EGU HydroTalks: Thom Bogaard on landslides and the IAHS-HELPING decade

[music] Hello everyone, welcome to the hydrotox podcast. I am Archetta. I am the incoming early career scientist representative of EG's hydraological sciences division and an environmental scientist and with me is uh I'm Melissa. I'm the outgoing early career scientist representative for the HS division and I'm an ecosystem scientist and in this episode we are really pleased to welcome Professor Tom Bogart from Del Technical University. Professor Bogart's research sits at the intersection of hydraology, geomorphology, and natural hazards with a particular focus on rainfall triggered landslides and regional landslide early warning systems. His work combines field observations, tracer studies, modeling, and interdisciplinary approaches that link ecology, geology, and water sciences to better understand and mitigate landslide hazard worldwide. Professor Bogat is also the chair of IHS helping hydraological decade and holds visiting professorship at Cassad University Bangkok, Thailand. Welcome professor Bukard. >> Thank you for having me. >> Brilliant. Okay, so we're going to start with a really basic question. How exactly does water contribute to landslides? That's very good question because for many hydraologists that's not that obvious. Um [snorts] so the basic thing is that rainwater infiltrates the soil or comes from other places to a certain location in the landscape and builds up a local groundwater level and uh the fact that there is a lot of water basically lifts the soil a little bit. So the soil grains are less connected reducing the strength and mobilizing a landslide. So uh contrary to a lot of people think like oh it's wet and it get very heavy and now it goes down. That's not the mechanism of a landslide. The mechanism of a landslide is that the internal strength of a package of soil is reduced where their water is basically buoyancing part of the weight of the of the of the slope. >> Okay. So besides just water because it doesn't seem that landslides are just water initiated, they're a mix of complicated and interconnected dynamic processes. And so there's hydraology, ecology, and geomorphological factors. And all three of those are quite sort of different disciplines that are all sort of working together. But how do each of those factors contribute to sort of triggering a slope failure or preventing it? Yeah, that's an excellent question and that makes landslide hydrarology such a super super nice and interesting field for hydraologists to work and uh it also links nicely to to my personality. I love to work on hydrarology in an interdicciplinary way and landslides are very very interdicciplinary. So vegetation is a very very important aspect of landslides mainly because roots um anchor and stabilize slopes. Uh root strength really helps keeping the soil together or keeping a soil layer towards the uh bedrock. That does not mean there is no landslides if there's vegetation on because then most likely the slip surface will go a little bit below the uh the root zone. For the rest, vegetation um influences heavily like we all know the water balance of the slope. It uh in it influences the infiltration capacity. It uh regulates um transpiration. Um and if it is a really nice ecosystem in the soil, it also has a lot of influence on how water flows and is stored in the subsurface. And that interaction indeed um influences then the timing and the magnitude of a landslide. The geomorphology it's you could say it's a consequence. I mean at some point but the slope of course is important in the Netherlands where I live. Not so many landslides as the slope is relatively close to zero. So that doesn't make you uh give you a lot of risk for for landslides. um the lithology. Yeah. That determines uh how thick is the real riolic riolic layer, how um permeable is the soil, how um much strength it can generate. So this whole interplay of natural factors is super important in in landslide work, but it also makes it extremely nice. And then we live in the anthroposenic times. The humans are playing a huge influence by making roads, steeper slopes, adding drainage systems which sometimes fail. Beautiful research uh recently on failed drainage systems which at the end of the pipe create landslides and debris flows. So this whole interaction makes it a really interdicciplinary field and um and fascinating to work on. and adding something else to the mix of factors that sort of bring forth a landslide. I was really interested to find out that water chemistry can actually play a part. So it's not just what the water does, it's how the water is. Can you explain a little bit more about that? Yeah, that was that was a really um also for me an eyeopener when I was a young early career scientist working on my PhD and I had this this idea of using water chemistry to see where the water was coming from a bit of classical hydrarology. If I have a Marley's area, I will have a bit of sulfate. If I have only limestone, I will have a bit of more calcium. you know it was it was a basic um idea of understanding the behavior of my landslide hydologically. [snorts] Uh but slowly I I got to see and to