Transcript

BRAD NEWBOLD 0:00
Hello everybody and welcome to We Measure the World, a podcast produced by scientists for scientists.

NING LU 0:07
But what from the society point of view, sometimes if we’re not just focused on one landslide, we’re focused on the region, we’re focused on the likelihood a particular area, whether there will be a landslide or not. So the challenge is how we can expand that from the single monitoring site to a region for those isolated ones. So that becomes one of the very challenging ones. So I think I’m looking forward is basically to develop some kind of upscale method to be able to predict a region or a watershed and that way we’ll be much better prepared.

BRAD NEWBOLD 0:43
That’s just a small taste of what we have in store for you today. We Measure the World explores interesting environmental research trends, how scientists are solving research issues, and what tools are helping them better understand measurements across the entire soil plant atmosphere continuum. Today’s guest is Dr. Ning Lu. Ning Lu is a Professor of Civil and Environmental Engineering at the Colorado School of Mines. He obtained his Bachelor’s in geotechnical engineering at Wuhan University of Technology, and both his master’s and doctorate in Civil Engineering at Johns Hopkins University. He is well known internationally for his work on stresses and variably, saturated porous media, with his primary research interests in seeking common threads among basic soil physical phenomena, including fluid flow, Chemical Transport, heat transfer, stress and deformation. And today, he’s here to talk to us about his work developing an equation to predict instability and shallow slopes and his vision for the future of the study of slope stability. So Dr. Ning Lu thanks for being here.

NING LU 1:43
Yeah, thank you. Hello, you’re welcome.

BRAD NEWBOLD 1:46
So first of all, we definitely would love to get into I mean, we will get into you’re talking about your research, projects and interests and other things. But we would love to start out in, in wanting to know more about your background and how you got into science in general soil science and, and eventually into civil and geotechnical engineering.

NING LU 2:07
Well I yeah, I grew up in China. And I went through the process of cultural revolution. And I was very lucky. And when I was a teenager and graduate from high school and China, we opened the door, and out the Cultural Revolution. So I went to the college in Wuhan. Wuhan University of Technology. And after that, I don’t have too much of a choice for what I exactly want to study. But I do get involved into the engineering, civil engineering and geotechnical engineering. So after that, I have an opportunity to come to the United States to start my graduate study in civil engineering, particularly in geotechnical engineering, first get a bachelor degree and then get a PhD degree. And area I specifically study is how the soil behave mechanically and hydrologically.

BRAD NEWBOLD 3:05
Right, right. Yeah. So as as a young man, did you have any interest in soil science or, or anything along those lines?

NING LU 3:12
Well, I think that, as I just mentioned, that when I was a young, I never thought about I will study soil more maybe late today, civil engineering, how to build structures, both above the ground or in the ground. I’ve never thought about that I would get involved into the understanding of the soil.

BRAD NEWBOLD 3:32
Right, right. We have several different questions about the various research projects that you’re involved in. In studying sloped stability, and landslide monitoring, and other things like that, or at least predicting stability or instability of slopes. You have a couple of different projects that that you’ve worked on here in the United States, in Colorado, and here in Washington State. I was wondering if you could take us through those those different research projects, or do you see them as as one whole project, were they were they kind of, I guess, looking at similar questions, or were they completely different to the one in Colorado versus the one in Washington State?

NING LU 4:15
I think I’d say that the largest of the big questions, the same question. Basically, when you have a rainfall and whether it’s by the liquid which you call a rainfall or snowfall, you’re eventually those water will go into the soil. That instability will occur. So both in Washington where you mostly have a water rainfall condition where in Colorado, we have a snowfall. By snowfall, you actually want melt but it all focus on how the soil is going to response. The slope is gonna be responsive to those conditions.

