 Well thank you everyone for joining us today for HydroTerror's latest webinar. Apologies for keeping you waiting, having a few technical issues. David's hiding there in the darkness. Really appreciate having David Hamilton here today for this webinar on the challenges and opportunities of remote sensing for water resource assessment. Moving right along, so there's a picture of David. Let him memorize that one given we can't see him live yet and that's myself, Richard Campbell. So David is our presenter and he'll do his best to answer the questions that have been raised as well. So David is the director of the Australian Rivers Institute and a professor in water science at Griffith University in Australia. He has held positions across engineering and environmental sciences at the University of Western Australia, School of Science at the University of Wokato in New Zealand and at the Australian Rivers Institute. He was the chair in Lake Restoration at the University of Wokato in New Zealand from 2002 to 2017. David was a foundation member with three colleagues of the Global Lake Ecological Observatory Network, GLION, which now has over a thousand members globally. David has researched interests in lake restoration, high frequency environmental monitoring and catchment scale water quality modeling. He has been involved in water policy development at regional and national levels in New Zealand and is a past president of the New Zealand Freshwater Sciences Society. He's editor-in-chief of Inland Waters and holds associate editor positions for several other aquatic journals. So highly qualified speaker we have today and actually having been through the slides it looks like a really interesting presentation with respect to monitoring and how to make the most out of that data. Before we charge into the actual presentation, we love getting your questions and many thanks for the early bird questions. By the way, we've got a lot this week. So David will endeavor to answer those questions at the end. If you want to ask a question during the presentation, please use the Q&A button at the top of your screen. And I will read those questions out to David and we will do our best to field those. Why does Hydrotera run this webinar series? Well, we love to share knowledge and a lot of other people do too. And we provide a forum for them to do that. We like to focus on topics relating to environmental monitoring and management. And this one certainly fits the bill. We're doing our best to facilitate education and we're also wanting to be an industry leader by sharing knowledge of the latest and greatest in the ways to do environmental monitoring. Brief outline of what David's going to present on today. So topic number one is getting ready to commit to high frequency sensing. How representative is a measurement. How does a measurement vary in time and space? How can we use technology to address this problem? And topic two, monitoring. How can we use monitoring technology to better monitor? What do we need to know about the sensors and examples of dissolved oxygen and chlorophyll monitoring? Finally, we'll move to the Q&A section of the webinar. And without further ado, I'd like to thank David for presenting today and we will hand over to David. Thanks very much, Richard. Yeah. Thank you. And apologies. Some technical issues on my end here. Not quite sure why the first time I've sort of presented without video, but we'll see how we go. And thank you for the introduction, Richard, and the opportunity to speak today. I hope that I can give you some insights into some of the real opportunities that we have for remote sensing. I'll be focusing on high frequency sensors in particular. And also some of the lessons learned along the way as much as anything. So I wanted to begin just with the classic sort of context for monitoring. And when we think about doing a lot of environmental monitoring, it's often got a very strong spatial component and also a temporal component. And this is just an example of some actual data of dissolved oxygen. It's not varying greatly through the surface waters, but consider perhaps another system where, you know, quite commonly you may see dissolved oxygen vary by three or four milligrams per liter over the course of a day or so. Obviously at nighttime through respiration only and during the daytime, the dissolved oxygen increases. And as a consequence, we get a peak during the day and then a fall at night when respiration is not compensated for by photosynthesis. But you can imagine these two dots here and they're only separated by three or four days. And so you go out one day, you think you capture the dissolved oxygen about right, maybe go out at 10 o'clock in the morning on the 23rd of September, and then you go out a few days later. And you might go out at 12 o'clock, for example, but you can see what I'm alluding to here that what high frequency sensors allow us to do is be able to have a continuous or more or less continuous record of the strong variations, particularly in constituents like dissolved oxygen and chlorophyll that are really key to and a basis for many of the water policies guidelines that we might see it. Next slide. The sensors that we've got available to us currently can now act at a range of spatial and temporal scales. Some of them go down to molecular and gene scale and you would have heard of the metabolomics or metagenomics that promises so much. And I think it's still got a little bit to deliver to live up to those expectations. But increasingly, there's a variety of different sensors that we're using they span a massive spatial context right from molecules up to global scale. And also, in terms of their frequency of monitoring, we can set them right down to seconds or so. And in some other cases, we may be using them to sort of inform decadal type variations and right up at the far end is our global case where we're looking at satellites that are orbiting the earth. Often on timescales of a few days and able to capture a huge amount of information, especially at relatively high frequencies and these are the opportunities that we've got now for the for the environmental monitoring to to really be able to capture observations from space. Next slide. So it's not necessarily completely easy. You would get the impression and particularly looking at some of the papers that it's pretty much plug and play a smooth pathway from sensors to data to understanding. So the objective being that you set up some sort of a sensor network. You use a wireless transmission. You then might use big data analytics or you might use some sort of a modeling strategy to inform to drive the information stream from the sensors and develop some sort of an understanding and ultimately to feed that into policy and understanding. And it's a big challenge because it crosses many disciplines. So when we're looking at some of these sensor networks, we're really dealing with electronics people we're dealing with people who should have, have often have vast field experience, but bringing together that combination of people is really quite unique