What your soil moisture data are trying to tell you

If you want accurate soil moisture data, correct sensor installation should be your number one priority. When measuring in soil, natural variations in density may result in accuracy loss of 2-3%, but a bad installation can potentially cause accuracy loss of greater than 10%. Poor installation is the most common source of error in soil moisture data, but there are techniques that will ensure a perfect installation every time.

How to get it right

Sensor installation expert, Chris Chambers, explains why you need a smarter soil moisture sensor installation and how to achieve it.
Learn:

• What good soil moisture data look like
• How various installation issues show up in your data (i.e, air gaps, a loose sensor, soil type change, depths crossing)
• How to ensure an accurate installation
• How the new TEROS Borehole Installation Tool reduces air gaps and site disturbance while improving consistency
• What other scientists are doing to ensure a correct installation

Next steps:

• Discover TEROS soil moisture sensors
• Check out the TEROS installation tool
• Request a ZENTRA Cloud live demo

Questions?

Our scientists have decades of experience helping researchers and growers measure the soil-plant-atmosphere continuum.

• Talk to an expert
• Request a quote

Presenter

Chris Chambers operates as the Environment Sales & Support Manager at METER Group, the world leader in soil moisture measurement. He specializes in ecology and plant physiology and has over 10 years of experience helping researchers measure the soil-plant-atmosphere continuum.

Transcript

BRAD NEWBOLD
Hello, everyone, and welcome to Soil Moisture: Why Installation is Everything and How to Get It Right. Today’s presentation will be 20 minutes, followed by 10 minutes of Q&A with Chris Chambers, METER Environment General Manager here at METER Group. If you have a question for Chris, type it into the questions pane at any time during the webinar. We’ll be keeping track of these to answer during the Q&A. So please don’t be shy and submit those questions. We’ll also be sending out a link to the on demand webinar, as well as the slides for you to review as soon as they’re available. So without further ado, I’ll hand the microphone to Chris Chambers.

CHRIS CHAMBERS
Thank you, Brad. So, as Brad said, my name is Chris Chambers and I would like to talk about installing soil moisture sensors today. And the reason why I feel it’s so important is because it’s basically the foundation for the data that you’re going to collect. If you have a poor foundation, or if you have a poor soil moisture sensor installation, then it really makes the rest of your data interpretation difficult. So we’ll spend some time on that today. And I just want to reiterate, as Brad said, please don’t hesitate to type in any questions for me, I’ll get to those at the end of the seminar. So our goals, I want to talk about some background information that I think everyone should be aware of when they’re installing soil moisture sensors. We’ll look at some data, some actual customer data that I’ve consulted with customers with over the years, and look at what installation trouble can, how it can manifest in your data. And then we’ll talk about some of the priorities to think about when you’re on your own soil moisture sensor installation. And then I’ll wrap up with the METER contribution and how we’re trying to improve this section of the science.

CHRIS CHAMBERS
So just a little bit about me, and I have 10 years of experience working with soil moisture sensor data from the viewpoint of data quality. I started with Decagon Devices about 10 years ago, and one of my main objectives is to help customers get the most out of their data, whether it’s soil moisture sensors, or weather data, canopy data. And so when the customer sees something odd in their data set, I’m one of the first resources here at METER. I’ve installed hundreds of sensors over the years, I want to say thousands, but I’m not sure if I’ve topped that 1000 point mark, but I’ve installed a bunch of sensors so I have some experience here. First off, a bad installation isn’t an indictment on you, it doesn’t mean you’ve done anything wrong in particular. So, you know, I don’t want to start from that end, but it does have enormous impacts on your data quality. I’ve personally made almost every mistake our customers have and I’m going to show you one of those later.

CHRIS CHAMBERS
So to start with a background, we’ll talk about volumetric water content sensors today and those units are in meters cubed of water per meters cubed of the total soil volume, and that’s a fraction between zero and one. It can also be expressed in percent volumetric water content, and that’s if you multiply the meters cubed per meters cubed by 100. And we’re going to talk exclusively about sensors that use the electrical properties of the soil to measure water content. You’ve probably seen these broken into categories like TDR, FDR, or capacitance sensors. These all have different methods for measuring the water content, but the basic principle behind them is the same. The physics are applied in different ways. So these sensors all emit an electromagnetic field into the soil. And the main factor that affects the electrical properties of the soil is the water content. The soil minerals don’t really change on a day to day basis, it’s the amount of water in the soil. And key to our discussion today is that most of the sensitivity of the sensor is within a few millimeters of the probe.