BRAD NEWBOLD 4:49
Right. Right. So let’s talk about the the project in Colorado first. So you’re you’re now at the Colorado School of Mines. And so this study we’re looking at along the I-70 corridor there in Colorado. What was the what was the setup of your project? And what was kind of the main, I guess hypotheses that you were testing there?

NING LU 5:09
Yeah, so in Colorado, we have this long Interstate I-70, interstate 70, which is the one that crosses the continental divide in Colorado. In fact, it is the major highway across the US in the middle part of the United States. And one of the challenges we have Landslides occur in the high elevation after snow melt in springtime, because the snow can fall and accumulate and for a couple of meters. And eventually in the spring, it will melt and within a couple weeks, right, and so that poses challenges for the I-70 for years, since I-70 was built, the cross Continental Divide called Eisenhower Tunnel, and right on the western slope, about a couple of miles there, there was a huge landslide body. So when they dig the tunnel, they clean that landslide body part of that, but then they put all the remaining of drill bit of the tunnel into the slope and build a six segments about 1.5 mile. And that became very problematic, initially, no problem. But after a while, and the highway start to settle and it settle in a very vigorous way. In the spring time, it was accelerated by that. But then it was slowed down by the fall, it was stuck, right and then we come back in next year. So for years since 1970, there has been this kind of settlement. And as a result, it’s quite a bit of a sediment. So there is worries about it, the people, and everybody we’re actually the driver is the bond of I-70 You can feel very bumpy on that road and it’s because of the landslide. So the concern that’s here is two of them. One is in the short term. And you know, it’s it’s very unsafe, right it could have a cost of some kind. Of the long run, maybe some catastrophe will occur. Right. So that is a big major concern. And very challenging is that because of the elevation is above 10,000 feet. And in that segment, you can you don’t have any alternative interstate so interruption or do a reconstruction become very costly. Right, that’s a that’s a where that they they think of the Colorado Department of Transportation particularly, they think that this has a lot to do with the water and how the water is going to affect the slopes stability is one of the big issue. So for that particular issue, I learned over the process. What kind of mechanisms are involved and eventually, we get a look at a some remediation techniques that may come out of there.

BRAD NEWBOLD 8:01
Right now up there, you’ve talked about the different characteristics there. Of I-70 across the continental divide, very high elevation, are there any other particular characteristics of I guess, of the of the the geology or or the the, you know, soil character? Yeah, the soil characteristics that are there that that contribute to this, you talked about them building the tunnels and and clearing away this landslide body, but still, there’s, there’s things that are going on, you know, the that, that potentially contribute to that movement of the soil

NING LU 8:37
As a yes. The mainly because the when the snow melted, it actually caused a lot of groundwater level increase rapidly. And our investigation of the costs of 10 years and we find out that it’s a particularly very serious under the highway, the water table, for example, under the highway, within a season can change at nine meters or raise about nine meters. And that changes a lot of stress in terms of the what we call the effected stress. And eventually we mobilize the existing landslide every year.

BRAD NEWBOLD 9:16
Right. And so I would assume that like you said every year there’s you know, as as the that water table, you know, it’s I guess waxes and wanes and fluctuates. Yeah. Yeah, that then that definitely would would affect it the the road surface and the underlying soil beneath that. What are some potential? I guess, ways to mitigate or remediate those effects?

NING LU 9:44
Yeah, so the groundwater is wildly important, water temperature is very important. People recognize early they do a typical remediation is to on the downslope side of the highway. Put a lot of so called a horizontal drain. So drain the whole into the horizontally tilt a little bit towards the valley direction, so that way will naturally drain the excess groundwater. And however, what they find is that that technique is not very effective. So and that, and they don’t know why. And then also try another different kind of mechanism is try to try to reduce the weight of the slope by drilling a extensive caisson to remove the soil and then a rock and then replaced by some lighter material. However, that technique also does not work. So, our investigation indicated that because the water table raised rapidly under the highway, the water is actually was coming from a different place than originally thinking it’s from the upstream of the highway, because you have a very larger the slope that have covered by the snow, and when it’s melted because of it melts so rapidly and most of the waters converted to ground under the highway. So that became a part of a major issue. And we investigated through the deformation monitoring multi years, we also conduct a groundwater monitoring table, which is a more extensive borehole to monitor that one eventually confirmed our conception. And so we also propose the sort of the new method, basically, to develop a drainage system in the upstream side of the highway, that could be effectively reduced the groundwater table.