the people that go right from the from the sort of grassroots, get the sensors into the ground or into the into the water, right through to the policymakers and that's where the challenge lies is that ability to be able to aggregate and collate the the potential information stream that we've got. Next slide. So this is my sort of word of warning slide if you like. Yes, be prepared for the challenge. And that challenge can be from right from deployment to policy. The case on the left hand side is a sensor that's, and it's not some sort of green monster there. It's actually a sensor that white there is a wiper that is wiping a light sensor, while the rest of the sensor apparatus, or supporting the light is actually covered in the growth of algae. And that isn't even under nutrient enriched conditions. So, you have to be aware that all of these things have a sort of maintenance time maintenance scale that's required to be able to deliver good information. Of course, you have to be prepared for the data deluge. And that data deluge relates to the fact that suddenly if you're even monitoring at 15 minute intervals you've got 96 measurements a day. Those 96 measurements where it might relate to only one parameter or one variable. And so you can imagine that quickly, you're building up stacks and stacks of data, you want to be confident in the QA QC, that is the quality assurance quality control of that data set. And you also want to know that you've got the right calibration procedures in place and I'll talk a little bit more about the graph on the right hand side. And that just gives you some idea of the relationship between the bio volume of algae, and one of the senses that that is used. And it's essentially showing that different species have different relationships for bio volume versus the sensor reading. And then of course, translating it into policy and practice, and what does it all mean. So we've got to have those intermediate people that are taking anything from AI through to models through to understanding to be able to convert it into something practical, and that's really meaningful. And that is goes right back to the senses themselves because good information at high temporal and spatial frequency is what is required to be able to drive good policy. Next slide. So we started 2004, four of us got together and started what we called the Global Lake Ecological Observatory Network. And as you can see from the right hand side, somewhat primitive. But sort of concept was that we could put out multiple senses and be able to monitor remotely at various locations and obviously the focus of those locations was in lakes. And a lot of lessons learned along the way. But we started to realize the potential of this network of bringing people together from around the world to be able to look at temperate systems, some tropical systems, tropical systems, and to be able to deploy a range of different senses and build up an understanding and knowledge of how those senses work. How that we could transmit the information back to us, fortunately through cell phone networks nowadays, and how we could start to use models and other tools to be able to interrogate that information and develop an understanding. Next slide. And it quickly expanded. Gleon has around 1000 members globally, probably a few more by now, with an objective to make high frequency observations of key variables, such as temperature dissolved oxygen, algal concentrations or chlorophyll, and be able to send that data to a web accessible space in near real time. And the beauty of that of course is that you may be able to become responsive to the particular conditions that exist in terms of water quality at that time. And then to provide opportunities to better understand the late dynamics, obviously through models as one of the things that I've mentioned, and just shown here are a variety of different platforms. And that sort of speaks of some of the uniqueness of Gleon, in a way that it's allowed a grassroots structure to take place and build up a sort of core of knowledge of what's working, what senses do we deploy, and how can we work to do it better. Next slide. And some of our colleagues have taken it to extreme examples. The Lake Canary to see if Galilee case actually put fridges out on the lake and have been attempting to do chemical analysis alongside some of the standard sensors methodologies. And it's probably just working just worth mentioning actually that the chemical constituents still represent probably the greatest challenge. For example, yes, we can get down to somewhat low levels of nutrients, not quite as where we'd like to be in terms of environmental monitoring. And of course we're only dealing with the dissolved nutrients, and in many cases it's the total nutrients, the total phosphorus and total nitrogen that are actually the constituents that are used in policy examples. And then you can see the Lake Biwa Japan case. Lake Biwa, a major water supply in Japan, and obviously they're really keen to do some good monitoring on that lake. What sort of investment was involved with that, David? Sorry, what sort of investment was involved in that? That's sort of an interesting question, Richard, because once you get the water supply treatment people involved, the investment tends to be a lot greater. And it's really this interface between what's scientific experimental work and what is actually the routine monitoring and where you get a major investment like that. It often brings the two together. And so we're probably talking, I'm guessing, $2 or $3 million for a setup, something like that. The other cases that you may have seen, the simpler cases going out on lakes, maybe of the order of a couple of hundred k, somewhere in that order, to get a multi-sensor array that potentially profiles the water column. Okay. Next slide. My colleague, Chris McBride, was really in Gleon from the outset. He's developed a huge knowledge. He's been deploying lots of different systems in New Zealand, and you'll see the Rosen McBride reference come up time and time again here. And he's really just showing perhaps three or four different setups of what you might put out. One of those would be just a simple point mooring, a single array of sensors not moving. And so just at one fixed point. And so often we've gone out and we've monitored at the surface waters of a lake and somewhere deeper in the near bottom waters. So the case might be where you start to introduce profilers. And the advantage of those profilers is that you don't need a lot of duplication of sensors, such as what you're seeing in the middle there. So the vertical spar buoy with two point mooring, multi probe, dual bottom and surface referenced. And the case in point and you can see there what is essentially being done is to represent fixed sensors for temperature, for example, that would be to give you effectively the whole water column as being captured for measurements. And in those cases maybe where you want very, very fine scale measurements you want to fixed location as opposed to the vertical profiler. But the vertical