CHRIS CHAMBERS
So here’s a conceptualization of what that EM field looks like. Now, this is an oversimplification but I think it gets the point across. So with an electromagnetic field, basically, we’ve got these field lines emitted in the soil, and note how they’re much denser right next to the sensor. This is in EC 5 over here in the corner and so this is an approximation of the relative strength of the electromagnetic field. So what’s happening right next to your sensor is going to have a larger influence on the response of the sensor than things that are happening out here. Now anything within this field is going to affect the output of the sensor, but it is important to note that what happens right here is going to have more influence on your data in the long run. And here’s another way to visualize it, this is one of our TEROS 12 sensors. And so the full field encompasses a volume about like this, and in this case, it’s just about one liter of soil.

CHRIS CHAMBERS
Now, one of the second things that’s important to understand is what the sensor is actually measuring. Our sensors produced here at METER are capacitance sensors, and the main thing that affects the capacitance of the soil is water. Now, I’m going to, I want to draw your attention to this dielectric permittivity. It’s a unitless measure, basically, it just tells you how good of a capacitor your soil is. And here’s the dielectric of your soil minerals and organic matter. So this is the basic composition of your soil and it’s not changing, this is going to be constant from soil to soil unless you have treatment, or compaction, or something like that, this is almost always going to be the same. In air, this is going to be the extreme end of what your sensors can detect and it has a dielectric permittivity of one. And water is on the opposite extreme. So this is basically your range of what your sensors can read from one to 80. And you see that air is going to draw it towards the extreme dry end and water will draw it towards the extreme wet and so bear this in mind as we as we move forward.

CHRIS CHAMBERS
Okay, let’s take a look at some data. And I want to thank Daniella Carrijo for letting me share this dataset with you. This is a rice field, and here we have three treatments. Over here we’ve got the volumetric water content in percent, and we have number of days on the x axis here. And so here you can see, we have three treatments, the solid black line is continually flooded, so it’s a rice field that’s flooded all the time, and then they have treatments where they’re drying it down to 35% water content and 25% water content. And let’s look at the beginning. So I’m kind of cheating a little bit here because this looks like a great installation. In the beginning we’ve got three different sensors, and these are 10 HS sensors installed about seven and a half centimeters into the soil, so we’re up near the surface. And here we’re basically just looking at the sensor to sensor variability, all about within 1% of each other. And you know, consistent soil type, this looks like a good installation right here. This is what you want to see if you have the same water conditions in the same soil type. And then we get a dry down. This is where, here’s our continually flooded field still trucking along at saturation, and then we’re pulling the gray lines down and then they wet it back up, they flooded the field again. And this is what Daniella noticed and had questions about and what’s happening right here.

CHRIS CHAMBERS
Notice that our saturation level of the treatments is now much higher than what we see with where they started and also with the perpetually flooded control. And you know as we progressed the soil type comes into questions like this. And it turns out that these are installed in vertisols and a vertisol has a lot of shrinking and swelling capacity. And here you can see this is a typical vertisol. These photos were provided courtesy of Leo Rivera who works here at METER and has worked in these types of soils before. So you can get these large cracks opening up as the soil dries down and that’s exactly what’s happening in Daniella’s data set here, is you see more water being able to get right next to the sensor. Remember that water has a dielectric of 80 and so when it’s completely saturated, you’re holding that higher dielectric value right next to the sensor. And the fact that those soils have opened up and provided a gap for more water to be right there, we’ve actually, this is a really good example of what can happen with air gaps in the soil on the saturated end.

CHRIS CHAMBERS
And, okay, let’s look at the dry end a little bit. This data set was shared by Quinn Campbell, who works at the USDA ARS Newingham lab in in Reno, Nevada. So we’re going to shift from a perpetually saturated system over to soils, that get very, very dry. So these are highly porous, coarse substrate soils. And here we have volumetric water content in meters cubed per meters cubed on the y axis, and time on the x axis. And this is a really typical way to view soil moisture data is a time series. And here, we’re going from October of 2015, all the way up to May of 2018, just recently. And yeah, the dark blue line is, these are EC 5s, the dark blue line is underneath the plant, and these light blue lines are interspatial just kind of out away from the plants. And I don’t see any warning signs in these data, they look look pretty good. Got a long stretch of very little precipitation here and we’re getting down to really, really dry soils, but for the most part, yeah, it looks like we’re trucking along great.

CHRIS CHAMBERS
And then on a nearby site with the same experimental design, now bear in mind that these are installed very near the surface, all of a sudden, we get quite a bit bigger gap there, we’re missing some data here, you know, paying attention to that, quite a bit bigger gap between what’s happening out here beneath the plant, and what’s happening here out in the open spaces, and I think this is a red flag. Here we’re dipping below 0 percent VWC here, and by definition, that should be the lowest that soils can possibly get. But bear in mind that air has a dielectric less than that of soil, so what this likely means is that we’re getting air influence in here. This can also be a calibration issue, but I think that less likely that it’s calibration, since we’ve got the same sensors installed nearby, and performing slightly differently, so I think this is an indication of either an air gap or it’s installed too near the surface. And these are very shallow sensors installed within the top five centimeters so it could actually be too close to the surface and emitting that electromagnetic field into the air as well. So I’m not 100% certain what’s happening here. but in any case, this looks like a case where the installation is either too close to the surface or there’s an air gap right next to the sensor.

