Mark Rosiek

Intro:

Mark Rosiek is a planetary photogrammetrist with an Air Force and a terrestrial remote sensing background. He worked on producing topographic maps but also made efforts on the archival of old planetary image documents. He speaks about missions such as Huygens, policy questions and the role of the military from multiple aspects of planetary mapping.

Rosiek:

I started working in 1980, and now 40 years later, things are completely different. Photogrammetry and remote sensing is my background. And so, I started out doing research–well, I worked for a beltway bandit for a while around Washington D.C., and then I went to work for the Air Force. And I worked at the Air Force for 15 years, and then I came here to USGS. So at the Air Force, it was more research and development of image processing, and working with satellite imagery, and things like that.

Hargitai:

So it was not planetary.

Rosiek:

No, it was not planetary. So from 1983 through 1998, it was focused on terrestrial images and just the transferring from hard-copy photographs to digital images. And just everything you see now on Google, the Air Force was doing back in the 80s and stuff.

Hargitai:

Is what you do more engineering or more science?

Rosiek:

I would say engineering. The science is more in the photogrammetry. We're making maps so people can do planetary science.

Hargitai:

Concrete projects. It's not in chronological order. The first is the Titan Huygens landing site. What's behind this landing site mapping project?

Rosiek:

For Huygens, the bad part about that is, one channel didn't turn on, and so they didn't get all the images back. Because the way the spacecraft was set up, to get the data back, they had two communications channels. And one of the channels didn't get turned on for whatever technical glitch that happened, whether it was a software programming error or what. So we didn't get all the images back, but we got some. And so, as it came down, the spacecraft was rotating, and so the cameras were clicking. And it was supposed to collect three types of images, low-resolution, medium-resolution, and high-resolution. And as it came down, the resolution would increase, too. And so, we got some images, but we had to piece them together. And Annie and Randy did most of that work, and I just helped out a little with that.

Hargitai:

And is there any map sheet output from that project, or that's just a scientific paper that resulted?

Rosiek:

Yeah, I believe it was just that we produced small, little sets of data. But there was nothing to do a map sheet with.

Hargitai:

Because…

Rosiek:

It wasn't big enough. And it was mainly, "What sort of topography information do we have?" And after a while, it gets away from having a map sheet, which everybody's used to. And now, it's more digital data, and it doesn't matter. It's just, put it into the computer, and we can manipulate it and use it with other data.

Hargitai:

Can you identify when this transition happened.

Rosiek:

I came here in 1998. And so, there was still this emphasis on map sheets. Before I came here, there was an atlas of the solar system they did, and it took a lot of time and a lot of effort to do that. And now, that seems that putting together, like, a book of an atlas or putting together a map sheet and having a map series isn't as useful because it's all on the computer now. And it's like having the digital data and having that properly annotated is what's needed. If you look for the United States, USGS puts out a product that's digital for the entire United States.

Hargitai:

So it happened in parallel with terrestrial mapping change, it also changed to digital?

Rosiek:

Yeah. Because it's like, as technology changes, everyone has to go along. Technology's pushed forward by different groups. And so, like, in the 80s, the gaming industry was really pushing digital processing and computer graphics work, and then everyone else could take advantage of that as that moved along. And so, as cameras went from film to digital, Kodak sort of went away because they didn't keep up with it, even though they have enough patents to.

Hargitai:

I only thought of the military as the driving force behind all of the high tech. But you say gaming is another.

Rosiek:

The games. Yeah, because once you get something to be commercially viable, it's like, "Well, we can make money with this, and we'll push the technology there." So, like, NVIDIA, they were a big gaming chip maker. And those chips got used more in graphics systems. And as the military needed graphics, it was like, "Okay, we can take advantage of what the graphics people are doing for gaming." It's like, "Where can people make money with this technology?" And that pushes it. And then, the military was trying to push it, but it was after a while, they realized they weren't in control of pushing it because there's all these other uses for it.

Hargitai:

Martian contour maps. I found many sheets from 2003 to 2005.