profiling has really been quite a revelation. In fact, you know, in the 15 or so years I've been with Leon vertical profile is a really commonplace now they're pretty reliable. And they're really saving on the number of instruments that you have to deploy through the water column. Next slide. I wanted to show you a couple of examples where our understanding is sort of really evolved because of some of the Gleon setups. And this particular one is up in the mountains of Taiwan, Yang Yang Lake. And it's showing water temperature, which the multiple colored lines that you can see starting on the 22nd of August. And interestingly, you can see that it was quite stratified varying from about 14 through to 21 22 degrees Celsius at the surface. The blue that you see at the bottom of the axis there represents the rainfall. And you can see what's happened here is that there's been a huge rainfall. It was actually a typhoon. It stranded all of the researchers that were actually involved with this this work. They were sort of stuck where they were for two or three days. But that precipitation event that rainfall event effectively displaced all of the water in the lake. And then the stratification set up again with the temperature varying obviously from about again about 14 degrees through the 18 degrees. But you could imagine if you just missed that two day event. You'd have no idea what the consequences would be for that lake. And a number of papers were written about this where the biogeochemistry of the lake the chlorophyll and nutrients actually changed completely because the whole of the water column was displaced. Like shallow lakes. I remember starting my PhD. And I was always told, don't worry about shallow lakes. They're always mixed. You don't need to look too carefully about that that whatever you put in, it's vertically mixed quickly because the wind will drive that mixing. And in fact, the more we look the more stratified stratification that we see in these shallow lakes and often they're stratified for 3040 sometimes even 50% of the time, even lakes that are as shallow as about two meters can be strongly stratified for substantial periods of the day. And if you just click again Richard, I hope this will pull up. Yeah, and again. The difference is actually enormous because here are two photos that were taken one of them when the lake was mixed the top one. And the other one when the lake was stratified with the warm waters of the surface being about 27 degrees Celsius. And what that allowed the cyanobacteria the green, the blue, this blue green algae in this lake to do was to be able to float up to the surface, instead of being mixed through the water column during the night. And so this is a huge driver of the dynamics of many of the shallow lakes and particularly those where there are algal blooms that occur, and essentially those those algae float. And the temperature is is a major driver through its influence on the stratification dynamics of these legs. Next slide. So I'll switch attention a little bit now to dissolved oxygen. And a classic quote here from Evelyn Hutchinson who was probably one of the original pioneers of lakes I think he wrote four volumes of a book. On lake processes and late dynamics, and he said, how a limnologist someone that studies fresh water can probably learn more about the nature of a lake from a series of oxygen determinations than from any other kind of chemical data. Next slide. Now, when we talk about dissolved oxygen. We're really talking about a massive transition of sensors here. The old days were to go out and do Winkler titrations to measure dissolved oxygen. So literally titrating to be able to understand those concentrations of dissolved oxygen. And gradually the sensors have improved they've got better and better. They've become optical sensors in many cases. They may have wipers across the optical part of the sensor to be able to stop biofouling. Or in this particular case on the right hand side a sacrificial copper anode that would also is also designed through the copper to be able to act as an alga side and prevent that biofouling. So it's actually been quite a quite an enormous transition to go from Winkler titrations. Where you just do a few each day to potentially taking hundreds, even thousands of measurements a day using an oxygen sensor. Next slide. And this is where it gets pretty difficult and I'm not going to try and borrow in too much today to talking about different sensors. But you've got a large range of large choice to be able to make a decision about which one. Some of these will be able to measure extremely high high frequency. And so you might use those in a profiling mode. Some of them will be extremely stable in the long term. And for those ones, of course, what you want to do is get them out there for over the long term with the wiper that prevents biofouling and be able to hopefully avoid some of the problems of drift. And potentially photo bleaching of optical sensors, for example, that that may occur. All of these senses will drift, but to varying extents, they'll have different levels of sensitivity to temperature. And they'll have different, there'll be different techniques that will be used to derive the data. But I guess one of the important things is that it's not dissolved oxygen, it's a voltage and it's a voltage that's converted to dissolved oxygen. And so we've always got to be aware that we need to play a part in doing that quality assurance for ourselves that dissolved oxygen is really going to be dissolved oxygen measurements with these senses are accurate. Next slide. When you manage to put high frequency data together over long periods of time. You get some remarkable insights into systems. And this particular case is Lake Rotorua in New Zealand with data collected at 15 minute intervals pretty well continuously from mid 2007 and in this case, and it does go on. But I'm only showing here to mid 2015. The blue being at the surface where you see a little bit of variation but most of the time the oxygens around 100% of saturation give or take say 20%. When you see these master swings in the dissolved oxygen at the bottom of the lake around about 20 meters. And so during the winter when the lake is mixing the oxygen in the bottom waters approximates that of the surface waters. But during the summer, you see these huge swings right down to 0% of saturation in the bottom waters. So you can imagine that any sort of benthic biota that relies on oxygen, of course, in those bottom waters would not be able to survive with these swings in dissolved oxygen. So let's just take a look in a little bit more detail in the next slide with. And sorry, I'm going to that red line represented just the detail of the of the following slide. So it's one of the events that we're now looking at. Left hand side, you see water temperature top and bottom. Sorry, you see water temperature. And on the right hand side, you see the dissolved oxygen. The blue is water temperature and the orange line is the bottom dissolved oxygen. Let me get this right. Sorry, I'll just recheck that. Yeah, the