BRAD NEWBOLD 11:37
Right, right, as you’re going through in researching and monitoring, what I guess what were some of the the instrumentation or the measurements that that you have going on there?

NING LU 11:48
Yeah, so we put a inclinometer with the Colorado Transportation Department, drilled a borehole inclinometer basically, to measure the soil deformation at a different depth from one borehole to a single borehole. And we also measure the groundwater rise and decline measure the water table. And we also put us on sensor to measure the soil moisture, because the the soil, change in the moisture can change in the weight of the soil material. And we have the information about the precipitation snowfall from the National Weather Service. So with those information then we conducted an investigation.

BRAD NEWBOLD 12:30
Right. What were some of the changes? So you talked about the fluctuations of the sales? Maybe just off the top of your head? Or what were some of the changes that you saw in the in the you know, with your inclinometer and with those soil moisture measurements?

NING LU 12:43
So the inclinometer basically confirmed when and where and at what depth the landslide occured? It happened at the interface of the bedrock. And those few material from the drill bit of the I-70 the Eisenhower Tunnel. So that comfirmed weather had occured and what a table basically tell us how much how much pressure that you’re going to generate on the ground and due to the due to the snow melting. So those are the very two important information allow us to put into the model to calculate and then confirm that the timing and adapt where the landslide can reactivate.

BRAD NEWBOLD 13:23
Right, right, right. Well, let’s let’s move on to talk about your similar research projects in Edmonds. Outside of Edmonds, Washington, they’re here in Washington state it’s a slightly different geology, slightly different climate. So instead of dealing with with seasonal snowfall, and the snowpack that’s melting that’s, creating these these fluctuations. I’m assuming that they’re in Edmonds, Washington, you have constant near saturation of of sediments and of the soil there. And how, what are the similarities and differences between those two projects there in in Colorado and in Washington?

NING LU 14:01
Yeah. So actually, they in Colorado, it turns out that eventually it’s the fader surfaces below mostly below the water table. So it’s completely saturated by the water table that just changed the pore pressure. In Edmonds case it’s the opposite. Even though the water table may not be very deep. However, the failure occurred in the above the water table, okay. And so what happened is because of the El Nino and La Nina change, and you have an excess of precipitation, the slope, sitting there over the years kind of a balance and with the environment, but with the excess rain fall during the El Nino, La Nina Year, that can change the stress there inside the shallow soil. So what’s challenging is, for for many years, we can understand quite well and characterize it quite well, how does stress change under saturated condition however under unsaturated condition, it is not that straightforward, it has a lot to do with the soil moisture, it also has a lot to do with soil suction. So, what I have done is to basically put that to information, the soil moisture and soil suction together to be able to assess what’s the stress change that will cause the landslide,

BRAD NEWBOLD 15:25
Right, right. And so, so is it similar? So, you’re talking about soil suction, so then are using so in In other terms, soil su- soil suction is a a civil and geotechnical term that’s similar to for those in our audience who might recognize water potential or soil water potential. Same, same, same idea just different different terms. Um, so are you were you using the same sorts of instrumentation there in Washington as you were in in Colorado.

NING LU 15:52
We use a quite different in a Washingtons case early the US Geological Survey put some kind of instrumentation to measure the monitor in the water table mainly to focus on the saturated zone, in the saturated zone we this is the situation that will extensively use soil moisture and soil suction to basically measure the pore pressure, and the under the unsatureated condition, the pore pressure become less than atmospheric. So sometimes people connectable values, so in Edmonds case, we extensively measure the precipitation record nearby, and as well as the soil profile of the moisture and suction change. And that become the one defining variables that we can use to forecast or analyze.