Rosiek:

Randy Kirk had an idea, and he showed it to people before I was here at NASA, about how Viking could be used for stereo mapping. But the problem was that the position of the images wasn't all that good. So as I started to do these map series, Randy had this database that said, "These images should be viable for mapping," so I had to get the images and look at them to see if they actually aligned up together. Because it could be like, "This image is not even in the map sheet," because the pointing on Viking was so poor. It's just the way the data was collected back in the 70s is that trying to position the spacecraft was not as accurate as they can do now. Lunar Reconnaissance Orbiter that was used in the 2010 to 2020 timeframe, that started off really phenomenally registered, and it got even better as they went along because they did things with the orbit determination, improving that with the gravity field at MIT, and then ASU did more, updated the orbit information on the satellite. And now, all that information comes together a lot cleaner.

Hargitai:

And what was the target audience for these maps?

Rosiek:

There were scientists who were mapping different areas for geologic maps and doing other studies. And so, the way it worked is that we would talk to people who were doing mapping for NASA and ask them if contour information would be useful for them. The data we were producing was okay. I'm trying to remember what mission started in early 2000. And so, they were getting better laser altimetry data. And so, that sort of says, "Viking was okay, but we can get better data from the laser altimeter."

Hargitai:

So maybe that was the last project with Viking.

Rosiek:

Yeah.

Hargitai:

Okay, so the choice of the region is determined by the subsequent geologic mappers regions.

Rosiek:

Right. Yeah.

Hargitai:

I found digital files, USGS lunar south polar DEMS, or maybe north pole as well.

Rosiek:

Under George Bush in early 2000, there was a push of going back to the moon. And so, Marshall started their program there, and we started working with them. And they had 50 sites that they wanted done. And then, also, it was this interest in doing things at the polar regions. And they had some data there, and so it was like, "Can you try and make that?" Because they're trying to go back to the shadowed regions at the poles. And I think it was the way the Lunar Reconnaissance Orbiter was set up is, they had better and higher-resolution information at the south pole.

Hargitai:

Why it never made it to a map sheet?

Rosiek:

I think mostly because it wasn't enough data to make into a series. And also, it's like a map sheet suddenly has no more use because everyone's doing things digitally. Elevation data can be studied easier in the computer.

Hargitai:

I found this Apollo flight film project. Where are these originals kept and how they are digitized?

Rosiek:

When I was here last–we had, like, a third or fourth generation here in the warehouse. The original Apollo film, when it came back from the moon, got frozen and put into–I believe it was at Houston. And by freezing it, it keeps the film from degrading. And so, we started talking about digitizing it, and Mark Robinson down in Phoenix, he got a project to digitize it. And so, that put all that film online.

Hargitai:

The originals?

Rosiek:

Yeah, the originally stuff that actually–the rolls of film that got picked up on Earth, flown to the moon, and exposed at the moon, came back, and went to Houston, and got put into a freezer. So when they were coming back from the moon, someone had to come out of the capsule, go back to the back part, and pull out the rolls of film, and bring it back, and put it inside the capsule so they could bring it with them. And so, it was like, "Okay, open up your camera, pull out the roll of film, but go outside and pick up a roll of film, and bring it back into the capsule so we can bring it down." So that film came down here and got exposed on Earth. They did the chemical processing to make the film negatives. And then, they made other copies of those, too. And they got distributed. But the originals all stayed at Houston, and that was what was digitized.

Hargitai:

I found a paper from 2008, and the title was What is Planetary Cartography, and Why Does it Matter?

Rosiek:

Well, in 2008, the technology transition is that Photoshop and other techniques, robotic vision, was starting to come on. And so, it's like, do you need positioning information that says, "These are the coordinates of this place?" Or do you just need relative positioning that says, "Well, I'm in this room, and this table is at this location, the chair is over there," so when you're on another planet, it's like, "Okay, whatever sensor or robotic piece of equipment is up there, it's located here. Where do you want to go?" And you get to go out there, look at a rock, and do you need to know the exact size of a rock or just a general description of it? To me, it's like, if you want to start collecting data over time, and make sure it's all in the same position, and you can do time studies of saying, "How did this area change, both geomorphology? Is this a new crater? Is this crater eroding?" You need to have positioning information.

Hargitai:

You say that planetary cartography extends to the rock-size scale from planet-wide size scale?