blue is the water temperature. And the orange is the dissolved oxygen. And what we're looking at is surface on the left and bottom that 20 meter sample on the right. So let's have a look at that that right hand side in particular and you see an almost linear decline in dissolved oxygen when stratification sets up. So as soon as you get even a small gradient in temperature between the top and the bottom waters. That means that the mixing is no longer occurring between the top and the bottom waters. We're looking at this gradient of dissolved oxygen, depletion of dissolved oxygen, and in this case down getting down to about three milligrams per liter in those bottom waters before a mixing event occurs. So for example, assuming that that was perhaps a wind mixing event that restored the dissolved oxygen to the bottom waters by mixing the bottom and the surface was next slide. And you can actually put these together to look at sequences over over years. So each individual event has been highlighted here as different colors. And we look at we're looking now at the number of days until that deoxygenation occurs. So the vertical here is the dissolved oxygen as a percent saturation, and the timescale shows how quickly that oxygen is depleted. Obviously, if the if there's a lot of organic matter, there's a lot of decaying algal cells and collapse of an algal blue, then the rate of decay of dissolved oxygen will be very high. By contrast, if it extends out much longer, then obviously that organic matter is is much reduced. And so what you're seeing there is actually a significant improvement between 2008 nine and right through to 2014. That was reflected in fewer algal blooms than in 2008 sorry 2014 compared with 2008. Next slide. So, up to now I've been talking mostly about the temporal context of monitoring. Let's have a look a little bit at the spatial context and you can see here that I've got an algal bloom. And many management authorities are going out and they're trying to figure out. Well, if I sample at the edge of the lake, I'm probably getting a good assessment of risks to human health because some of the cyanobacteria that creates blooms can be toxic. If I go into the middle of the lake, is it representative. And Richard, if you could just click on the on the right hand side to show simulation. Yeah, if you could just click on that one. So this is actually a model simulation because it's obviously almost impossible to measure at the time scales that we're showing. But it's a very accurate three dimensional model. And it's been coupled with an assessment of the number of cells per mil of cyanobacteria. Imagine if you were someone monitoring and trying to monitor this lake. It's a lake in Paraguay, actually. And where on earth do you start. It's got huge variability in time and space. And this has been one of the confounding factors in much of the cyanobacteria work that's taken place around the world, much of the blue-green algal monitoring that's taken place around the world. So if you're scratching the head, they go, for example, to the site shown in the photo at the left hand side and say, yeah, I can see there's an algal bloom there. And often if the wind changes or something else, some of the mixing dynamics change, there's no bloom. So it's really difficult to and cyanobacteria actually represent one of the most challenging as the monitoring that's done. How can high frequency sensors help? Next slide. Well, if we start to think about how we can combine these sensors with other measurements, so we may have something on the ground, or in the water rather, and that may be a chlorophyll or a phycocyanin sensor. We may have a satellite that gives us a snapshot every 12 or 14 days, perhaps. We may also have some aerial, other aerial sensors, some of them perhaps fixed that are looking, essentially looking at the water and looking at the color of the water to be able to interpret where those blooms are occurring. So you're starting to see here that we may be able to use multiple tools to be able to better inform ourselves about the distribution and the dynamics of these blooms. One of the things that's clear is that some of the manufacturers are not fully engaged with some of the scientific requirements. And so when I look, for example, at some specifications pulled up here, you know, chlorophyll down to .02 micrograms per liter. I mean, that's almost laughable. It's just not realistic. It's only bacteria at 150 cells per mil, but we don't even report in cells per mil when we go out and measure. So we have to be really aware that there's a difference here. These are proxy proxies that we're using to inform ourselves of concentrations that we're interested in. And so automatically converting them through as the manufacturer has done here to to sort of concentrations is really quite deceiving, certainly in terms of reporting in cells per mil. It's just not realistic. Next slide. When I say it's really difficult and challenging, but it does offer a huge amount of hope for us. And here what we're looking at for chlorophyll is that chlorophyll has a certain absorption spectrum. And it also has a fluorescence or emission spectrum. Now, when we shine a bright light on phytoplankton on algal cells, then about 1% of that bright light is actually used to make the cells fluoresce. And similarly phycosine, which is contained in cyanobacteria specifically, as opposed to all phytoplankton. But in cyanobacteria, the blue-green algae, we can use phycosine as a proxy for those, for cyanobacteria. And the photo on the right just shows a sort of breakdown of cells in the scum in an algal scum. And you can see the blue, which represents the phyco, the phycosine of the cyanobacteria and the green representing the chlorophyll at the surface there. Next slide. So specific wavelengths for excitation and specific wavelengths for detection, the green, the chlorophyll, the blue, the phycosine. And so here you're typically looking at an excitation wavelength of about 470 nanometers of blue light and a detection wavelength of around about 670 or so nanometers for chlorophyll. Phycosine, and you're looking at exciting it at about 600 to 610 nanometers and then a detection again, pretty close to the chlorophyll detection. So you can see some of the challenges. And I think the next slide should show you what I mean by those challenges. Sorry, not quite. I've moved straight on to the satellites. Sorry about that. No, no, that's, that's fine. So, so that was a little bit about the sensors, the satellites are not too dissimilar in some ways. So these satellites are now being sort of oriented towards picking up a greater resolution of the light spectrum. And so you're now starting to see satellite specific satellites like Sentinel to that is designed to specifically pick up phycosine. So you're starting to ask, how can we use these sensors in combination with those that are sitting in the water body in a lake. For example, every time the satellite goes over. So we've got measurements every time the satellite goes over to be able to really get that snapshot and extrapolate