, and you measure the electrical conductivity, or electrical resistivity of the gypsum material, and from that infer the water potential. The nice part of these sensors is that they’re really inexpensive. There’s just a little block of gypsum and a couple of electrodes. The readout device is pretty, pretty simple. It’s just a resistance meter. The real drawback though, is that gypsum is soluble in the soil solution, especially at high pH. And so you might only get one to two years of service out of the sensors. Also, because you’re measuring electrical conductivity, the output from the sensors, or the electrical conductivity is pretty sensitive to salt in the soil, to the soil salinity. And because you’re measuring electrical conductivity, and electrical conductivity is highly temperature sensitive, you have temperature sensitivity. So those are some pretty substantial drawbacks when using these sensors for research. They’re pretty good, and a lot of people use them for irrigation scheduling, as long as you’re willing to basically waste them every every year or two. But they’re so inexpensive, you can do that pretty economically. Measurement range is pretty good on the dry end, they go all the way to permanent wilting point, but a little bit limited on the wet end, they only go up to negative 50 kilopascals. And then of course, the accuracy on these, these really don’t even come with a calibration. It’s kind of a qualitative measurement that will give you some information about the water potential of the soil, but not anything quantitative unless the sensors are calibrated by the user. And so not necessarily a great research instrument, but not a bad instrument for irrigation. Another similar instrument is the granular matrix sensor. These sensors also use electrical conductivity to understand the moisture state of the porous material. But in this case, the porous material is more of a sand material and not a gypsum material. But there is still a gypsum capsule in the sensor to try and buffer the effects of soil salinity. So it works to some extent, but if you’re in a high salinity environment, then even that fails, and also that gypsum capsule will dissolve over time and eventually you’ll have no more gypsum and the sensor becomes not worth as much. And you also have temperature sensitivity because you’re making an electrical conductivity measurement. With these granular matrix sensors, the measurement range is really optimized for irrigation, so zero to about negative 200 kilopascals. And these instruments do come with a calibration but there’s no accuracy specification given, so the accuracy of the sensors is a bit of a question. But certainly these are widely used in irrigation. I think that they are used in research some but it’s a little bit unclear what the error bars are as far as accuracy.
DOUG COBOS 29:21
The second type of porous matrix sensor, instead of measuring electrical conductivity of the porous matrix to determine its water state, these sensors use the thermal conductivity or heat dissipation properties of the ceramic to understand the moisture state. One of the big advantages is that these sensors are typically made of some ceramic material that is stable in the soil environment. So first of all, it’s not affected by salts, and it doesn’t doesn’t dissolve in in the soil solution. One of the big drawbacks, though, is that these instruments are a little bit more complex to use. You have to have a data acquisition system that’s capable of adding a known amount of heat or a known current to the heating element and then measuring the temperature rise and the temperature fall and then doing some analysis of that time versus temperature curve and determining the heat dissipation and understanding then the thermal properties. It also, to do this correctly, you need to implement a somewhat complex temperature correction because a lot of heat flow in porous media is carried by latent heat flux, by water vapor flow away from the heated surface toward the cooler surfaces. And both the diffusivity of water vapor and air and the latent heat of vaporization are temperature dependent. And so you have to correct for that to get accurate measurements. If you can do that, if you’re well set up to do that, and you’re, you know, for instance, a guru in Campbell Scientific logger programming and you can do this, then these instruments become really good instruments. And notice the measurement range is quite nice, about negative 10 kilopascals to negative 2500 kilopascals. I’ve seen papers where the measurement range has been extended well beyond negative 2500 kilopascals, all the way out to air dry with these sensors. So it is possible to do that. But of course, the measurements become more qualitative, and less quantitative, bigger error bars as you go out toward the dry end. The real difficulty for most users with the sensors are that each sensor requires individual calibration. So these come without an accuracy spec and without a calibration and so you have to go through the steps to calibrate these yourself. And the accuracy that you would get out of each sensor depends on the quality of the calibration that you implement. And so this is fine if you have, you know, some skilled graduate students or technicians that can do this. But it adds a lot of time and expense to actually determine those individual calibrations for each individual sensor. There are also a couple of sensors out there that use the heat capacity of the ceramic to measure the water potential of the soil. So instead of measuring the thermal conductivity, you measure the heat capacity. Again, these should not be affected by salts or dissolution. And I have to throw this up there because I was looking on websites the other day, you know, in preparing the seminar, and the manufacturer of these sensors claims that you should be able to measure all the way from zero to negative 1 million kilopascals. So all the way from saturated to oven dry. I’m dubious, but maybe it’s possible. You may get some sensitivity throughout that whole range. But again, though the accuracy is not known as there’s no accuracy spec given on the sensors. So that’s one of those measurement range specifications that makes me scratch my head a little bit but we’ll leave that for you guys to decide. You guys need to publish some papers on those and then say what the actual measurement range is.