Rosiek:

Yeah, if you're going to up and actually be on a planet–like, here on Earth, when they want to build a bridge, there's a certain type of surveying crew that goes in to make sure the bridge gets built right. So if you're going to go up to a planet, and put people on there, and start building things, just having a habitat land there is one thing. But if you start trying to build up a base on the moon of people and buildings, you're going to have to have some surveying to see how things are connected. And that can be done in different ways. If you're trying to map the entire body, you need to have an absolute position system. But if you just want to map a chunk of ground, how do you say what's level? When you're doing hydrology studies, you need that. And so, when you're coming to a planet, it's like, "Okay, how do we decide what's level?" Because there will be gravity. And so, how do you know you're going uphill versus downhill? If you're there, you physically can feel it, but how do you map it when you can't see it? Unless you have an absolute position, you don't know what to level it to.

Hargitai:

Is there any major project that I didn't mention?

Rosiek:

Let's see. For the moon, there was Clementine and then Lunar Reconnaissance Orbiter.

Hargitai:

What's the significance of the Clementine mission?

Rosiek:

After not doing any missions for a while, that was sort of a thing between, I think, the Navy or someone in the military was running it, because they couldn't test things on Earth, they decided to test it at the moon. And it was like, "If we're going to do this, we'll do some things up at the moon." Well, that provided some multispectral stuff and provided more global coverage. It was a lot of small images that had to be pieced together, and so that made it difficult. And so, when I started working it, they had some mosaics out and some networks out. But as I worked with it more, things started to show where they had some errors that they didn't detect before. And as you start using it with the Lunar Reconnaissance data or trying to get prepared for going back to the moon and getting this data to join together from the different missions. You could see that things weren't aligning. So there were certain errors in it. Clementine came up with a network, and then that suddenly got surpassed by everything Lunar Reconnaissance Orbiter did.

Hargitai:

Today it's not used?

Rosiek:

Some of the multispectral data was still useful. Saying, "This is how the moon looked at this time," so if you're doing crater studies, it would still be useful.

Hargitai:

Going to more generic questions, first about training. The skills or knowledge that's needed for your work, the planetary one, how did you get that? Was it in-house training, or were university courses enough for this work?

Rosiek:

There's a lot of on-the-job training that goes with planetary, and it changes with the missions because the different missions use different technology. And so, I went to the State University of New York's College of Forestry, and they had a Photogrammetry and Remote Sensing Department. That taught me about the metric information, trigonometry, doing aerial triangulation. And when I went to work for the Air Force, that just expands to working with satellite data because that became more important. And so, as a technology changes, you have to keep up with it. So there's a lot of stuff you have to learn as you go along, too.

Hargitai:

Just on your own.

Rosiek:

Well, on your own and–at the Air Force, they were good about providing training, and there was some training I could get here at USGS. But for the planetary data, because it's not as big a field, there was less official training and more just talking to people about, "How does this work?"

Hargitai:

And then, you talked about your colleagues here. Was there any organized–like, a small class that wants to learn?

Rosiek:

It was mostly a small group of people talking, and someone would have a problem, you would talk with other people, and it's like, "Okay, I'm doing this, and you worked on this mission 20 years ago, what was going on?" So, like, with the Clementine data, when I started working, with it, it was like, "Okay, here's the people that worked on it," and we'd just go up, and try to arrange a meeting with them, and talk with them. Or if you had a problem, you would talk with people. The benefit of textbooks is, you can read something and then apply it to what you're doing. It's like, "Okay, I'm having this difficulty, or I'm trying to solve this problem." You can do a literature search and see what other people are doing. And even though it's not planetary, you can adapt it to what you're trying to do with the planetary data. But when it gets down to, like, "Here's the Lunar Reconnaissance Orbiter. How was it calibrated, and how do the images work?" You have to talk to people about that. Because it's like, "Oh, yeah, we changed that software, so that's why that's not working anymore. You're going to have to update it here."

Hargitai:

So it's the mission team.

Rosiek:

Yeah, it's the mission team. And with the papers, that helps disseminate it to everybody so everybody can keep on board with it, to make sure there's a record of how has this mission changed and how the processing of the data's changed.