it over the, for example, a whole of a lake or over multiple lakes. Next slide. But we do need to again caution on various issues that can can occur. One of those is obviously biofouling wipers, various biocidal compounds have been used. The wipers as far as I've seen have been very successful in increasing the stability of many of the sensors. But you do need to realize that if you're not using them, then it's very easy for these biofouling agents, particularly algae to grow on the sensors and impact on the readings. Next slide. Typically what we do, sorry. Typically what we do is we'd, we'd try and relate chlorophyll fluorescence measured from a sensor to an extracted value. And this would be from a grab sample that would be taken back to the lab and analyzed for its pigment content in terms of its level of chlorophyll. So we're still sort of stuck in a way to rely on calibration and the calibrated values that we would typically expect in the field, maybe of the order of 70% of the variation explained perhaps. There's always going to be that inherent variability when we're dealing with field samples. Next slide. But even in the lab, we can see that here we've, we've looked at the, the bio volume, which is basically a measure of biomass of cyanobacteria on the horizontal axis. And we've compared that with the readings from a sensor measured as phycocyanin. So different elbow groups actually have actually vary quite hugely in terms of that relationship. So it's very important that we do maintain this focus on our quality assurance in terms of calibrating, validating our measurements at each step where possible, so that we build up increasing levels of confidence in these senses. Next slide. There's just one. It's a, it was a pretty good relationship that you saw. The other, the other one is around quenching. So many of these senses do have a quenching response. In other words, as you increase concentrations, for example, of chlorophyll, then the, you eventually reach a point where the sensor responsiveness decreases. And so this quenching response is obviously very important. And you have to be aware that you want to be using that sensor in the linear part of its range, or else be able to readjust the range that the sensor operates at. Next slide. Quenching also relates to what we call non-photochemical quenching, which is very common in these types of senses. And particularly in profiles, it can, it can be a real problem because essentially some of the algae or phytoplankton at the water surface are exposed to very, very bright light. And this particular case was when you can see the temperature one morning went from 10 to 11 degrees or 11.5 degrees Celsius. That was indicative of a very thin surface layer. And I'm sure you've been swimming and you've noticed this, this very thin layer that sometimes exists in a, on a calm lake or calm ocean. And that means that the mixing is only for a few centimetres below the surface. So those phytoplankton were exposed to very, very bright light. And the consequence of that in terms of the chlorophyll fluorescence was that the cells completely shut down their response. Essentially, they were protecting themselves from that bright light. And the result is that there's no linearity between chlorophyll and the biomass that we want to measure. So we also have to be aware and cognizant of that, that issue. Next slide. And you can see it really clearly here, beautifully demonstrated here. What we'd expect to see is a continuous or semi continuous green line, maybe with a bit of a trend. We just need to be careful with the solar radiation. You can see there, because that bright light basically drops the fluorescent signal quite strongly. Now, you would say, okay, let's use our nighttime measurements. And that, that's, that's certainly a good, good compromise to, to the sort of case. But if you're going out and you're using these types of probes during the day, you need to be very, very aware of that non-photochemical quenching that will basically completely disrupt the, some of the phytoplankton at the surface in terms of their fluorescence response. Next slide. And you can actually see this in a beautiful profile across Lake Taupo. One of the, or the, the largest lake in New Zealand, 160 meters maximum depth. And the gray here represents the bottom of the lake. And the colors represent the levels of, of chlorophyll fluorescence, which approximates to the chlorophyll concentration. So it's a beautiful clear lake. And what you see is this amazing deep chlorophyll maximum sitting at about nearly 50 meters deep. So that's where all the algal activity is. And that's what these sensors can do. They can pick up these amazing insights into, you know, whereabouts is all the activity happening. But the other thing to point out is actually the non-photochemical quenching that's occurring as we went from 6 a.m. around sunrise to 1 p.m. Then you can see that if you look across the very surface of the lake here, you'll see where you've got an increasingly blue sort of low chlorophyll signal right at the top, at the zero meter case as you look horizontally. And so that's indicative, in fact, of increasing quenching of the signal under bright light. Can you acquire correction factor to that, David? Yeah, we've been working on that actually, Richard. And the conventional models don't actually do very well because they don't keep a memory of what happens to those cells. So in other words, if you work a typical, what's called a Eulerian type model, it basically throws everything in together and forgets about any memory of the light or the nutrients that the cells had experienced over each time step. So the way in which we've been doing it is using individual based models to try and get at this problem. But it's really thrown up for me that our models are actually not good enough to pick up this sort of thing at the moment. Next slide. Yeah. So I'm not sure how we're going time wise, Richard. It's good. We've got plenty of questions. So if you might want to just write the key points, we'll go to the early bird questions. Yep. So my message in a way is know your instruments, know your network that you're trying to set up know the systems that you're trying to work to measure the data. You will get malfunctions, you'll get connection issues, you'll get outliers in the data. It's going to be a challenge. And if you think really that these sensors are an easy way out. Well, maybe they are for some of the temperature sensors. But in a lot of cases, they require quite a lot of work, but it's worth it, in my opinion. So you've got to know about temperature compensations quenching lifespan, particularly for some of the complex measurements like chlorophyll and phycocyanin. And you generate a lot of information, lots of individual observations each year. So it's going to be a challenge to manage that. Take away number two, data collection and quality assurance. Calibrate the sensors according to the manufacturer's recommendations, for example, a two point or more calibration measurements of temperature and atmospheric pressure, for