CHRIS CHAMBERS
Okay, let’s talk about a good installation a little bit. I would like you to all meet Kevin Hyde. Kevin works with the Montana Climate Office and we chat quite a bit about installations, about sensor, about weather. Montana Climate Office is a mesonet scale, where we’re working with them in a lot of instrument instrumentation across the state of Montana. And Kevin prefers the open pit installation method, which is fine, it has a lot of virtues, and here’s a sensor, as you can see in here, installed in a profile, the cables are all nicely bundled away and protected by conduit from rodents, anything like that. And if we go to the next slide, check out this soil to sensor contact here, this is really what you’re looking for. So I don’t worry about Kevin’s installation when we talk about data quality issues. And this is the single most important thing, if you do nothing else, if you focus on nothing else during soil moisture sensor installation, getting good soil to sensor contact is the most important thing you can do and I can’t stress that enough. Because again, those first few millimeters next to the probe are going to have the most influence on your reading.

CHRIS CHAMBERS
Okay, this is what not to do. This is a recent installation that I did. I’m not ashamed of it, these are rocky soils, I pushed them in by hand and, you know, field work just sometimes throws you curveballs. And here, this sensor in particular, I just couldn’t push it in there enough, you know, it was too hard to push in by hand. And here I am with this air gap that’s close to and not quite an inch, but this is definitely going to have an impact on the quality of my data coming in here. And so, you know, this is the best I could do to get the sensor in, I tried moving it around in different places, hit a rock every single place I went. So just kind of unfortunate here, what I had to do in the end is just try to pack soil around it. That’s not an ideal situation but in some cases, that’s just the best that you can do. Because having these, just kind of filling in my hole around these, this is a four inch borehole, is really, you can’t rely on that to get the job done. So I want to hammer home that idea of sensor to soil contact. Again, they should look like this all nice, tight, snug, haven’t wiggled them or anything like that, you just push them straight in. And this is not what you want to do, okay, this is what you want to do.

CHRIS CHAMBERS
Okay, now I want to talk about the METER contribution and how we’re trying to improve this aspect of soil moisture data. So a couple years ago, we actually had a little contest where we turned our engineers loose and gave them some time to design something to install sensors in a profile, and they came up with some pretty crazy stuff, but this worked out really well. We’ve designed this installation tool to install in a four inch borehole and drive sensors into the sidewall so that you get maximum soil to sensor contact. Here’s a TEROS 12 down here at the bottom. It all fits snug right in there and it drives the sensor straight into the sidewall, no wiggling. It really is a remarkable way to install sensors. And we have a ton of mechanical advantage up here, not literally a ton, I’m talking about a figurative ton here because I haven’t actually measured it, it’s a lot of mechanical advantage. So you get this really easy lever to push, the whole thing is two meters long, just to give you some context, and here we’ve got a guide so that you can set it at one meter, 60 centimeters, two meters, whatever, and then just push the thing down into the hole, drive your sensor in, raise this part up or move on to the next plot.

CHRIS CHAMBERS
Not only is it better installation, it saves time as well, and we’ve had great success on installations with this tool. So minimal disturbance, and that’s something that you saw with Kevin’s installation, and that’s just something that comes along with the pit install. This is the installation tool in a four inch borehole and you can see that its impact on the overall site is much smaller and you can get all the sensors installed into the native soil in just one borehole. Here we have our extremely photogenic Leo Rivera displaying it in action. And so it just drives it right into the sidewall. It’s really easy to use, it gets a good installation, and like I said, it saves a ton of time.

CHRIS CHAMBERS
Now it’s not the perfect tool, I do want to talk about one important thing about the trench method. I had this chat with Kevin, not too long ago. And I said Kevin, you should use our installation tool. And, you know, saves time, maximum soil to sensor contact, blah, blah, blah, all that stuff. And he’s like, yeah, but I need to know the soil type. And that is not you know, he’s right, you do have to know the soil type. Soil type always is important to the data that you’re going to interpret for water content sensors. And that’s something that is also important to bear in mind in your installations and you can see here, you know, you can see his whole profile as he’s installing sensors here. Now that’s not out of the question with a borehole either. Here’s a core soil that I made a borehole in and we pick up a type change here, this was about 45 centimeters down and so this is something I’m not familiar with this particular soil. So it’s something that we bagged up to bring back to the lab, to run a soil moisture retention curve on it and get a little better idea of what’s happening here. It’s coarse, I can tell you that. But, you know, so you still can, and I recommend do, collect soil type information or dig a pit nearby so that you’re still getting that soil profile and know what you’re installing your sensors into.