Hargitai:

So the project participated in, was there any one of them that was significantly more difficult or more complex than the others?

Rosiek:

It's always changing, and they all had their own complexity and things. When I first came here, it was like, "I'm just trying to learn about how does this planetary data work and stuff." And then, the technology wasn't as good, so you had to make some workarounds. And then, when you started doing stuff with Mars, there were other issues coming up with data, and calibration, and things like that. It was like each mission is unique, I would say.

Hargitai:

Was there any cooperation in your work with other groups who work here, like geologists, or the airbrush group, or other programmers?

Rosiek:

When I came here is when the airbrushers were going away, and they were all being given the opportunity to be trained to become digital processors. So that was in 1998 or so.

Hargitai:

And is there anyone who transitioned?

Rosiek:

Yeah, there were a few that did transition type of thing. Some people weren't able to pick up on the digital–because they were used to doing physical work, painting lines and things like that.

Hargitai:

The new work would be, like, image processing and computer?

Rosiek:

Yes. It would be like using Adobe Illustrator to do a map layout and having to do digital lines instead of sitting there at a map sheet and making layers of paper that went into the printing press and things like that.

Hargitai:

Did you have a more direct connection with the geologists, for example, that requested things?

Rosiek:

Yeah. Here in the building, Ken Herkenhoff, who's a geologist, he was doing a lot of stuff with Mars that I worked with. And so, we would talk with him, and it's like, "Well, we're having this issue," or we would make a map, and then the geologists who were driving the rovers were saying, "Well, we're at this hill, and we want to come off the back site. How certain are you of the slope coming off of that?" Because it was like, "Okay, we're going to drive the rover down, but we can't see over until we get to the edge. And if we get to the edge, and it's too steep, we can't back up." And so, it was like, "Go back and check this data and tell us how accurate do you think it really is." Or there was one time, just the way we drew the digital elevation model, it looked like there was a small ravine somewhere, maybe two or three meters deep over a 10-meter area. And so, it was like, "Well, if that's true there, then that's significant. We want to go see it. Is that really true, or is that just a shadow type artifact?" We said it's probably a shadow.

Hargitai:

You have mapped many bodies and many regions. What's your favorite as if you were an astronaut, and you could land in any of those? Which one would you choose?

Rosiek:

I'm trying to think of the book from–Mao or someone. There was a monk who was doing maps when people were going around exploring the world. And he sat in the monastery making all the maps. And I never got the book, and I wanted to because that's how I felt, that I'm sitting here on Earth, and I don't have to go anywhere, I can explore all these places by just sitting here in Flagstaff, Arizona. And all this information is coming here to me, and I can sort of get a feeling for that area. And so, to me, the moon, just because I see it every day, is interesting. And because it's someplace I feel like I could go to someday. I could go spend $2 million and fly around the moon or something. [Laugh] Not that I have $2 million, but you could always win the lottery or something.

And then, Mars, just because the features and the canyons were just so interesting. With Huygens going to Titan, it was just so different. And so, it's like, some days, you go in, and one crater looks like another crater, you're just really frustrated. Other days, it's like, it's just fascinating to sit there looking at these images, and thinking of the shape, and the size, the body. You go up the Grand Canyon, you can imagine what that looks like, and you look at these other canyons that are so much bigger, and deeper, and different. You can sort of visualize that and feel like you're there.

Hargitai:

In your choice of being a photogrammetrist, and later going to planetary, can you go back in time and find the early influences that influenced you going towards this field?

Rosiek:

Sure. I was sitting around a campfire when I was in high school, talking to somebody about what I should go into. And we were just talking to him, and he says, "Well, based on what you're doing," he recommended the College of Forestry in Syracuse to me. And when I was there, I didn't like fluid mechanics, and so I went into the photogrammetry and remote sensing operation because you could go outside and do surveying, and that just seemed more natural to me. And I liked math, and trigonometry, and statistics, so that fits in with the field also.

