The HYPROP and WP4C provide the ability to make fast, accurate soil moisture release curves (soil water characteristic curves-SWCCs), but lab measurements have some limitations: sample throughput limits the number of curves that can be produced, and curves generated in a laboratory do not represent their in situ behavior. Lab-produced soil water retention curves can be paired with information from in situ moisture release curves for deeper insight into real-world variability.
Co-locating water potential sensors and soil moisture sensors in situ add many more moisture release curves to a researcher’s knowledge base. And, since it is primarily the in-place performance of unsaturated soils that is the chief concern to geotechnical engineers and irrigation scientists, adding in situ measurements to lab-produced curves would be ideal.
In this brief 20-minute webinar, Dr. Colin Campbell, METER research scientist, summarizes a recent paper given at the Pan American Conference of Unsaturated Soils. The paper, “Comparing in situ soil water characteristic curves to those generated in the lab” by Campbell et al. (2018), illustrates how well in situ generated SWCCs using the TEROS 21 calibrated matric potential sensor and METER water content sensors compare to those created in the lab.
Our scientists have decades of experience helping researchers and growers measure the soil-plant-atmosphere continuum.
Dr. Colin Campbell has been a research scientist at METER for 18 years following his Ph.D. at Texas A&M University in Soil Physics. He is currently serving as Vice President of Environment. He is also adjunct faculty with the Dept. of Crop and Soil Sciences at Washington State University where he co-teaches Environmental Biophysics, a class he took over from his father, Gaylon, nearly 25 years ago. Dr. Campbell’s early research focused on field-scale measurements of CO2 and water vapor flux but has shifted toward moisture and heat flow instrumentation for the soil-plant-atmosphere continuum.
Dr. Gaylon Campbell shows how to measure the thermal conductivity of wet insulation with a 5-minute read time. These previously impossible measurements are now possible through advanced electronics and analytical methods.
How researchers are combining in situ, drone, and satellite measurements to extract key information and how these data can be connected across scales.
In this 20-minute webinar, METER scientist Leo Rivera compares available methods and teaches how to combine the latest technology to generate full, accurate curves with hundreds of points in only a couple of days—instead of a couple of months.
Receive the latest content on a regular basis.
COLIN CAMPBELL 0:08
Hello, and welcome to today’s virtual seminar. My name is Colin Campbell. I’m a research scientist here at METER Group. And today I’m going to be talking about comparing in situ soil water characteristic curves to those generated in the lab. This was a project I worked on with scientists down at Brigham Young University, Neil Hansen and Brian Hopkins and several of their students along with some lab technicians here at METER Group. Soil water characteristic curves are important for many reasons. They’re a critical piece of knowledge to understand hydraulic properties of soil. Also, we can understand soil type from them. And we can predict mechanical performance and stability if we’re looking at them from a geotechnical angle. And I’ve got a typical soil moisture release curve or soil water characteristic curve right here on the right hand side.
COLIN CAMPBELL 1:12
We simply compare the water content here on the y axis with the suction or water content on the x axis. Traditionally, we could only do SWCC or soil moisture release curves in the lab. This is how we might do them. This is a chilled mirror dewpoint sensor here on the left, that would create or we would be able to measure soil water potential or soil suction in the very dry range using this instrument. And the wet range here on the right, we have a wind Schindler technique called a HYPROP that measures soil from zero to a few 100 kPa suction or negative few 100 kPa if we’re talking about water potential. We have to combine these measurements however, with a measurement of soil water content, which we can do fairly simply using a balance and and it’s shown here. So as I said, traditionally we simply combine these two, and we get a soil water characteristic curve here as shown on the bottom. Now, before I go any further, we have a lot of people who might be interested in this and I’m bouncing back and forth between soil suction, something that geotechnical engineers or civil engineers use to describe what traditionally soil scientists would call water potential. And there’s essentially the negative of each other. So while they have the same units, the negative 10 kPa and water potential would be a positive 10 kPa in soil suction.
COLIN CAMPBELL 2:59
So one thing that we asked was could we add field derived SWCCs to lab generated curves? This is something we were really interested in because you can only make a certain amount of SWCC’s in the lab. It just takes some time and take some lab effort. What if it were possible to simply bury sensors out in the field and create one just by taking data over time? So we went to the literature and started looking at other studies that might have done this. And what we found was that these studies have been done at least a couple of them in the last 10 to 20 years. But generally there was a reasonable lack of agreement between things found in the lab and things mentioned in the field. So we wondered, is this still the case?