learn that besides using it as a tracer to really look at the water balance of a slope because a landslide is basically a water balance you know water in water from the sides in and also how it drains. So it's it's about filling, storing, and draining. And if the drainage is fast, you will never have a landslide. If the infilling is slow, you will never have a landslide. So it's all about how much water is stored. But besides that, the strength of the soil can change a little bit because of chemistry. And the most famous examples are the quick place in um in the Scandinavian and northern areas of the world where we have a lot of um marine deposits which slowly get a bit more fresh water because of rain infiltration and because of it the internal strength reduces enormously and then with really the slightest trigger it can generate big big landslides. So [snorts] chemistry has an has an important influence of the internal molecular structure and the internal what they call then angle of internal friction the strength of the of the soil. It's [snorts] a really nice geotechnical um uh type of analysis a lot of lab work typically done uh but I for me it was by accident I came across it by using it as a tracer. >> That's very cool. It's a very good accident to have. [laughter] >> It's a nice accident to have. Definitely. Yeah. Yeah. Yeah. >> So, is it possible to predict a landslide? And if it is or if it isn't, how is it how can you do that? And is it challenging to predict a landslide? >> Yeah, if it was possible, I think everybody would already have done it. So, the answer is it's very hard. Um but of course this is the key business of geotechnical engineers and environmental engineers working on on slope stability and on predicting landslides in in broad lines you could see that it it there is a lot of similarity with flooding. Um so what you see typically if we see a landslide as one big part of a slope that fails you will see we can uh we can um classically use physical deterministic models um calculate infiltration capacity do a lot of lab test and slowly start to understand and be able to numerically model a slope uh and then you could predict it based on the amount of rainfall more practical in uh in this type of big deep-seated slope and slope failures is that we use surface monitoring points and have early warning systems related to it. So that's that's for the the single deepseated landslides. >> [snorts] >> much more challenging and a field that I I really like to work on for the last years is can we predict it over a region then the exact location is less important. But you basically say this basin with a certain amount of rainfall and a certain amount of anticedent hydological conditions is more or less likely to get to see landslides. And then if you overlay that one with um the susceptibility map. So the map where we know it is more likely to have a landslider. So the spatial probability of a landslide to occur then uh we can have warnings. If you would say would that be a good prediction? Now the landslide world predicts many more landslides than occur. So most of the time we say like there are 100 landslides and then maybe 0.101 occurs. So we see a lot of potentially unstable places in in the earth. And that's I'm happy with it because we we don't want to predict too many landslides correctly. A landslide is still a um an anomaly. It's still something which happens um uh it's an extreme event. And that makes sense. You know if you look at hydrarology and geomorphology the landscape is still in some kind of equilibrium with its climate and also the vegetation the land use is somewhat in equilibrium with its climate. So that's why still at this stage most of the time landslides are anomaly not happening that often but we have some tools to predict it. So, sort of leading on from that, you've been active in Thailand on near realtime landslide early warning systems and you've been using weather radar information for that and could you sort of explain how the weather radar information plays into that and whether or not it's a common practice that this is sort of implemented? >> Yeah, that's a really excellent question. Um so one of my joys in in academic life is is working with my colleagues in Thailand. You know it's a real big uh problem over there. The infrastructure really suffers and the climate um has big high intensity rainfall events spread all over the place and those are accelerating and increased in number. Now I always thought like you know you just take the weather radar and you have the numbers but first of all weather radar typically gives a scale like low intensity to high intensity does not give a number. So to go from a reflection percentage an amount of energy getting back from the weather radar you have to to calibrate that one to a real number. Is it 10 mm per hour? Is it 20 mm/ hour? What is it? Now, specifically in Southeast Asia, not only in Thailand, but in Thailand, and we do most of the work, you will see that even frontal rains have this high intensity cells. And only the high intensity cells really matter for landslide occurrence. They are the places where flash floods occur, where landslides occur. And you