example, may affect calibrations. You need to keep on top of the sense of performance and the sense of outputs. And that means that there'll be an age to be a lifespan of the sensors, they will drift. So there's lots of lots of different things that come into play here. And I'm going to use the people around you, I guess, because it is a very multidisciplinary field. And it's also requires a lot of networking, a lot of communication to be able to optimize the opportunity that we've got. So that shouldn't be sensory that should be sensor networks and data harmonization offer the potential for unprecedented insights into water resources with capability to inform management and policy. And I hope I think I've got one more slide. Yeah. So it would be great to piece it all together wouldn't it. And in some cases we're right at the tip of being able to do that. In other words, we've got sensors in the water. We've got our satellites that are able to open up new opportunities for spatial information every couple of weeks, which can be calibrated effectively by sensors in the water. So we've got new equipment coming out all the time that allows us either to take that equipment through a lake and get a good spatial representation or to be able to leave it in the lake. And also complement these measurements and if you just click on the bottom right there, Richard, hopefully that'll throw up a movie. You know you can see the sort of thing that we're looking at here where this is water temperature. We're going through a geothermal area in this particular lake out into the main body of the lake. We've got lighter or we've got various sensors that are giving us the bottom of the lake. We've got an instrument that goes across the lake taking readings and then we've got a boy in the middle of the lake also taking temperature readings. It's a new opportunity that we've got here and I think it's pretty exciting. It's interesting looking at that. We've been involved a bit with extrapolating between measurement points and I know that Syro were working closely on extrapolating soil moisture across paddocks in real time. Looking at that in DVI imagery and matching it up together. It sort of does come down a bit to how many of those points you have as a point of truth, doesn't it? To correlate against it. So how many sensors are really required out there versus what you're relying on from your spatial imagery. Exactly. That's right. And so if you are putting in the investment to have reliable, if I call them in situ sensors, then it just opens up this world of opportunity for the satellite interpretation. And with AquaWatch coming up with CSIRO, their new AquaWatch program. And with the level of resolution that's possible now for particularly for agricultural soils and looking at soil carbon and methane emissions and so on. It's really an incredible opportunity, but it's extremely data demanding. And it does require our ability to be well organized for databases and for our interpretation skills. One thing we've noticed just before we jump to the questions is a lot of people are putting networks in that that sort of structure around actually running and monitoring sites and the rigor that comes with those standards like the Bureau of Meteorology have for how to run these sites just isn't being applied very well out there in a lot of cases. So I suppose there's a couple of things I was looking at in your presentation. I was thinking, well, we've seen similar things, but some of those checks on the environment that the sensors in against the sensitivities of the sensor itself are really important because you can be chasing a lot of furfies that have been affecting, no sensor readings are affected by those secondary factors and you think they're real, but they're not real. So I think there's a need for a fair bit of work in that space. You mentioned the use of proxies and sort of questioning almost questioning the ethics of the sensor manufacturers about the units of measurement that they're reporting out of their sensors. I think there's a lot of work required there, you know, it'd be great to have something like the standard around that. It's a bit of a hole in the industry. And as you say, there's so many new sensors coming out all the time and different approaches. I'm not sure if in your organization, David, you're focusing on that at all, but I think the industry needs something like that. I agree fully, Richard, that sensor standards should be a really big focus. But there is a sort of danger here to you that, you know, someone picks up a picks up 100,000 of sensors puts it puts them in a stream or in a lake or something like that and then becomes disillusioned with what they've got and perhaps hasn't fully understood what was required in order to be able to get the data to a trustworthy stage. And that's my point that that we're at the sort of threshold where those that are doing it and going down that pathway and becoming sort of expert at it are really extremely knowledgeable and such valuable resources for us. And then there's others that where we can see that disillusionment with what they've got, you know, and perhaps hadn't quite appreciated, you know, just what was required to be able to set up and manage the system that they might have. And unfortunately, there's a lot of old sense and junk in lakes around the world. And, you know, that's that's a sad case, unfortunately. There's a bit of a lack of training opportunities for people around this. Like there's absolutely a dearth of good training on say things like true data provenance out of senses, you know, what what does a measurement really mean. Running some training sessions on that. There's a certain complexity that I think some applied training courses need to be developed as we sort of there is an explosion and senses you know how to tell us this is more senses but how you integrate it all together and that side of things is pretty challenging. And I'll put that slide early on about how things link together. That's the journey. I'd return is on and it's, it's a massive challenge. But we better move on. You've got some questions coming in and we've got a lot of early bird questions. Well done to everyone who's sent in early bird questions. We do have another webinar starting in a couple of weeks by Nigel Murphy on biochar. So keep an eye out for that. Questions. So I'll read these out David and then if you can have a go at answering them. So question number one, I'd like to know more about aerial satellite sending and its current application for ecosystem health and impact assessment. I think you've covered some of that. Sure. Yeah, Richard, the thing I didn't cover was was perhaps the opportunities for drones and putting, you know, doing spectroscopic measurements on drones. I mean, that is another possibility. You know, it's being it's, it's sort of a pretty sexy looking opportunity. When you put a drone up and you have a spectroscope or something that's able to detect color on it. But keep in mind also that there are fixed sensors for those as well. So, you know, they can be all be all be used in