Eyes on Earth Episode 57 – Landsat and the Great Lakes

JANE LAWSON:

Hello everyone and welcome to another episode of Eyes on Earth. Our podcast focuses on our ever-changing planet and on the people at EROS and across the globe who use remote sensing to monitor the health of Earth. My name is Jane Lawson and I'm the host of today's episode, where we're talking about satellites data usefulness in monitoring and restoration on the Great Lakes. The Great Lakes represent roughly 20% of the surface fresh water in the world and 90% in the United States. More than 3,500 plant and animal species inhabit the Great Lake Basin's open water tributaries, wetlands, forests, and dunes. More than 30 million people in the U.S. and Canada call the basin home. The health of the region faces threats that range from climate change and invasive species to pollution and development. Identifying and addressing those threats can be a daunting task. Our guests today are here to talk about how Landsat and other remote sensing tools has helped with some of these challenges. Bob Shuchman is co-director of Michigan Tech Research Institute or MTRI. He earned his PhD in Oceanic Sciences and Natural Resources. And is also a research professor. Mike Sayers is Research Scientist with a PhD in Optical Oceanography. Bob, welcome to Eyes on Earth.

BOB SHUCHMAN:

Hi. Happy to be here.

LAWSON:

Mike, welcome to Eyes on Earth.

MIKE SAYERS:

Thanks Jane.

LAWSON:

First, let's talk about some of the challenges to the Great Lakes basin that you have worked with. Bob, do you want to start.

SHUCHMAN:

Yes, I would be happy to. Really what Mike and I at MTRI have focused on the last decade is water quality via remote sensing in the Laurentian Great Lakes. And there are three forcing functions that have greatly altered water quality over the last couple of decades. The big one, climate change, anthropogenic forcing, and invasive species, particularly the mussels that arrived in the 1990s and greatly altered the ecosystem. Now if we drill down a little more on what does water quality entail for us, it entails chlorophyl, dissolved organic carbons, suspended sentiment, water clarity, and light levels. But the other big thing that we look at are harmful algal blooms, particularly in the western basin of Lake Erie. So those are the issues that keep Mike and I up at night. Mike, would you like to add to that?

SAYERS:

It's been interesting for me coming into MTRI as a young scientist that was frequently recreating and doing things in the Great Lakes and seeing how these problems actually manifest in person, and then being able to use remote sensing and science to better quantify those.

LAWSON:

Why is remote sensing so useful for some of those research and monitoring efforts?

SHUCHMAN:

The short answer, Jane, is think about the remote sensing satellite time series that we have. It is a time machine. It allows us to go back a couple of decades now and look at some of the water quality changes. A lot of times we don't have the resources or the intuition to go out in a research vessel and measure invasive species five years before we really realized that there was a problem. So by using Landsat time series that goes back to the early 1970s, that allows us to fill in measurements vis-a-vie remote sensing that frankly we didn't have enough sense to measure during the period that the problem was occurring. Time series chlorophyl for example, benthic Cladophora water clarity all help us better understand what is the health of the Great Lakes. Where have we been? Where are we now? Where are we going?

SAYERS:

I'll just also note that remote sensing is an extremely useful tool in an environment with the scale of the Great Lakes. So the Great Lakes are an extremely large area. It's really impossible to think about sampling these things effectively throughout the entirety of the Great Lakes Basin, so remote sensing allows us to get that synoptic view of the Lakes and see how different areas are changing at different rates or are forced by different things.

LAWSON:

Let's get a little more specific now about the project you worked on to map submerged aquatic vegetation, or SAV, with Landsat data. I know a lot of that is the Cladophora that you mentioned. So why is it important to know where SAV exists in the Great Lakes.