COLIN CAMPBELL 3:44
Because things have changed in terms of technologies quite a bit since then. And we have the option of bearing some pretty accurate sensors. Now, this is an example of what I’m talking about the MPS 6 by METER Group, which is now called the TEROS 21. This is a sensor that has a wide range of suction so that it can measure water potential, so it can measure and it can now do that pretty accurately. The benchmark for measuring water potential we call a tensiometer. It just equilibrates water and in the soil across a ceramic matrix. Now, the problem with a tensiometer is it can’t go very dry. In fact, the driest it can go is negative 80 kPa in terms of water potential or positive 80 kPa in terms of soil suction. But we need those sensors to go over a much wider range to create a good moisture release curve. So our question was, can we make this technology good enough so that is possible? With the MPS 6 we now see that compared to a tensiometer we can get a plus or minus 10% accuracy across 80 kPa, which was exciting in terms of the opportunities to install these in the field and get good measurements. The water potential or suction calibration is actually stored in the instrument itself.
COLIN CAMPBELL 5:15
So we simply get an output of water potential or suction directly from the sensors in the field and create these SWCC’s directly. Now the idea of doing this experiment actually came out of a completely different thought or idea where the problem was that here at the Brigham Young University turf farm. They left for the weekend of Memorial Day a few years ago and came back to find all their grass had gone into dormancy because their irrigation system shut off and it was extremely hot that weekend. They called me up and said, hey, isn’t it possible to just monitor things in the soil so we could set an alarm to know if our irrigation system goes off so our plots won’t die? And of course it is. So I went and together with scientists down there we installed systems, two in the north plot and two in the south plot, so that we can monitor the moisture in the soil and make sure to alert somebody if things broke. Now, in these two plots, we installed as I said, two locations where we put sensors into the north plot that had almost 97% sand and a very small fine soil fraction and to the south plot where we had more clay and silt fraction, and you can see how these graded out there and their texture classifications. These are actually given for the ASTM standard.
COLIN CAMPBELL 6:50
If we use a USDA standards, these would be a top one, the South turf plot would be a sandy loam, and the North turf plot would be a loamy sand. So because this was an experiment to try to understand the status of water in the soil, we decided to install sensors around the middle of the root zone at six centimeters and at the bottom of this turf grass root zone which is about 15 centimeters. And we decided that it would be a great opportunity to combine both the water content and water potential or suction at each of these locations. What we could do with this was that we could when measuring both see what water was available to the plant and how much water the plants were using. So, combining these really made sense. We also put another sensor, the GS3 at the bottom of this engineered soil region. So the top 30 centimeters or so, sometimes 25 centimeters in the soil profiles all engineered soil based on the turf growing needs of this grass. The bottom soil there, the clay loam, this soil is actually native soil to this area. And we wanted to see what water was actually leaving the root zone. It was going down into this clay loam. So here’s a picture of us actually installing the sensors in the sidewall of a trench in that turf grass. And as you see, we buried all the sensors and then we very carefully put back the soil in the way that it had come out including the grass on top to preserve the nature of the system.
COLIN CAMPBELL 8:44
So here are all the soil moisture data from the entire season in 2015, where we are making measurements. So we started in June, and we end ended in the end of September. Now these measurements are ongoing. So we still are getting data from the sites. But I want to look at this in particular because what we’re trying to do here was create a situation where we looked at different performances of the soil, depending on different irrigation techniques. So on the left hand y axis we have the soil volumetric water content, ranging from zero up to about 40%. On the right hand axis we have matric potential from zero to negative 1500 kPa. So this is water potential. If it were a suction, it would be zero to 1500. Then these curves here we have sensor measurements at six centimeters. That’s the dark blue. At 15 centimeters, that’s the light blue and then at 30 centimeters, which is this dotted blue line here. These are all water contents. Our water potentials are up here, and they are at both six, this dark blue dotted line and 15, this light blue dotted line. And as you see, in the early portions, this water potential or this suction doesn’t change, why doesn’t change? Well, we actually were able to run kind of three different periods controlling the irrigation. This first period was what we called the calendar based irrigation. We actually didn’t touch what was already being done at the turf farm. So this was being managed by their irrigation specialists, and we were just running on the calendar of trying to refill the soil profile every two days. As we see during that period, as I mentioned before, the matric potential didn’t change. The reason it didn’t change is the sensors we were using, the MPS6s are now the TEROS 21s, they do not read above negative nine kPa because that’s the area and potential of the sensor. So from zero to negative nine as it dried, or zero to nine suction, we didn’t get any change, we won’t see any change essentially in the sensor. After the calendar irrigation we took over control of the irrigation system and started doing fixed dried periods. So we would allow the soil to dry down a certain amount before turning the irrigation system back on. That allowed us to see decreases in water content here at the six centimeter and 15 centimeter level. We started to see decreases in water potential, which means that we were actually starting to deplete the water in the root zone. Toward the end of the season, the grass was gonna go dormant anyway. And we just had these drying periods that allowed the grass to dry all the way down till it didn’t take up any water. Now why am I describing this all to you? Well, I just want to familiarize you with the data. It’s not really important as it relates to the experiment, but I want to show you that we actually did collect several different periods during which the soil dried in several different ways. And that because of the situation we’re in that we had a nice pull from evapotranspiration from the soil using the turf grass, that we’re able to create a wide range of water contents and water potential.