And so, then when I graduated, it was like, "Well, where can I go to work without experience?" And so, talking to people I went to college with, it was like, "Oh, there's an opening down here in D.C. They need a photogrammetrist." But it was with a private company doing work for the military. And then, I wanted to move out of D.C., so I went back to graduate school. That didn't work out as I planned, and so I started working for the Air Force up in Rome, New York. So that was really fun because that was a lot of research and development. And it's like, "Okay, how do we do digital image processing? How does digital image processing compare to processing hard film? Are we getting the same metric accuracy?" and things like that.

And that was good, but it was starting to mature, and it was like, "Okay, this is all getting automated. We solved all the problems." And then, there was an opening here in Flagstaff, and it was nice because it was more of a production thing. But it's still research because with each mission, there's problems to solve. So it isn't like a production, like with Google Eye or one of the Earth-based stuff, where the same product comes down every day, and it's, "How do we produce it quickly?" Here, it's, "A new product is coming down. How do we do something new and different with this to solve this new and different problem?"

Hargitai:

And this opening here at USGS was specifically planetary?

Rosiek:

No, it was specifically planetary because it was for the astrogeology group. When I was at the Air Force, I did work on one project that some people from astrogeology were on. So it was like, "Well, that's a fun place and sounds like a nice place to be." And I was sitting back in Rome, New York and in the journal, they had a little ad for it, and I said, "Oh, I know about that. That's something that would be good to do." It was a nice transition at that time to go from really focused on research because around 2000, a lot of the research for digital processing and digital mapmaking was done for the Defense Mapping Agency. It was becoming much more of a production thing. Global positioning systems are coming on, inertial navigation systems were becoming smaller and lighter. So now, you have drones. And because of the inertial navigation systems and the GPS systems, you can fly those around and collect all sorts of images on the Earth.

Hargitai:

With physics, or maths, or surveying, can you remember any early influence, like books that you read that made you think, "This is the thing that I like"?

Rosiek:

At the College of Forestry, it was a small group. And so, for the people doing surveying and mapping, there were just six students, and then there were several teachers. So, like, the surveying thing was very popular, so there might've been 40 or 50 students in that class. But for the photogrammetry and remote sensing group, they were doing things all through the college with different students. But the students who were actually doing the coursework, there were about six of us in our graduating class. And so, it was just working with the people and working with those teachers more on a one-on-one basis that made it seem a lot more fun.

Hargitai:

So it's more personal influence than media, like books.

Rosiek:

Yeah. Yeah. When I was in high school, I knew nothing about photogrammetry, remote sensing, or cartography.

Hargitai:

Can you explain who were the people who most strongly influenced the way research is done here? Who were the most influential leaders here?

Rosiek:

Okay. Well, when I came here, Wes was the head of the group, and he would interact with someone from NASA. And so, it was like, NASA would say, "Okay, here's what we need," and Wes would say, "Here's what we can provide," and they would sort of negotiate it. And before I came, like, in the 60s, it was like, "Here's what we would like to do," and NASA had lots of money, and anything you did was new and important. Then, it became much more production. It's like, "Okay, this is how the program is doing and what we want to do at NASA. What can you do here?" And so, it was all this talking between astrogeology and the planetary person at NASA. It still continues, but it changed over time. It used to be astrogeology had a little more free rein on what we did, and then it became, "No, we have to do things differently," budgets got tighter, things like that.

Hargitai:

How are universities part of this work group that produces planetary maps?

Rosiek:

Yeah, that's changing, too, because robotics was more into the university stuff, and Ames was doing more robotics. And that's becoming more of how people figure out where to go and what to do. Robotic vision is different than sitting there, trying to make a metric map of something. And then, universities could do things differently, and as the technology became more accessible, universities could start doing this stuff. It wasn't like you needed a big fancy plotter that you needed for lunar mapping. Suddenly, you could do it on a PC type computer.

Hargitai:

So then what's the role of USGS in this ecosystem? Or what's the changing role of USGS?

Rosiek:

Yeah, that's a good question. And since I've been gone for 10 years…

Hargitai:

Not today, but historically.

Rosiek:

Yeah, historically. Like, for the lunar mapping in the 60s, it was like, you needed a group of people who understood photogrammetry and remote sensing, and you needed the technical expertise. And now, that information is becoming much more easy, and so it's like, it changed from, like, you have to have specific experts doing something full-time to, "Oh, this technology is something a university student can pick up and start doing on their computer." You need some group of people to keep things managed, to say, "Okay, here is the body we're studying. Here's how the coordinate system is defined," and to understand that, and to make sure everyone's doing things right. So it becomes more of a technical management issue, I think.

