Profile | Frank Kelly, Director, U.S. Geological Survey’s Earth Resources Observation and Science (EROS) Center
With Landsat 8 now seven months into its planned five-year mission, NASA and the U.S. Geological Survey (USGS) are moving out on a White House directive to come up with a long-term strategy for continuously collecting the type of moderate-resolution imagery of the world’s land masses that U.S. Landsat satellites have been gathering without interruption since the early 1970s. On Sept. 18, NASA formally petitioned industry for ideas what the agency is calling a Sustainable Land Imaging Program.
NASA’s Goddard Space Flight Center tapped Orbital Sciences Corp. and Ball Aerospace & Technologies Corp. in 2008 to build Landsat 8 — originally dubbed the Landsat Data Continuity Effort — after the White House overturned plans to fold Landsat-type instruments into a troubled civilian-military polar-orbiting weather satellite program ultimately canceled for cost and schedule growth. Just a few years earlier NASA had tried without success to find a commercial partner to share the cost of building a follow-on to the Landsat 7 satellite it launched in 1999.
The White House has made clear that it wants NASA — which is leading the effort with input from the USGS — to consider alternatives to buying a dedicated, government-owned and -operated satellite to replace Landsat 8.
EROS Center Director Frank Kelly, a veteran government scientist who joined the USGS in 2011 after stints with the National Weather Service and the U.S. Air Force’s meteorological program, wants his NASA and USGS colleagues to give serious and immediate thought to building a clone of Landsat 8, which was formally handed over to Sioux Falls, S.D.-based EROS in May. As the main repository and distribution hub for Landsat imagery, EROS has a clear interest in maintaining the continuity of its 40-year-plus data archive. NASA, as the senior partner in the White House-ordered Landsat follow-on study, has said it will give fresh consideration to commercial data buys, international partnerships and other alternatives to spending upwards of $1 billion for another free-flying Landsat satellite. The USGS, as the junior partner on the study, represents the interests of a Landsat user community that, according to Kelly, already knows what it wants out of the program near-term: a minimum of two Landsat spacecraft in orbit.
Kelly, who comes to the latest Landsat strategy debate with decades of experience in government weather-satellite programs, spoke recently with SpaceNews staff writer Dan Leone.
With the White House again pursuing an overhaul of the Landsat program, what is the top priority for Landsat data users?
We want to have two birds on orbit at any given time, and that goes way back into the history of Landsat with Landsat 1 and Landsat 2. With Landsat 7’s fuel running out in the next three to four years, we need something up there. There have been discussions that I’ve been involved with, both with government and industry, where a lot of people said after Landsat 8 launched, “Boy, we need to do the next one right now and there’s nothing on the table. Let’s build a clone and put that up as soon as we can.”
So you advocate building a clone of Landsat 8 as soon as possible?
What I’m an advocate for is getting a mission up there to be able to have two birds on orbit after Landsat 7 meets its decommissioning date, which will come at some point. That’s what I’m an advocate for. Does it need to be a specific clone? I think a clone in the near term helps us to maintain some efficiencies in our ground system and the way we operate the spacecraft. It would be very similar. We wouldn’t have to change a lot of things, like ground systems, to be able to support that. So I think there are some cost efficiencies in being able to do a clone. My experience in working operational programs — I was on the Defense Meteorological Satellite Program — is that you want to build follow-ons that are very similar. Your launch risk is lower, your operations risk is lower.
Why is it so important to have two Landsat spacecraft?
We have a mid-morning orbit with an equatorial crossing at roughly 10 a.m. That way, if you’re doing things with one spacecraft and you’re getting a 16-day repeat, you might only have two chances in a month to get over any point. And it might be cloudy when you get there. When you look at the way in which the land changes, the land doesn’t change, usually, on a day-to-day basis, not like your weather missions. So we don’t have to image at as high a temporal frequency as we would for a weather mission, but we want to do it often enough so that we can understand what’s going on with the plains, with crops, or how things are progressing with floods. By having two satellites, you get eight-day repeats, which meets, I believe, 65 percent or 75 percent of our users’ requirements.
One of the key directives from the White House, with the latest Landsat strategy study, was to find a way to reduce costs. Given that directive, is there any room to improve upon Landsat 8’s capabilities?
I’ve been in the government for several decades now, and it seems like I’ve had a boot heel on my head, in terms of budget, for a long time. As a scientist, I would want to see how we could use something like hyperspectral imaging, and how that could help us do many additional kinds of things. But some of those technologies still have to be proven before an operational program might really look at using them. So for now, we have to trade revolutionary technology for the continuity of the mission.
Has NASA held any meetings about the long-term strategy for Landsat?
There have been several meetings. There are two or three people we plan on having on that study from EROS. My concept is to use people who are spending time on Landsat 8 and taking advantage of some of their expertise and some of the confidence they’ve built up with NASA recently, going forward.
Since USGS made it free in 2008, there has been a swell of requests for Landsat data. But what has the free data policy cost EROS in terms of revenue?
This was long before I got here, but I think that was on the order of $6 million or $8 million during any particular year. Prior to the free data, EROS had the opportunity to have an income stream that it used for any number of activities. You could invest some of that into the infrastructure, or you could invest it into the science.
Would the EROS facility in Sioux Falls have to be upgraded in order to handle the data that would be generated by a Landsat 8 clone, should one launch in the next several years?
I don’t think our network is totally maxed out, even with another Landsat. The last addition to EROS was in the mid-1990s, so the center was built for computers of that era. We have a situation where our server and computer footprints are pretty small, so as far as space goes that’s not really a concern. The network that we have here is pretty well sized. We could accept additional data that would come in from a Landsat. We’re in the process of figuring out what we want to do with Earth imaging data from the European Space Agency’s Sentinel 2 satellites, when and if that’s made available to us in the not-too-distant future after they launch next year.