SHUCHMAN:

Well again a good question. Cladophora which is a submerged aquatic vegetation. It is classified by EPA and USGS as a nuisance algal. It is a filamentous green algo that grows on hard substrate, and one of the issues is during the spring and after Memorial Day, Cladophora will grow in length, and then when we have storm events, we call it sluffing, the Cladophora will get ripped off the hard substrates and then float to shore. For example, at Lakeshore Dunes you will have it on the beach, it smells bad and there are issues where you can get a botulism growing and it's resulted in major kill offs of shore birds. But really, it's such a nuisance for the tourists that they want to figure out is the Cladophora problem getting any worse? Now, the plot thickens also in respect to Cladophora. And I'll use the word "heaven forbid." Heaven forbid the Asian carp really get into the Great Lakes. One of their favorite foods would be Cladophora. Now if you're evil you may go "oh, that's good, let Asian carp eat the nuisance algae." But we really don't want Asian carp in our Great Lakes. And again, knowing where the large amounts of submerged aquatic vegetation, which is mostly Cladophora, will help us think about how could we remediate that. How could we better manage that? Because again being a plant much like Chemlawn, you have to be sustentive to nutrients going into the lake.

SAYERS:

Our work and our maps have also been utilized extensively by the research community who are really aiming to build models and really understand how the Cladophora grows, and the ingredients needed and the right mixtures in order to sort of optimize growth. So they use our maps to understand where these beds of this SAV are. And how thinking about hydrodynamics and things like that, and how as Bob mentioned nutrients are important, how does that get delivered to these systems? They really need to know where these Cladophora beds are to really understand the areas where these more physical lake characteristics are going to impact growth. And then where areas of problem may occur in the future due to point source run-off and things like this.

LAWSON:

Do you want to tell us a little more about your original mapping effort with Landsat and the recent update that you did?

SAYERS:

Early in the Great Lakes Restoration Initiative, the GLRI program, we were funded to generate initial SAV distribution in the Great Lakes for roughly the 2010 period. And to do that we utilized Landsat 5 & Landsat 7 data in conjunction with a depth-invariant index algorithm. The idea is to make reflectance from the bottom let's say sand in this case in two meters of water look the same or be the same value as it is in ten meters of water. And if you don't do that the bottom maps won't work in the sense that the different bottom types will be mapped incorrectly. That was really useful. That allowed the community to see where and how much SAV was within the Great Lakes. The EPA came back and wanted an update, essentially to utilize the same methodologies we did before, and remap the SAV and see differences. For that we utilized Landsat 8 which had recently come online for the community, and we were able to generate SAV Distribution Map for the 2018 period, which allowed us to look at some changes over time.

SHUCHMAN:

The only thing I'll add here is that our initial set of maps, they were well, well received by not only the resource managers in the Great Lakes Basin, but the research community. They would use our maps to do then very directed focus studies. In particular kudos to our Canadian colleagues, as well as our USGS colleagues here in the Ann Arbor area.

LAWSON:

What were some differences you saw then between those two time periods? About 2010 and 2018. 

SAYERS:

The differences were mostly localized in the sense that we didn't see any broadscale changes within a given lake. So we didn't see a generalized loss or gain of SAV, but we did see gains and losses in individual areas throughout the lakes. One of the interesting examples of that was the Sleepy Bear Dunes National Lakeshore in Lake Michigan, which is an area with an extensive shallow water area that is conducive to Cladophora growth and one of most heavily populated areas of Cladophora in the entire lakes where many of the fish kills or bird kills that Bob spoke about earlier occurred. In the update saw significant areas of gains and loss, which was first a bit unexpected. And so, being the scientists that we are we loaded up a boat and went out to verify that, and the areas that we showed as loss were predominantly related to mussel shell beds. So, as Bob mentioned earlier on the invasive species of zebra mussels and quaga mussels had come in and colonized on the bottom. And if you remember what Bob said, Cladophora requires a hard substrate, the shell beds provided just that. And so in that 2010 period we saw extensive SAV growth, Cladophora growth on these shell beds, but by the 2018 period many of these mussels had died and the shell beds were degrading and being covered by sand. So, we saw a lot of loss associated with these shell beds going back to a natural sand environment. So that was an interesting finding. We also saw some areas of gain which we actually associated with the increased water level between the two time periods. We saw roughly I think it was a meter and a half more water than in the 2018 period. We suspect that that allowed for Cladophora to grow into these areas that previously were too shallow, which allowed the wave action more velocity at the bottom so the Cladophora could stay attached. It was not conducive to growth. It also perhaps served as a purpose to regulate the amount of light at that depth, so the Cladophora was sort of in its optimal area.