BRAD NEWBOLD 16:46
Right, right. And that’s what I was gonna ask you. I mean, it’s, I guess it’s, I shouldn’t say it’s easy, but but after the fact, when a landslide occurs, and yes, you can analyze and kind of work backwards to see what happened, where are those points of failure. It’s another thing to forecast what I guess maybe moving forward in this, what have you learned that might be applicable to to better forecasting? Landslide or slope failures?

NING LU 17:14
Yeah. So in in, in Edmond’s case, actually, it’s very important for my career, because at that point, we know a lot about the suction and the soil moisture, how the suction and moisture will turn into stress that will change the landscape of stability is one of the big question. So I developed some theory, but never put into work. And so in Edmond’s case, the monitoring give us the record all the way to landslides occurs we are very lucky that the our station was all gone because the landslide, but the lucky part because we have all the data after the moment of the landslide. Yeah, so that’s very unusual. And I remember that one day my USGS called me and tell me they call me say Hey Ning, you know, the your theory works, and she said, I used the moisture record, the suction record, and use a simple equation calculate exactly the depth and the timing of our observed landslide. So, that gave me a lot of confidence about the theory I build for that moving forward and the framework basically now can equal to use for any other place. The reason is, no matter how complicated of geological or, the slope, the angles of your material is what caused landslides is a simple matter of the driving force versus the resistance force. When the driving force that exceed or equal to your system force and the slope give up. So from that point of view, in the framework, I have a builder that is basically saying, under this unsaturated condition, what is the driving force? What is the resistance force, those are very different than the saturated condition? Those all require us to use the soil moisture, use the soil suction to quantify.

BRAD NEWBOLD 19:10
Right, right. So I actually let’s, let’s dig into that because as part of this, you talked about the theory you’ve talked about now you’ve experienced this, and this has kind of led you into creating these mathematical models to describe and potentially predict slope stability and slope failure. Can you go into a little bit more detail about about your, these equations that you’re you’re trying to create here and so on.

NING LU 19:34
Yeah, so physically, if you have soil, the soil can be a liquid, a solid, or an air in general under a saturated condition, the if your soil only have a water basically or the void occupied by water and what makes the soil become a very a material to resist the the driving force is the skeleton. The stress in the skeleton will co-effect the stress and the saturated condition that could be quite straightforward. If you measure the pore pressure, you will know the overburden of the total stress, you subtracted the polish from the total stress you had effected stress or scarring stress and unsaturated condition that became challenging because you no longer can use the pore pressure or the suction, as a stress remover from as a matter of fact, that was so under unsaturated condition, because of the development of capillarity. Because of development or absorption, the soil edge of the effect they just become higher, which means more stable. However, when the rainfall come, you re-saturate them or you go to the wetting direction, it actually reduce the magnitude of that stress. So that because the slope becomes vulnerable and eventually fails.

BRAD NEWBOLD 20:53
Right, right, now now with that is, I guess I’m just trying to think of the various variables and parameters that might go into a model like that. You’ve explained it very simply, I’m assuming that there’s there’s much more much more detail that goes into that. How does how does, I guess I was thinking, you know, slope angle, or how does the the soil characteristics play into into this model there?

NING LU 21:16
Yeah, so the slope angle is the one that provide a driving force, because of the steeper that you have the slope, the higher that you have a total stress. So that provider, you said that either you analyze a slope or with a slip or not, you just look at the driving force and resistance force, the driving force is the soil self way. And the resistance force is the material property, sometimes we call soils, the cohesion. Friction angle that internally to resist those driving force. And on top of that, one is the what maybe they’re resistant, the resistance force is this additional due to the soils suction due to the soil moisture content, and they actually become part of that, to be the effective stress to resist that other one. And that is why when you have a water content, change it, you suction change it, you change it that that stress it may turn to cause suction stress, which is basically replacement of suction of the water pressure under the saturated condition. So when that quantity changes, or reduces specifically to certain thresholds, and you’re resistant force is no longer be able to sustain for the driving force much longer.