want to predict them correctly. Now what we do in Thailand is that you um every six minutes you get a new acquisition from a radar or sometimes two. You directly interpret them based on hourly data from MET stations. So after one one and a half hour you have a kind of bias corrected intensity spatially distributed over the domain of the weather radar. Now you do that a few hours and then you can start predicting where it's going. You know the things that we use in the west to know if we bike um dry or wet to work there. We then use it for uh for forecasting where the water will fall. We can do that three maximum 6 hours in advance. Uh and that one is now used to have a kind of early warning prediction for landslides in regions in in Thailand. um you cannot prevent a landslide from it. You cannot stop it. I mean, but at least you can issue the warning to get out of the way or to not go on the road. Uh stuff like that. And but it's it's amazing to see uh with what type of energy and and technical skills the colleagues from from Thailand are really working on this type of of um products and how they help to make their country safer. That's really very stimulating to collaborate with >> a cheetah. >> Yeah, I that's very interesting and I have two follow-up questions. So one is predicting where the rainfall is going to happen. Is that enough or do you also need to predict the amount to get a better um prediction? And the second one I had in mind was how well is the prediction system like how many percentage of time does it work or not work? >> Yeah. So normally regional and it's it's an excellent and and very very uh correct question. Um so of course you need to know where the high intensity cells are because those are the the the the trajectories where all the waterfalls that you have to have your your issue. Um you could link it with a physical deterministic model, overparameterize it and try to um basically calculate per 30 with 30 m how much groundwater you have and what is the um stability of that piece of a slope. That's very time consuming and computational heavy. But more and more we work in that direction. Traditionally and much more practical, we look at a section or a subcatchment and we basically say this cell will enter this subcatchment and then the whole subcatchment is under risk. So we don't discriminate then the exact location. We just say like this area or this subcatchment or this part of a community is under higher risk. Um how often are we correct? um we very often say it is likely that there will be a landslide and then nothing happens. So we are um in the in the landslide world especially for the regional assessment of there will be a higher probability of landslides we are very often wrong. So what the current state of research is not to predict the landslides because typically there always will be one or two that we are correct but we cor predict a lot of times when there's nothing and nothing really happened. So to reduce the amount of false alarms. So traditionally if you look at lumped regional landslide early warning systems we overworn and that should be reduced from a communication perspective and that's where currently the lot of research is going and maybe to add on to that that's where hydrarology is so super important because also now we discussed the rainfall as a predictor but rainfall itself is is a really you know it's an okayishious predictor but the real predictor is the combination of the antecedent hydraological condition plus the rainfall. >> Now you're all hydraologist listening to this talk and you know it's not easy to predict how much water is stored where in the catchment. So you get a kind of of probabilistic framework of saying how wet the catchment is and then if the catchment is wetter it's more likely that a rainfall event will trigger a landslide. Now that combination that's that's where hydro where what I call landslide hydraology. >> Cool. It's really interesting to hear all this. Um landslide disasters can sort of still cause major losses when they do happen and that can be sort of in the loss of human life but also in terms of loss to a nation's GDP or productivity. Um, where do you see some of the biggest gaps being in terms of mitigating the losses from landslide disasters? Like is it to do with scientific understanding of landslides and how they occur or is it policy or planning or sort of what's your take on this? >> Excellent. Um you know we are in science and we always try to say that we have uh lacks in our and gaps in our knowledge but to me I think currently uh and already for maybe a decade. We are at the stage of how do you get all that information across? How do you communicate it and how make do you make sure that actions can be taken to be prepared or to prevent? Um first of all a flood a landslide a flash flood can never be fully uh prevented. We have to know that uh a gem of process is that at some moment in time part of the slope will end up in the ocean. You know that's the long-term geological process that will be there. But you want this process to take place in a way that it's not damaging or influencing um us as a society. Now bringing near real time. So I maybe I have