COLIN CAMPBELL 12:20
So let’s actually take a look at what happened in the field. So the first question is, can these soil water retention curves or SWCCs, can these create or be created from a measurement of soil suction and water content directly in the soil? So and can we learn it, are they useful for anything? So I took all the data from that four month period, and we put it into a graph here and so these dots you’re seeing here are, one is for the Southwest five centimeter site and that’s the light blue dots. The dark blue dots are the Northwest five centimeter points. So these are in two different soils. These are in the little finer soil, a little coarser soil, coarser soil on the bottom. And then I went ahead and simply used the models generated from Campbell Chiazawa to model what a standard soil type would be if it were a silty sand, a well graded sand or silt. Or if you’re using the USDA system, this would be a sandy loam, a loamy sand and silt loam. So I overlaid the SWCCs, we created in in situ with these models, SWCCs. Because I wanted to know, can we actually describe the soil type? Does it really give us information about the soil type if we generate these moisture release curves? And the answer is in fact, it does an okay job I suppose. So the south plot again, the slight dots, notice it’s matching up reasonably well with the silty sand designation here. So here’s silty sand. This is our silty sand model and our actual field soil. SWCC matches fairly well. The North plot, this well graded sand, here’s our well graded sand that light line, and not as close, but it does match up if we squint a little bit. So it looks like that to a gross approximation we can predict, so we did a reasonable job of guessing at the soil type from our SWCCs created in the field.
COLIN CAMPBELL 14:45
So what about matching up these SWCCs created in the field with the laboratory approach? So we took many, many samples out of these field soils using the HYPROP technique, this Wind Schindler technique, and we joined them up with our in situ data that this graph right here again, we just took the field data and plotted it here. Now I’ve taken up out all the wetting events, except for a few of these you can see here just for example, when the the soil is wet because the sensors don’t respond at the same speed, so the matric percentage potential sensors are much slower than the water content sensors, when they don’t respond at the same speed, then we have these dots out here as the soil wets and they’re not very useful to compare, because of course, this is also a drying curve as well. So they don’t compare very well right here. This is because our entry point to that sensor, the MPS 6 is -9 kPa or nine kPa. And so you see a different difference there. If it did respond well, I expect these slides would match up quite nicely. As you see, as we decrease in water potential, and I need a negative sign here, because we’re talking about matric potential, we see this curve match up reasonably well. And this is actually the best curve we were able to find in the experiment.
COLIN CAMPBELL 16:17
So if that’s our best, what else did we find? Well, things were pretty good. Here’s the Southeast plot six centimeters, they are close as we drive toward 500 kPa, we left the line a little bit. At Southeast plot at 15 centimeters, again, matched reasonably well, but on the low side, here’s southwest at six centimeters matched well, this is the one I showed you before. The southwest plot at 15 centimeters didn’t match very well at all, we got a big deviation. One of the things we learned from that, from these differences was that there was an issue if we let the roots dry. And so this is what I want to talk about briefly right here at the end. Can lab sampling cause problems? So when we lab sampled some of these soils, and we didn’t test them right away, we saw something, of course, the roots died no matter what, because the plant’s, the growing portion of the plant, obviously removed, and the more roots got removed from the sample, the worse our comparisons gotten. This is a sample where the technicians who ran it actually removed all the roots. And now we can see the large difference between the field soil and the lab sample. So that gives us some idea that we need to be pretty careful when we sample to even hope to get a good comparison here. How would it have done without that? Well, we did collect data, and they were much closer. But of course, we do see in all of these things, if we if we go back one slide that we underestimated in these cases when there was a difference. So if there was a difference, we always underestimated and I think that’s because the roots actually do hold some of the water in this coarser textured soil, essentially forming more of the soil matrix. So some conclusions that we can come to, it looks like these field SWCCs that we collected did describe the soil texture fairly well. So that was an interesting thing. And that’s something that we want to do more work on. It also appears that that soil and the field and lab curves agreed fairly well. Remember, we were in a course and to medium textured soils and we think that those would agree best of all, especially as we may not get as even a evaporation from finer texture soils. We’ve done this test a little bit and found that we still maybe are not quite there with this, but we’ll be coming on finer texture soils, but we’ll be coming out with more data on that soon I hope as we do more tests. So I guess the end result is at least from this experiment that it may be possible to augment these lab soil moisture characteristic curves or SWCCs with field curves. But really, we need to take more data. I’ve been talking to others who have been doing this and looking over their results. And it’s not completely simple to do this, but with a little bit of careful analysis, it seems like it can be done. So we’d love your comments or questions. If you’d like to contact METER Group’s support people, they would love to hear your experiences and also maybe help you with what you’re trying to do. So I hope this has been useful to talk about this and look forward to more conversation about creating in-situ moisture release curves.
