Interview by NOAA-CREST Undergraduate Researcher Valentina Rappa
Image: Wetland Variability in Radar Backscatter with Changing Seasons, created by Brian Lamb
Today I am interviewing Brian Lamb, a PhD candidate at the City College of New York studying remote sensing. Brian is here with us to discuss his past research projects, the trade-offs between different types of satellite data, and how drones can be used to augment information collected from space.
Valentina Rappa: Thanks for taking the time to meet today. First off, I wanted to learn a little about your past education. Where did you get your undergraduate and masters degrees from?
Brian Lamb: Well, I got my undergraduate degree from Cornell University. I was a Natural Resources major, which is a blend of environmental policy, management, and science; I even took an environmental philosophy course. As far as the actual science goes, I was mostly taking ecology classes. After undergrad, I got my masters from the City College of New York in technically geology, although it was more of an Earth and Atmospheric science focus as far as the workload and my research.
VR: So how did your past education inspire you to pursue such a prestigious degree in remote sensing?
BL: [Chuckles] Actually a big part of my undergrad was an illustration of the fact that I wasn’t well suited for doing the management and policy of environmental work. I wasn’t interested in that track nor did I do well in those classes, while in contrast I found the actual environmental science classes to be rather interesting.
Like I said, my courses were primarily ecology, although in my senior year I took a course on Remote Sensing for Environmental Resource Inventory. I remember the professor, Steve DeGloria, was a great guy and made the course material very intriguing; it was a really neat blend of computer-based work, physics and fieldwork as well. So I thought [remote sensing] would be a cool sub-discipline to go into.
VR: That’s interesting. So last summer you worked with high-resolution optical data to study the Chesapeake Bay for the U.S. Department of Agriculture. Can you tell me a little about that? Why was this type of data desirable for the project?
BL: We chose high-resolution optical data because we were looking at a small section of Maryland, rather than all of North America for instance. In addition, this project focused on mapping crop residue and analyzing agricultural processes within agricultural fields. In the Eastern United States these agricultural fields are pretty small so we needed high-resolution data to accurately observe the variation within those fields.
VR: You worked on a research project called Changing Hydro-logic Land Surface Dynamics of North America, where you used coarse-resolution data sets rather than high-resolution. Could you tell me more about that?
BL: Yes, of course. So that project, as you said, made use of coarse-resolution passive microwave observations with a spatial resolution of 25 kilometers.
For this resolution I was dealing with very large pixels, as far as looking at a grid of that data and imposing it onto a map. The course resolution data was really useful for looking at large scale hydrologic processes over large areas. In that project for instance, our domain of focus was the entire North American Continent, mostly in the United States and Canada.
We were looking at cryospheric hydrologic interactions in our data sets and essentially trying to derive climate indicators based on the length of the frozen season, the length of snow cover, land surface inundation and the amount of days where there was a transition between frozen states and thawed states. This transition generally meant that the ground, for instance, would be frozen during the daytime when there was sunlight incident on it — we were observing the amount of days necessary for the ground to become permanently thawed.
VR: Do you have a preference for this type of high-resolution optical data in comparison to course-resolution microwave data?
BL: Well, in the best-case scenario you would have large spatial extent and high-resolution data, but one of the concepts that really solidified when doing both of those projects is that there is always a trade-off between spatial resolution and spatial extent.
If there is not that tradeoff then there will be a price to pay as far as the amount of processing and lack of capabilities, as well as for data transmissions to be able to relay those satellite observations from ground stations to various servers and then to your computer where you are processing the data.
Considering satellites cannot tell you much about what is below the surface of the water, data collected from UAVs can really compliment satellite data. So yes, there is a clear trade-off between having a large extent, having a high resolution and having adequate processing capabilities or data storage.
VR: I see that you are currently using radar data for your PhD. research. Can you explain the major differences between coarse-resolution passive data, high-resolution optical data, and radar data?
BL: So in either case you can really think of the satellite as you would almost your own vision. With the coarse resolution data it would be analogous to you being at a far distance from a target and being able to see a large area, while high-resolution data is comparable to you being close to a target and being able to see it in great detail, although you can no longer view a large special extent. Satellite data resolutions are not very different from that.
Passive microwaves make use of the Earth’s natural emissions. As a physical body of a given temperature, the Earth will emit microwaves, while high-resolution optical data is just essentially measuring the radiation that reflects off the Earth’s surface from the sun. They are both passive measurements in contrast to the radar data, which is an active measurement.
A radar will emit a signal to the Earth’s surface and then will measure what is returned — unlike measuring the sun’s reflectivity, it is measuring the reflectivity of its own signal. I’m currently using active radar data rather than passive data because I’m studying wetlands a
nd the spatial extent is not very large, so you need a finer detailed instrument for studying the variability within them as far as the vegetation and hydrological dynamics.
VR: Do you think your future with remote sensing will expand towards UAV [autonomous aerial vehicle] data rather than solely satellite-based data?
BL: If I continue on the trajectory I’m on, going from upland agriculture to wetlands and then further down into the ocean, perhaps that could happen. The wetlands remote sensing project that I am currently working on ties into a larger-scale carbon cycling project. That project is focusing on the lateral export of carbon from marshes into estuaries. I am using remote sensing to understand how we can image different points in tidal cycles because this is shown to be a primary factor driving the export of carbon from marshes into adjacent estuaries and in turn into the ocean.
As far as where the UAVs could tie into this project, I would say definitely in the actual monitoring of carbon in the marsh or estuary water. It would depend on what instruments are on the UAVs and their capabilities, but that’s a possible application given carbon dynamics are the real main focus of the overall project.
Outside of the scope of my current project, I find that UAVs could be a really good way to validate satellite data, especially when looking at water and ocean coloring. Considering satellites cannot tell you much about what is below the surface of the water, data collected from UAVs can really compliment satellite data.
VR: How do you plan to use your doctorate degree in remote sensing for your future career?
BL: I would love to continue with remote sensing, preferably in a research position. I guess it doesn’t necessarily matter if it is academia, federal or even private, I just really like the field. I have found that physical science and computer work compliment each other very well and is something I truly enjoy.