to start with the beginning. One thing which we do and is I think yeah it's implemented in most places in the world now is that we make landslide susceptibility maps. So we basically say where the risk is higher or lower. You know for a flood it's around the river typically not in the mountain top. For a landslide it is typically in areas which are between the 12 and 30° slope. If it's too steep you get to rock falls. Uh so you can make maps where we expect them and which are trained on the databases that we have um uh collected over the last many hundreds years. And now you see that remote sensing is really helping us to improve those spatial maps. So the spatial planning is one thing to really to really work on and then hopefully society uh will implement and and make sure that people really build according to those roots. I mean that's a challenge. Let's be very very honest here. Then on the prediction part uh I think we mainly are working on this hydrometerological thresholds regionally uh to make sure we have less false alarms and that it becomes easier for a civil protection agency in the country to take action. Um good examples are already the the for example impact based forecastings of typhoons. You know typhoons will create landslides. They are so extreme they will create landslides in in north Thailand in north in China in in you know in many many many places. Uh but those you can warn now we really 3 4 days advance can say how likely it is that this typhoon with a certain amount of rain will enter an area and then you don't need me anymore to say there will be landslides because that's so much water it's it's risky. So the gap for society to make society safer is really about using the data. It's really about using the data and ask scientists to help making sure we are not too often wrong in our predictions. >> Okay. So in terms of how the world is changing as a response to climate change and how that is shaping rainfall patterns and maybe shifting hydraological cycles, how do you see this influencing landslide frequency or magnitude or timing? And are there particular regions of the world that you think are most at risk? >> Oh yeah, definitely. um uh any geomorphologist I I hold a chair in hydrarology and geomorphology and any geomorphologist will say you know a landscape is in equilibrium with its climate. I mean that's the basic that's the basic start start point um and whatever reason but we have a huge change in our equilibrium and um so the forcings are typically different and I'm going to give you an example that people maybe are not even thinking of that often. So we know that climate change has now also led to for example floods after drought. So you have a drought period and then you have this high intensity rainfalls. That's a combination we typically did not see in many regions. Now we did some really nice work on on clay slopes and this was then in in in China but where you could see that a large dry period will create desiccation cracks will change a completely hydraological there's a feedback of that drought to the hydraological behavior of the subsurface. If you then have a high intensity rainfall, the desiccation cracks, they do not close yet because it's a lot of water in a short time. You get a fast preferential flow. Landslides are always preferential flow hydraologically h towards a potential slip surface. So what you now see is that the risk of a landslide in those areas is much much larger than it used to be because you know it used not to be so dry and then so wet. Now this type of of of changing conditions between the hydraological forcing and um the hydological forcings the climate in general that really really really changes and I would basically say all over the world of course as long as the total amount of rainfall is increasing in places where it's getting drier and drier and almost no water then of course the landside risk reduces but as we So overall the hydraological cycle is augmented is larger nowadays. So overall you could say that landslide risk is increasing. Uh then that's the hazard part. And then we are all living in places where we maybe could better not have lived but for good reasons need to live, need to have a road. And that means that also the risk part is really increasing. So we will see more and more um risk related yeah landslide risks because people live also in mountainous areas and for good reasons. You know it's not that everybody can live in a flat Dutch landscape and by the way then you have floods. [snorts] >> Okay. And you've also done some work with nature-based solutions in terms of water management and disaster risk reduction. And could you briefly describe actually what a nature-based solution is and what are some of the challenges in sort of using nature-based solutions in a scientific sense? >> Yeah. So let me start with so a nature-based solution is is making sure to solve a societal problem. It's a solution because we have defined a problem and the moment you have the problem you try to do it in a way which is more coherent with its natural conditions to say it in my in my own simple words. Um, that means that you would like to have a bit less impact or at least make