BRAD NEWBOLD 22:31
Right, right. Right. So, so what are the next steps? I, I assume that you’re not just going to sit on your laurels? And but that you’re going to move forward and try to is there? Is there room for improvement with this model? Or? Or are you moving on to other things, I guess, what do you see as the future of of this current research that you’re working on?

NING LU 22:49
Yeah. So we have been, I’ve been working with the US Geological Survey in Golden Colorado, which is on our campus. And we over the last 20 years, we have a developer, many of the sites in the United States, in the East Coast that we have North Carolina sites, because that’s where the hotspot for the hurricanes. And also due to Pacific precipitation in Colorado just mentioned that and then the west coast, the we have extensive from Washington, to Oregon, and to Northern California and Southern California, as a monitoring pro- project. And the monitoring will give us the all the detail information, both the space and time how the moisture, suction and suction stress are going to change. And moving forward, those are, basically provides some kind of a validation point to continue to refine the theory or to back up about the framework. But what from the society point of view, sometimes we’re not just focused on one dense layer, we focus on the region, we focus on the likelihood of a particular area, whether there will be a lens or not. So the challenge is how we can expand that from the single montoring side to a region for those isolated one. So that becomes one of the very challenging one. So I think I’m looking forward is basically to develop some kind of upscale method to be able to predict a region or a watershed. That way we will be much better prepared.

BRAD NEWBOLD 24:21
Right, right. Definitely something especially like you mentioned down in North Carolina. I mean, you’ve talked about dealing with in you know, snowfall mountain ranges in, in rainy climates in the Pacific Northwest. But that was another one that maybe in the future we’ll get to talk about North Carolina and or dealing with monitoring those kinds of issues with uh…

NING LU 24:43
Certin areas it’s an area where the public nature and landscape intersect yeah.

BRAD NEWBOLD 24:48
Yeah, that’s, that’s the other thing. I mean, the the intersection, I guess this this kind of gets into the, you know, the Human Geography of it all as well and, and how, you know, of course, the There’s gonna be human impact with these landslides. The, I guess the closer they are to, to human occupations and those kinds of things, if they’re landslides out in, in the middle of nowhere, it doesn’t matter too much. But But definitely, if we’re dealing with landslides, they’re on I-70 Or if we’ve got landslides due to hurricanes in North Carolina, or especially as we’ve seen, in California, along the Pacific coast there, where you have that, you know, potential arid environment that then gets inundated, and you have lots of people living in in hills or long slope sides, you’ve got lots of highways, freeways and and other things that are there that then get get impacted. And so, any type of I guess, monitoring or prediction, slopes stability or slope failure prediction would go a long way to to minimize the the, I guess the human toll or the impact there as well. What are some current challenges that that you have faced in trying to just I guess, simply characterize soils within geotechnical engineering? And are there ways that we can, you know, make those those measurements or those characterizations better in your opinion?

NING LU 26:13
In terms of the in terms of what in terms of landslide prediction or in terms of the how we can better understanding the soil soil behavior, hydrological mechanical? What?

BRAD NEWBOLD 26:25
Yeah, so I mean, let’s talk about the first the first just in general terms, and then maybe we can talk about the landslides in more specifically, yeah.