CHRIS CHAMBERS
And then the last thing, it’s not a silver bullet. The sensor down on that installation tool is in between a rock and a whole heck of a lot of mechanical advantage. So you can actually mash sensors, it’s a great tool, but you do have to be aware of rocks as far as the sensor is concerned. So you know what rocky soils are just challenging in general, as I displayed earlier when I was trying to install one by hand, so rocks are challenging.

CHRIS CHAMBERS
Before I get to questions, I want to acknowledge Kevin, Quinn Campbell, and the Newingham lab at USDA ARS, Leo Rivera and Colin Campbell within METER for their slide contributions, I drew on Colin for our theoretical depiction slide, and of course for Daniella at UC Davis for sharing her data, which is just a great data set, that’s hard to find something similar to that. And that work was published in Field Crops Research just recently in 2018, so I recommend that you guys get in there and check that out. So feel free to contact me directly if you have any questions, my contact information is available down here, or you can always reach us at [email protected]. And I want to thank you guys a ton for tuning in, and I appreciate the questions coming in.

CHRIS CHAMBERS
Give me a second here to take a look at these and see what we’ve got. Okay, so first question we have, is it true that you can rent the installation tool, you send it to us and then we return it?

CHRIS CHAMBERS
Yes, we do have a rental program for that, we realize that it’s not a trivial expense to purchase an installation tool, and many people need it for two weeks to do their installation and then are done. Right now the cost is $500 per month and we have discounts for quantities of sensors, and all the shipping expenses are paid in that. So if you request it for a monthly rental, we’ll ship it to you, you can use it, and then there’s a shipping tag in there to just send it back to us once your installation is complete. So feel free to take advantage of that opportunity. That’s only for the US at the moment, but contact your local distributor if you’re not in the US, because they may have this available as well.

CHRIS CHAMBERS
So another great question here is how moist, soft, does the soil have to be so that the sensor can be driven into the soil?

CHRIS CHAMBERS
This is a great question and I’m not gonna lie to you. I don’t know the full limits of it yet. So far, because you’ve seen the one slide, you have enough power in there to match the sensors, I’ve never matched a sensor without installing it into a rock. But we’ve got August coming up here pretty soon, we’ll get some more installations in then. I don’t have a complete answer, and it does matter. For those of you that installed many soil moisture sensors, installing when, like for here on the Palouse and in many places where you don’t get a lot of summer precipitation, or if you have dry clays that can just get super hard, installing a sensor in before your seasonal rains come can be extremely difficult. And I don’t think it’s going to stop the installation tool, I think even in August you’re going to be able to install it. I personally haven’t done that yet though so I’m going to stop before I make any outrageous claims. If you install the tool and bend one of the sensor pins for a TEROS 12, is it easy to detect by looking at the data? Most of the time. Most of the time, you’ll be able to tell if the sensor is actually broken. If it’s bent just a little bit, it’s probably not the sensor, the sensor may continue to work, and it will give you good data. We have actually tested that some, but I haven’t had any cases where we’ve bent pins with TEROS 12 without hitting rocks.

CHRIS CHAMBERS
There’s a couple questions here about recording the presentation. Yes, it is recorded, it will be available, it’ll be available online. There might be some delay but you can come back and check out the presentation, send a link to your friends, particularly if they’re installing soil moisture sensors for the first time. But yes, it will be available later.

CHRIS CHAMBERS
Let’s see we have a specific one here. Sometimes there is empty cell or asterisk asterisk asterisk stars instead of data. I am going to refer you to our support, please contact [email protected] if you encounter that.

CHRIS CHAMBERS
Really interesting question here about what would data look like if you installed just next to a rock, given that rock dielectric is different? So rock dielectric, it actually isn’t that different unless you have something weird, like a lot of iron content or something like that, but if you have a lot of iron content in your parent material, you most likely have a bunch of iron content in your soils as well, unless you have less or soils that are wind transported or water transported or something like that. So, you know rocks can be tricky, but if your sensor is installed next to a rock, it most likely won’t make any difference, except that it’s displacing pore spaces for the water. And really that’s real data. We’ve actually run some calibrations in the past where we’ve run a calibration with rocks and without rocks, and there was no appreciable difference between those calibrations. So it’s not it’s not a blanket issue. But in many cases, rocks won’t actually have any effect on your soil, other than the real effect of not having displaceable pore spaces. So I really appreciate that question, it’s an interesting thing that we do run into from time to time.

CHRIS CHAMBERS
So we’re out of time. Like I said, I want to thank you all for tuning in, I really appreciate your interest in this matter. And you can always reach [email protected]. Hit up our website, and we will get to all the other questions that have come in via email. Thank you very much.