sure that there are other ecological benefits that society and the world has. And that can be biodiversity, that can be temperature, that can be clean air, that can be our mental well-being for the fact that we like to live near a park and not in a concrete um in a concrete desert. Um so I'm a big advocate of doing it and we know for example that if we plant um good example is Switzerland they have protecting forested which forests which need to be there to basically capture rocks that fall down and protect the roads from the rockfall. So you can make a net you also can have a forest and that's a really sustainable way of having so you cannot touch those forest. Now this is an example in hydrarology we have many ways of storing water in cities sponge city concepts etc. Now the biggest scientific challenge is not this concept. I mean the concept is there but is the fact that in my honest opinion and maybe a bit bluntly said the time scale at which we look at their effect is typically that of one political cycle. So it's a few years and that worries me. It worries me because forest if you have a mono forest after 30 years or after 50 years or after 60 years it will not have the same strength as it has when you planted it. If you have a water storage area a swill in a certain area. It at some point will not have the same hydraological function as it had before. So this long-term effect of nature-based solutions is really worrisome and we do not spend enough time to see what are the long-term consequences and effects and they can also be positive but there are feedbacks known and unknown feedbacks and the feedback can also be for water storage and human health via factor-born diseases. So there are all kind of consequences and feedbacks that go beyond the one solution for one problem. That's the beauty. As a scientist and as somebody who typically says I'm an interdicciplinary hydraologist, I I love it from a scientific point of view, but from a societal point of view, we really should not overpromise and also look at the effects on a scale larger than five years. We do a lot of work on that. We really try to to push that. We also have long-term uh open air lab experiments really looking what happens to a soil after 5 years. How does it change its behavior? How is the co-evolution of the new ecosystem with the new hydraological conditions that we partly made ourselves because we brought the water there. uh this type of of feedbacks and interactions I think it's a gold mine for the new generation to work on scientifically. >> That's fantastic. Um we would also like to talk a little bit about your work as the chair of the helping hydraological decade. So in the hydro talks podcast before we had the chair of the panta uh hydraological decade and from the helping we had coordinators of the co-creation and the science communication working group. So this is a quite nice segue to this particular topic. Um so with your IHS hat on can you tell us a little bit about what are the main objectives of the helping decade? Yeah, thanks. And I I really enjoyed listening to those interviews. And let me first of all say that I am super proud of the fact that I'm allowed to do this. Uh and honestly, the task is relatively simple because of the enormous enthusiasm that is there in the in the bottom up grassroot um initiatives that this really is. um what I see helping focusing on at the moment um maybe go back you know when we had the prediction in ungage basin there was this this correct question like it's nice to do science when all the data are there but please can we do a bit more if not all the data are there and we got the pup decade after that we got the changes decade which really said how can you imagine that there is a kind of stationerity things are changing what we also just discussed on the long-term effect of this naturebased solution. They all fitted nicely if I look back to the to the to the things society was discussing and doing. And currently you see that the young generation, you ladies, but also all your colleagues are really and also me, we want to work a bit more for and with society with solutions. And this is a decade where we still do science, but the focus is really on making sure something is done with it. So I always say it's the solution decades. I also think Ben Howard in the previous um episode said the same. This is about solutions. This is about interacting with local stakeholders. And that is what really is the core. That's the core. We work for solutions with society all over the place. You know we have every hydraologist knows that a certain flood problem, a certain water resources problem has local dimensions and a global driver which is the land use and climate change um factors that we currently see. And so do you have any reflection on how these uh solution focused um objective of this particular scientific decade has connections to the previous pant scientific decade. >> Yeah. So what you really see is that science for science. So the the enthusiasm for science to [snorts] go for kind of physics or chemical automatic level of of of research has shifted a bit to interdisciplinary work and um you know when I was a PhD student I