combination for sure. Okay, question number two. Why aren't there more river and flood remote monitoring devices deployed. Early warning of sudden river rises can be life and death. Yeah, well, very true. I guess, you know, when you think about the IPCC and you think about the climate change models and you think about the quality of information that's now going forward in meteorology. Well ahead of a section. They're able to be using really good information, reliable information and admitted the easier information to measure than what we're doing and able to communicate that in terms of well the storm coming or those types of things. We're sort of moving towards that for floods. We're not there yet and we just need a bigger investment because if you look at the climate people. You know, I'm sure it's at least 95 to five that we in terms of investment going into climate monitoring and not so much into water monitoring. You know, I'm told 90% of the climate change impacts occur through the hydrological system itself. But only 3% of the funding goes that way. So, you know, that might be part of the answer with respect to this question. Do you think we get more value out of having more fixed location continuous monitoring sensors or more the style of like those run of river surveys that do on the Murray River where you run it and you get that sort of spatial context. Well, again, I'd say that both of the sort of tools that we need one of them gives you a sort of Lagrangian view of the river if you like and the other one will give you here's what happens in the course of a day or a week or and so on. So I think they need to be used. All of these things need to be used by people that are competent in them in a complimentary fashion wherever possible would be my answer to that. It is amazing what they've done. I don't know if you've seen that nanotem survey stuff that they did on the Murray River about 15 years ago you could say the groundwater surface water interaction. Pretty clearly. It's impressive. I would better move on to the next question. Number three. Do you have an insight on the ability and accuracy of remote sensing to detect water bodies refuge pools beneath the canopy cover. I have heard and and John Marshall is the person that put the paper in front of me from Queensland Government. That there's now a capability to look almost look through vegetation to be able to get digital elevations for riparian areas. So, all the time these these methodologies are improving. Would it look at refuge pools beneath canopy cover. Not sure, but certainly the ability to look through canopy cover I mean that's to me is pretty amazing. And we should differentiate here a couple of different types of satellites one of them which is basically simply measuring the color that comes back to the satellite, but others are able to look at sending out a signal and the retrieval of that signal to be able to do some things that I would never have thought would have been possible. And you've got others such as grace, which is a sort of two satellite system. And that's able to measure large scale groundwater changes. So, yeah, just amazing development of technology and certainly the if the aqua watch satellite or satellites get up. That will be a revolution for inland waters, particularly around phytoplankton and so in a bacteria. What is the status there, happy with the technology and now they're looking for money to get the satellite network up. I'm probably not quite the right person to answer that or though I wasn't a workshop two weeks ago with with them and you know it's, I think they're looking for confirmation of the funding for that mission. And if that goes ahead, then then there will be a major push now to implement some of the the satellites that they've been quietly developing over the last couple of years. So just to clarify something you said earlier, so you're saying that they can use satellite data for detecting groundwater movements or large scale groundwater essentially it measures in effect the mass of the earth. And so the major variation on that at the scale it's measured by those satellites is actually the amount of groundwater. So in effect it's it's an indirect measure to look at groundwater levels. That's something I thought would never happen. Well, it's amazing. See some huge opportunities here. I presume they're a little way off getting the accuracy of the pressure. Well, that's the idea is certainly that you'd be using a whole range of sites, large groundwater systems and be monitoring through time and be able to cross validate with what was actually being done with peaceometers. Yeah, that sounds amazing. Next question. Is the increasing availability of remote comms combined with new sensors going to impact the ability to ground truth remote sense data? The answer is yes. Yeah, in simple words yes. Maybe a slightly more. I think the beauty of these senses is that, you know, sites that are very hard to get to that may not be monitored very often if you if you've got that ability to capture continuous data. It's such a leap leap ahead. Are there any efforts going on to better rate other sensor networks that are already out there to sort of combine that to help to accelerate ground truth? I think we're putting up more and more satellites is the ground truth in keeping pace with it. And the answer probably be almost be no, I think. You know, that we, we certainly need to be able to ground do the do the ground truth thing and one of the aspects of satellite remote sensing has been that the barrier to entry now is much lower. You've got Google you've got various tools that are really at a fingertips which are or that basically take away a lot of the computation that previously was done by specialists in the field. That is obviously a good thing but it's also just something to be a little bit careful of and so some of the some of the aspects to that are not should should be understood because they're not necessarily applicable in all instances. I'm just trying to think of the example of an example of that but but checking carefully for example for cloud cover checking for atmospheric correction those types of things can be sometimes quite specific and a good justification that you should be looking looking at the data pretty carefully as you receive it. Yeah. Next question this is someone who could do with some monitoring in real time I think how are our fish populations going and what fish passage efforts are underway. I'm sure I'm the right person to answer that but I think that all of us would agree on the first point, not good. What about the fish passage efforts. I would say that we've we've now got acoustic sensors, which are very good at being able to monitor under very turbid conditions. That's a real bonus for fish populations that for being able to detect fish populations and they're getting very good at it because in some cases they're not just saying well did a fish go past the sensor. They're actually able to record well what was the size of the fish. And in some cases even now they're using AI to detect fish behavior so you