DOUG COBOS 33:23
So now we’re arriving at the sensors that I want to talk a little bit more about. These are the porous matrix sensors that use dielectric permittivity to measure the water status of the ceramic material or the porous matrix. And these are the sensors that we make. And I always, you know, hate to talk a whole bunch about our sensors and really zero in on these but the reason that we’re here today is to talk about some of the advances in measuring field water potential. And some of the advances that we’re familiar with, of course, are the advances that we’re making. So we want to tell you a little bit about what we’ve been doing with these sensors and what that means for your research. So before we get into that, with these sensors, the premise is the same as the others. You put a ceramic in the soil, and it equilibrates with the water potential of the soil, and then you measure the dielectric permittivity, which is an electrical property of a material. And that’s proportional to the water content and therefore the water potential. Measurement range on these sensors is about negative 10 to negative 100,000 kilopascals. And I’ve put a little star there by negative 100,000 because we get sensitivity to negative 100,000. But it’s not going to be particularly accurate at negative 100,000 kilopascals. And again, with any of these indirect measurements, okay with any of these porous matrix sensors, the accuracy really depends on the quality of calibration, and that’s what I want to talk about a little bit. So our first iteration of this sensor type was the MPS-1. And some of you guys probably remember the old MPS-1 sensor. This was a sensor that we started working on back in probably about 2000, and brought this to market and found out that while most of the sensors shared a common calibration, and you can see, for instance, in the graph down here on the bottom right that a lot of the sensors fall on the one to one line, which is what we want. You’ll see that some of the sensors deviate substantially from that one to one line and basically have large error or really poor accuracy in the measurement. And so we started looking at this quite a bit more and trying to figure out where this loss of accuracy was coming from. And it turns out that it is extremely difficult to make a ceramic material that each ceramic disc has the same moisture characteristic curve as all the other ceramic discs. Okay, so what we’d like is to have a universal calibration that each sensor shares the same calibration. But to do that, each disk has to be exactly the same. And I’m sure that most of you have never tried to manufacture a ceramic with a wide and controlled pore size distribution before. And so you’re probably thinking, how hard could this really be? Well, let me tell you from experience, it is hard, okay. It is really difficult to make a whole bunch of ceramics and have them all have the same pore size distribution, which is what essentially affects the calibration of the sensors. And so we decided, well, we’re not going to be able to control this as well as we want. So we need to calibrate these sensors and give it an empirical calibration to increase the accuracy. So the second generation was the MPS-2, and the MPS-2, as the name indicates, had a two point calibration, where we calibrated the sensors dry and we calibrated them in a vacuum saturated state. And that helped quite a bit. But we still had only plus or minus 25% accuracy. And plus or minus 25% is probably good enough for irrigation management, but it’s not going to be good enough for most research. And, you know, when we published the 25% accuracy spec, we took a lot of heat, you know, from a research friend saying, you know, Why is this so broad? You know, Why is this such a bad accuracy spec? And we said, Well, it’s because we publish an accuracy spec. If you look at any of the other porous matrix sensors out there, none of them publish an accuracy spec, and there’s a good reason for that, that this is not an easy measurement to make. But when we were evaluating this, we, you know, figured out that to get research grade accuracy, we need a multi point calibration for good accuracy. And we originally started trying to do a multi point calibration in the pressure plates. And this was a, I would classify that as a failure, okay, which certainly could not get this done. The equilibration times in pressure plates are ridiculously long, talking about leaving sensors in there for several weeks to come into equilibrium at multiple points. And the reproducibility was horrible. And, you know, we’ve spent a little bit of time looking at the problems that went on there. And the problems with equilibrating soil under positive pressure instead of suction are pretty well chronicled in the literature. And so we’re pretty well convinced that the pressure plates are not a good way to calibrate these instruments. And so we needed to go back and basically build a better mousetrap. And really what we needed was some calibration device that can calibrate multiple sensors simultaneously, okay. We need a lot of throughput. So this doesn’t take a lot of time and add a lot of cost. We needed something that equilibrates rapidly, and we needed something that calibrates under suction, because the equilibration under a positive pressure has a lot of intrinsic problems with it that may not be overcomable. So I want to talk a little bit, this is where we come into the recent advances and show you a little bit of some things that we’re kind of proud of actually, because this has been now 15 years in the making that it’s taken us a long time and a lot of effort to try and come up with a good field water potential sensor. And this calibration device was a big part of it.