NING LU 26:33
So I think the measurement of soil moisture, it’s, it’s a technology that has been in my view, quite to the degree that we can feel quite confident to understand the soil moisture, but the suction is not so. Soil suction become remaining, still very challenging, particularly in the field condition. For now, we only have a director method called the Tensiometer that only give us the suction range in a few 100k kilopascal. However, the nature the suction can vary, vary over six order magnitude, everything from zero to gigapascals order. So the sensor in that aspect is somewhat limited. And many of the phenomena could probably require much more wider to to be able to categorize soil condition or suctions. So that’s one of the areas that I think that remains can be improved. And doing that would greatly improve the ability to monitoring slope and provide the other thing is because of Mother Nature is complicated, now, the two slopes are the same. So providing some kind of forecasting what early warning system, it become more challenging. However, we we have a developer in different discipline, different kinds of techniques or methods to recuperate for that, for example, we have a much better idea now about the more sensing to know that topographic and topographic can provide a very powerful information that geological material found in most sensing and that provide a very good about the history of those particular areas. Right. And those information probably can also be used that to, to forecast landslide. There are some of the areas these are the things I think that can do a better job. Okay, in the future.

BRAD NEWBOLD 28:37
All right, I guess I should have maybe asked this a little bit earlier, but you’d mentioned talking about effective stress as kind of being a center point for your research. When When did you move or when did you first decide you wanted to get involved in in exploring this concept of effective stress?

NING LU 28:54
Yeah, in 1990s 1993, I moved from Washington, DC area to Colorado. And so that is actually the first time I start to expose a swelling soils and slope stability. So, before that, I have a silent limited exposure to unsaturated underground nuclear isolation. So, it really the issue about the slope of stability, the issue about how the soil can swell, cause the damage to civilian and residential house is the wild turning point I would say that for me to start to look at more serious about the effective stress, because those both of the case I just mentioned, swelling soils, slope stability, in many cases that occur under pairtal saturated condition. So that require extensive knowledge about the about how we can develop an effective stress, which has been very much a limited on the conservative side limited to saturated soil condition, but in reality, many of the infrastructure is under partially saturated conditions. Right. So that’s, that’s the turning point back to 1990. So I started there.

BRAD NEWBOLD 30:10
Okay. All right, in talking about being able to monitor, it’s there a lot of a lot of these projects, for various, you know, municipalities or local or state governments might be a bit more cost prohibitive to be able to install monitoring stations all over the place, like like we would prefer, right. I guess it just kind of was was crossing my mind when it when you’re dealing with these these models? Are their primary drivers. Would you say for instance, that soil suction would be a primary driver, say, if people wanted to do a, a simple monitoring solution? Where installing, you know, soil suction sensors or something along those lines? Would that be a shortcut? Or is it is it much more involved in when you’re dealing with monitoring and potential prediction?

NING LU 31:07
I think the from my point of view, the most of the monitoring is just give us the firsthand to develop a predictable theory. And ultimately, the predictable theory should be the variable that very easy to be measured. So from that point of view, and, you know, in the future, I don’t think everybody can, every slope can put a lot of a sensor right. So we need to develop a some kind of easily access information. Of course, the accessibility depends on technology. So for now, let’s say I just mentioned that topographical, the geological material, the vegetation, those information could be a handily obtained for most of the region. And with the high resolution, if you have a technical remote sensing, so eventually, under precipitation, also to the climate change, its somewhat predictable in a way, at least in the short term. So eventually, I think what will be the ultimate goal is to develop a tool, the model, confirm what they’re using those access model information. And of course, soil suction and moisture, was still be as a rule to as a tool to build those information.

BRAD NEWBOLD 32:19
Right, right. I think that’s our time for today. Okay, I really appreciated the discussion. I know that I’ve learned a lot and didn’t know a lot about this. And so it’s been great to be able to talk with you. So thank you for coming by.

NING LU 32:30
Thank you for having me.

BRAD NEWBOLD 32:32
And if you in the audience have any questions about this topic or want to hear more, feel free to contact us at metergroup.com or reach out to us on Twitter @meter_env. And you can also view the full transcript from today in the podcast description. That’s all for now. Stay safe, and we’ll catch you next time on We Measure the World.