know it's it's progressing but you know the details of it. It's not my area. Sounds like you know a bit about it though. Number six, just answer that one. Yes, there will be a recording of these webinars. If you go to our website on the right hand side under about hydratera. If you wait until Monday afternoon, there'll be a recording up there of this session. Number seven, plastic influence on different types of ocean systems, ecology, climate change, I think maybe focus in on the plastic. Yeah, so we, I am one of the authors and I say one, one of many. The paper coming out in nature very shortly, probably in the next two or three weeks. And it's quite a nice example of where the people from Leon got together and monitored or took samples from lakes around the world. The outcome of it is has been that plastics are so ubiquitous. It doesn't matter whether you go to, you know, the Alps of Europe or even some more remote places like Tahoe where there's 45 billion plastic micro and MISA particles in the lake. You know, plastics are enormous. And I don't know whether you could possibly monitor them remotely, but certainly that was a nice example where all of the methodology was standardized. It was a very well directed sort of project and the leader of it was able to bring it all together and synthesize a sort of overview of what's the state of our lakes in terms of plastics and it's, it's pretty horrific, you know, anything from the 45 billion particles in Tahoe that I talked about, but all of them right across the river. Have you seen evidence that that's really affecting the whole food chain of it now? Yeah, that's a good question. And that wasn't what we were tasked with. But I know that one of my colleagues Fred Loosh and Shima Jigarami are working on the sort of ecotoxicology, ecotoxological effects of plastics, but again just right at the edge of my area. Yeah, we had an interesting presentation from the Centre for Anthropogenic Pollution Investigation and Management, CAPM, by Susie Reichman and Plastics in Soils. She was talking about the impact of the micro plastics on the guts of the nematodes and how it just basically cuts them up. So it's a bit more like an asbestos sort of physical problem, not so much the chemicals in the plastic, but it's just killing off these various nematodes and things. Have you seen any of that or that's not really part of what you've focused on? Well, I think the next step after sort of quantifying what's in the water column would be to start having a look at the content of, you know, get some idea of content across different biota. And I'm thinking there specifically in terms of the nano and micro plastics of filter feeders. I mean, those filter feeders are going to be taking in huge amounts of those nano and micro plastics in some of those lakes such as Tahoe and in the ocean obviously as well. So, you know, what impact? I don't know at this stage. Yeah. Would you say that's the scariest thing that's going on at the moment in terms of ecosystems? I find it a bit alarming that we're still increasing, you know, some of the rate of synthesis of chemicals and the amount that they're used in the environment is quite worrying and obviously we've got a PFAS issue at the moment. And, you know, we can sort of monitor for potentially acute effects, but what are the long term effects and you might see that a little bit in terms of algal toxins, you know, that we certainly test and look for the acute effects of algal toxins. But, you know, the neuro genitive diseases seem to have been linked in some way to algal toxins are pretty concerning. And of course, there are other things such as liver cancers, kidney cancers associated with that have been associated with populations with have been very high levels of algal toxins. So, those those aspects are much, much more difficult and only arise through time as opposed to the typical, you know, toxicological tests that we might might do over short short time periods. Yeah. Okay, next question. Please discuss more on applying Sentinel to monitor surface water quality. Right. I'm just trying to remember Sentinel to has a band specifically tuned for cyanobacteria, which is really fantastic. The only trouble is that from my recollection of Sentinel to I think the grid size or the pixel size and it is about 300 meters. So it's very coarse. It's great if you're looking at Lake Erie or some of the some of the great lakes and large lakes, but it's probably not quite as well suited to some of our smaller systems. And that's sort of where the CSIRO aqua watch mission with its own satellite tuned specifically for inland waters and picking up multiple different wavelengths. But we would I think be a real bonus to for particularly for Australia to have it would be just be a huge transition in their ability to monitor water quality. We're very good at water quantity. Generally, we're not so good at water quality. And so these that's where the satellite technology is really going to play a role. Yeah. Hello, time David, if you got your mind hanging on just to answer the last three questions. Yep. Thank you very much. So Leon van der Linden says great summary of the challenges David so that was an easy one. And the other one managing expectations and communicating the results is a real challenge. I couldn't agree more. Absolutely spot on Leon. Yeah, it's a big challenge and those expectations are also very dependent on the capability of the groups that's working with these these senses. Do you utilize pH changes to interpret changes on blue green algae growth? Yeah, pH is an interesting one. It does tend to be they do tend to be a little bit more temperamental than many of the other senses. And they also prone to drift. Could you use them for blue green algae growth? Yes, but there are other factors that come into play there and one of them is alkalinity of the water. So could you, but you can kind of adapt an approach that might give you some insights into blue green algae by on a lake specific or water specific water body specific case. Yes, you can. Again, it's another complimentary measure to reinforce the what the other senses that you might measure that are more direct. It's a new pH sensor technologies that have come out fairly recently that require the same rigor around calibration. Are you familiar with those ones? No, but I have seen a big, really big change. And, you know, we, I mean, when we first started out, we really didn't bother putting pH sensors on the probes just because they quickly drifted in. But now they're certainly going going on to the probes. So they're probably the exact case you're talking about, Richard. Well, David, we've got through it all. Thank you very much for that presentation. I thought it was fantastic. It's really interesting. And thank you to our viewers who've held on for a prolonged webinar. But really great to get so many good questions and great to have a presentation from a true specialist. So many thanks, David. Thanks, Richard. I'll see that I get the camera up and coming next. I think that was really good. So thanks very much. Thank you. Bye.