DOUG COBOS 39:32
And so the concept that we had was that we have a computer that controls the suction or the vacuum, the negative pressure on multiple sensor chambers, and that that suction or vacuum is adjusted in steps and the sensors are calibrated at each of those suction levels. And so this is, you know, an artist’s rendition or a SolidWorks rendering of what that might look like. But this is what it looks like in practice. So we have a vacuum pump and a reservoir to provide vacuum to the system. And then we use a computer that controls the suction that’s applied to a manifold. So we can control that suction or control that vacuum at whatever level we want. And that manifold controls this suction on either 30 or 60 individual sensor chambers. So we have 60, 30 to 60 sensors that are calibrated, all at the same time in their individual sensors, which have individual water potential control. And so this is what it ends up looking like if you plot the output of the sensors, as we’re calibrating them. And what you’ll see here is that we can calibrate between 30 and 60 sensors in about 24 hours. So this is a huge step forward from the pressure plates that might take a week or two weeks to reach equilibrium at these different steps. So it’s a relatively, I mean, 24 hours seems like a long time, but in terms of water potential calibration, it actually is a relatively short period of time. So you also see here that these are the uncalibrated output from the sensors, and there’s quite a bit of scatter in that output. But after we calibrate at the six points, then we end up with really quite nice accuracy. So we now feel like we have a research grade instrument with accuracy that all the sensors fall within plus or minus 10% of the reading in the moist range. Okay, so this is a moist range calibration, and we expect all the sensors to read extremely well in that critical zone of water potential that governs water flow, and it spends most of its— that spans most of the range where irrigation might be controlled. We also have some indications from people in the research community that our performance at the dry end is is good as well. So you can see in the upper graph, that the agreement amongst the sensors is really nice, all the way down to permanent wilting point. And that we have reliable and interesting qualitative data down even below permanent wilting point. So as you go drier, you get further from where we’ve done most of the calibration, and we try not to put error bars on that, because we’re a little bit uncertain of how the calibration should hold there. We do use a scientifically defensible calibration procedure in that dry range. And so we expect that the accuracy should be pretty good, but it’s really hard to put error bars on that. And we’re working on that presently, trying to figure out just how accurate we are, as we get all the way down to air dry. So what does this mean for your research? Okay, so I want to back up just a little bit and talk about, you know, what we talked about originally, the limitations of field water potential sensors. Well we talked about one of the limitations, especially with tensiometers, is that what we want ideally is a maintenance free sensor and these porous matrix sensors, all of these except the gypsum blocks, are pretty much maintenance free and should be stable over a long time. So the MPS-6 is maintenance free. It’s not ever going to cavitate, you never have to refill it, it’s not going to dissolve in the soil unless you put it in maybe some pH one type mine tailings or something like that. We haven’t looked at that really carefully. But I wouldn’t suggest putting it in super acidic soil. So we have a maintenance free sensor. We have good accuracy at the wet end. Each of the sensors is individually calibrated, okay. From about negative 10 to negative 100 kilopascals, we’re really confident in the accuracy. It’s not tensiometer great accuracy, but nothing else is. Tensiometers are as accurate as you can get in the wet end. But this is accuracy that’s quite nice and research grade. And then unlike the tensiometers, we do have an extended measurement range. Say that we have good accuracy to negative 100 kPa, fair accuracy up to permanent wilting point, and then certainly meaningful measurements out to air dry, all the way to negative 100,000 kilopascals. And so we’re feeling pretty good about this. Of course, we’d like to improve some aspects of this. Some of you guys are probably asking, Why is it limited to negative 10 at the wet end? Why not zero? Well, the answer is that there aren’t any really large pores in that ceramic that drain between zero and negative 10 kilopascals. So some of our next steps are certainly to extend that wet end range all the way to zero kPa. We’re working on that right now because that is an important range for water flow in the soil and also for the geotechnical engineers to understand slope stability. Generally the slopes fail when they’re at, you know, saturation near zero kPa or even at a positive water potential or ponded conditions. We also want to spend some more time working in the dry range and try and confirm the accuracy that we think we have at those dry water potentials. And if we’re not getting the accuracy that we want, then we try and implement a calibration point in the dry end. But that is not anywhere nearly as easy as we would like it to be. The temptation there is to use pressure plates, but our experiences there have been such beastly failures that we aren’t going to go down that path again. So we probably have to design something, you know, based around our vapor equilibration instrumentation to try and get that dry end calibration point.
DOUG COBOS 45:52
So this is what I had for you today. I hope you guys found the analysis of the different instruments interesting and you know, understand maybe now, some of the limitations of field water potential sensors a little bit better. And I also hope you found interesting the parts that we were talking about, maybe we’re bragging a little bit, I realize that, but we are somewhat proud of what we’ve been able to do with the MPS-6 sensors and we hope that you’ll find those useful for your research.