Hargitai:

How is IAU then? Because IAUs work in groups. The IAU is an organization that the actual work is done here at the USGS, or is there something within the IAU…

Rosiek:

Well, see, all over the world, people are doing different planetary cartography things. And it's like, if you want the entire world to follow a similar system, everyone has to be working together. And so, you need some sort of organization–you need organizations within each country that are sort of defining how things are mapped and keeping track of things. And so, it's like, for the download stations and things like that, you need people understanding that. For people who are tracking bodies and things like that, you need that. And so, somehow, there's this generic need for people to understand what's going on. The thing is, how much of that expertise is still needed? In the past, it was recognized as, "This is the only way you're going to have this expertise is having people trained, and studying, and doing this work on a professional basis." Now, is there enough information out there that you don't need a professional group of people, but just keeping the information out there, and people can come in from a university, study it, work with it, go? The trouble is keeping it up to date and unified.

Hargitai:

Why do we need planetary maps?

Rosiek:

If you're going to be doing things on a body, you need to understand where things are positioned. And in order to talk about it, everyone has to be talking in the same system. Because there are things where sometimes craters got double-named because they didn't realize these images were of the same place, just because the positioning was off in the past. And now, going into the future, if you want people to study the planets, what is the reason for studying the planets? It's like, how do planets evolve, how are they shaped, what is out there? And if you all want to talk about it, you have to have some way to collect all that information and say, "Yes, we're talking about the same piece of dirt on this planet."

Hargitai:

So what's the argument today that they still need to give public money for more planetary mapping?

Rosiek:

I think as the technology improves, you want to go back to these same planets and study them more. If you're going to have a planetary program of exploring the solar system, you're going to want to be able to bring that information back, and you need some sort of metric understanding of it. You also need radiometric understanding for wavelengths and things like that. Because if you just start bringing this information back, and you don't really have a way to control it and understand it, it's like, "Well, how big is this crater?" If you're trying to do some analysis of it, slope information is needed. And unless you have some consistency in your metric information, it starts to become a hodgepodge of stuff, and you can get errors introduced. I think if you're going to want to study stuff, you have to have some sort of way of doing the bookkeeping.

Hargitai:

Is there an optimal endpoint of this, making it work more precisely [inaudible] metric?

Rosiek:

There probably is, but there's a need to update it. If at some point, you could say, "Well, we don't really need to do this, there's nothing interesting to do"–because it changes. If you look at technology, it's like, "Okay, this area is really interesting in technology." Then, suddenly, it changes, and something else becomes more important. There are new craters being formed. If we want to go, and explore, and put people on the moon and Mars, there's a need to keep up the information. But after a point, we have enough. But you still need to update it and keep it fresh. And when you go sending a spacecraft out there, and you're collecting new information, you still want to be able to join it to other information, so you still have to do some sort of metric processing and radiometric processing of the data.

Hargitai:

Is there any project that you wanted to do but you couldn't for any reason?

Rosiek:

There was a time when we were putting together a proposal to map the moon with a frame camera. So most of the cameras they're sending to planetary systems now are line-scanner cameras. As the spacecraft is moving, this linear scanner's filling up with data. They stop it after nanoseconds, and they collect that information, and it keeps moving. So that's not as stable as if you have a frame camera, you take a picture of the entire ground at one instant in time. And so, it has a little bit better metric properties. I was always interested in trying to do something with framing cameras on the planets.

Hargitai:

Is there any plan to?

Rosiek:

No, I think it's basically much cheaper to make a linear sensor because it's maybe 1,000, 2,000 pixels by 1 pixel wide. If you try to make a chip thing for our phones and things like that, it's easy to do on Earth, but when you send it in outer space, it has to be hardened for spacecraft. And you could start losing pixels. And so, I don't know if it's worth their effort to make a space-hardened framing camera.

Hargitai:

Thank you very much.

Rosiek:

Sure.

[End]