SHUCHMAN:

We also, at EPA/USGS request, identified three areas in each of the five lakes, and we went really back in the remote sensing time machine into the 70s. And then each year we mapped changes at those local areas. And it was just fascinating findings, because there we saw direct relationships in respect to Cladophora growth as a function of increased light level, i.e. water clarity, juxtaposed against the reduction of nutrients do to the bi-national compact agreements where they really stopped putting a whole bunch of fertilizer into the Great Lakes. Just think about when you go from the early 70s to the present in the Great Lakes, the water clarity has increased so much that we are seeing much more of the lake bottom from space, from the satellites. So we had to normalize the area we saw to really separate out the cause and effect.

LAWSON:

So, really for a project like this; I know you work with a lot of different remote sensing methods. But Landsat was kind of key for this particular project.

SAYERS:

Yes. Landsat was really our workhorse for this analysis. And it's really because it had sort of perfect combination of spatial resolution. We could resolve these patches or Cladophora beds throughout the lakes as well as temporal revisit. So we were able to get enough looks throughout the year to be able to put together a full basin-wide map. And that's a bit of a challenge in the Great Lakes, because we're a cloudy region in a lot of cases where we only have something in the order of roughly 60 days of any location in the Great Lakes, roughly 60 days of clear weather each yea, so that limits our ability to use passive sensors to see the bottom. But Landsat provided us enough looks, especially when combining 5 & 7. And then 7 & 8 together. To be able to see that. And so, it gives us also the ability to look at the whole thing synoptically. So the tiles are large enough to give us a large area. It also provided enough spectral information in terms of the bands available and the sensitivities to be able to map the SAV well enough in this case.

LAWSON: 

You mentioned earlier that the Great Lake Restoration Initiative had funded this project. And that began in 2010. How's this initiative influenced your work? And how influential do you think the Initiative has been in being able to restore the lakes since 2010?

SHUCHMAN: 

The GLRI is frankly a godsend to cleaning up and better understanding the Great Lakes environment in general. As you mentioned, it started in 2010 and continues today. And what it's done is provided a steady funding source. EPA hands the money out, but USGS, NOAA, Fish and Wildlife Service, state and local governments all are recipients, as well as NGOs, some industrial partners and academic organizations like MTRI. Without this, I like to call it steady eddy money, we would be forced to do three-year studies here and there. For example, NASA does a great job funding satellite studies, but they fund them for approximately three years on a given subject and then they move on to some other interesting problem. And that's their charter. But the GLRI has really provided the federal partnerships and state partnerships and academic partners like us this constant kind of funding that we can now rely on. And it allows us to do a lot of these legacy assessments. And again, GLRI is not a research, they're not the NSF. That's not their job. When we use the word research a lot of times, what we're really saying is, we're developing a new remote sensing base techniques to do a better job of assessment. And that's really what GLRI needs to do their remediation and clean up the Great Lakes.

LAWSON:

Does either of you have any closing thoughts?

SHUCHMAN:

I guess my closing thought is again back to GLRI. I am very happy that it is a bi-partisan funded activity. Much like the Everglades has, Chesapeake Bay has, the Pacific Northwest, and there is another one in Alaska, where these activities are annually funded congressional initiatives. I just hope for the sake of the entire Great Lakes community that GLRI remains well-funded and continues.

SAYERS:

Yeah, and I will just mention some coming attractions, if you will, of things that we're doing. As hyperspectral sensors become more available and I'll say the norm, as we move forward, we're working to adapt some of our methodologies we've used here to utilize this hyperspectral data. Think about taking us from just mapping the distribution of SAV to thinking about doing more intense biomass types estimates. Or even classification of a particular other species. We're primarily looking at Cladophora but there are many species of similar aquatic type vegetation in the Great Lakes. And we're hoping that hyperspectral data may be able to move us into that sort of realm in the future.

LAWSON:

Well, thank you Mike and Bob for joining us for this episode of Eyes on Earth. And thank you to the listeners as well. Check out the EROS Facebook and Twitter pages to watch for our newest episodes. You can also subscribe to us on Apple podcasts. This podcast is a product of the U.S. Geological Survey/Department of the Interior.