lived for according to the to what Furry said like preferential flow is the rule rather than the exception. That was when I was studying soil hydrarology. What I would say that the helping decade maybe is is interdisciplinary work is the rule rather than the exception and helping is doing that and with all the research we are doing it maybe seems less fundamental but it's not. It's interdisciplinary and the fact that we have this enormous focus on interdicciplinary working in co-creation like you had um in looking at several aspects and learning from other disciplines is really the core achievement that helping currently has and it's also really a core of the competences the new generation of hydraologists but also other scientists need to have. You know you had Stan Gross as one of the first ones. This relation of water quality with human health is something we also work on a lot. I visited him two weeks ago. And this interaction and this interdisciplinarity that's what that's the driver and that's what brings us to solutions. So solutions is not like oh it's a practical thing. No, it is basically a solution monisciplinary is relatively hard to achieve. It's an interdisciplinary part. So drumming on it a little hard maybe. But yeah, that's really what I believe that helping is about. >> Yeah. And are you proud of any particular achievement that you have had since you've become the chair in 2025? definitely not for my own contribution but I'm super proud for the energy uh and the grassroot culture that I has that has been achieved by my uh predecessors and and it's it's amazing to see uh and in that respect the role of a chair and even the role of leaders of sub leaders is modest and I think that uh that I like that also uh I am super proud of the energy we give to subgroups and if you now come up with a new initiative and you write a small proposition for it you can start with it and then you can create in EGUS a session on it you know that that that type of dynamic that's fantastic I'm super proud of being part of that that energy >> that's fantastic now we are almost at the end of the interview we have two last questions which we really love to ask all our guests. So the first one is in your view what are the biggest breakthroughs in hydraological research in last 10 years and what do you think are the biggest trends for the next 10 years? >> Yeah, that's a hard one. I heard that this was a question we always got. I prepared and I still find it difficult. So clearly an important uh breakthrough is the fact that we uh have um uh we have moved hydrarology and water management from a more physics or chemical um uh monodisiplinary work to a multid-disiplinary work. I think that's really a breakthrough. That also means we work with um with feedbacks. We start to really see the feedbacks and quantify those feedbacks. um the breakthrough is typically really the amount of data we have and we are going to to to use. So that's that's really important. That's also where we should work on for the next 10 years definitely. However, we still have steps to take to make our uncertainty quantification um um better. Uh I I really still think that sometimes and also how the scientific publishing community works uh we are sometimes overpromising our outcomes and we have to be really careful there. Uh I really think that that is something to take care of. [sighs] Um but the biggest challenges ahead are the unknown. But now I preach a bit for my own for my own church but the unknown effects of climate adaptation works in the world and also water resources availability in a changing climate and in a changing society. and we don't oversee all the consequences and having that quantitatively correct I think that would be a a major challenge to do with all of us. Maybe that would be one of my one of the 23 questions of the unsolved questions. I would go in that direction. >> So fantastic. So the last one is for our early career listeners could you share the best and worst piece of career advice that you have ever received? Yeah, I also thought of that one that was easier. The worst advice I ever got was that I needed to focus for my scientific career. You know that is that I have always lived to the fact that I just followed where I was enthusiastic for and uh that was sometimes a remote sensing data for landslide early warning system on a larger scale and the other time it was a synthetic DNA particle to trace water. you know, you sometimes have to have the guts to um to follow your heart. And then added to that really really um work with people where you have a click. I think um then Howard said a bit the same in your previous episode, but it's about science is about humans and not about careers. So if you manage to find the persons that you really really are connected to for a scientific question, then go for it. But if you then have another another brilliant idea or intuition or question which goes a little bit off path, please do it and do not listen to the person to say, "Oh, that's bad for your career." Fantastic. Thank you so much, Tom, for your time. >> It was a pleasure and thank you for having me. It was [music] really um stimulating and nice to talk to you.