Skip to main content

Dr. Kurt Thome

Talking about 20 years of NASA's Terra mission and its significance with Terra Project Scientist Dr. Kurt Thome.

For more than 20 years and more than 100,000 orbits of Earth, the five instruments aboard NASA’s Terra satellite have compiled an unrivaled data record of Earth and its interconnected processes. Dr. Kurt Thome has been part of the Terra mission since before the spacecraft launched on December 18, 1999, first as part of Terra’s instrument science teams and, since 2012, as the Terra Project Scientist.

Dr. Thome oversees the work of the satellite’s five instrument science teams, two of which (ASTER and MOPITT) are international partnerships:

  • Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER)
  • Clouds and the Earth’s Radiant Energy System (CERES)
  • Multi-angle Imaging SpectroRadiometer (MISR)
  • Moderate Resolution Imaging Spectroradiometer (MODIS)
  • Measurement of Pollution in the Troposphere (MOPITT)

Terra instrument data comprise nearly 20 percent of the 34 petabytes (PB) of Earth observing data in NASA’s Earth Observing System Data and Information System (EOSDIS) collection. More than 3,000 Terra instrument data products are currently distributed by EOSDIS Distributed Active Archive Centers (DAACs), with a majority of Terra data products (98 percent) located at NASA’s Atmospheric Science Data Center (ASDC: CERES, MISR, MOPITT), Land Processes DAAC (LP DAAC: ASTER, MODIS), and Level 1 and Atmosphere Archive and Distribution System (LAADS DAAC: MODIS).

Terra’s instruments continue to collect science quality data after more than 20 years in the most hostile environment known, and their data have contributed to more than 20,000 individual peer-reviewed papers that have helped explain the science behind numerous Earth processes. As Dr. Thome notes, there likely will never be another mission quite like Terra.

Terra was planned with an expected design life of six years. It now has been in orbit more than 20 years and all its instruments are still collecting science-quality data. What are some of the reasons behind the success of this mission?

Circular logo with the word

Some of it is luck, but most of it is great engineering. And by engineering, this is not just picking the right parts, it’s also properly estimating the lifetimes of these parts. The fact that these parts have over-performed on Terra really means that these parts performed properly, and in their proper performance they exceeded their expected lifetime.

You have to be careful not to adjust your expectations. There are the expectations that we started with on Terra, and we have far, far, far exceeded those. Part of this is because we have great engineers who built it, we have great engineers who are running it, and we’ve done a great job managing our power and our fuel, which are day-to-day decisions. There are a lot of things we do to optimize all the systems as much as we can.

When you’re watching from the sidelines you’re tempted to say, wow, 20 years. When you’re watching from inside you say, yes, this is what we expected to happen.

How has the retrieval, processing, distribution, and archiving of Terra instrument data evolved over the past 20 years?

The changes in data access have been amazing, and we’ve had to adapt to these changes. It’s not acceptable to have to wait for data anymore; data need to be available in real-time or near real-time and even the downstream, higher-level products have to be available shortly after they get collected.

Full disk image of Earth in super high resolution with North America in the center created from multiple true color images of Earth.
Image Caption

NASA's Blue Marble image of Earth created from Terra MODIS data. Data acquired February 8, 2002. NASA Goddard Space Flight Center image by Reto Stöckli; enhancements by Robert Simmon. Data and technical support: MODIS Land Group; MODIS Science Data Support Team; MODIS Atmosphere Group; MODIS Ocean Group. Additional data: USGS EROS Data Center; USGS Terrestrial Remote Sensing Flagstaff Field Center; Defense Meteorological Satellite Program.

Also, people don’t look at just one or two images anymore. They look at thousands or even hundreds of thousands of images, whether it’s 20 years of images from one site or globally for just one day. For example, the Blue Marble image of Earth would never have happened 20 years ago. By this I mean the data were there, but the ability to get the data and use them in the way that was required to produce this image took a few more years to mature.

One key [Terra data] strategy from the start was our use of NASA’s Tracking and Data Relay Satellite [TDRS] system. This is a way that NASA can communicate with a satellite without having to wait for a ground station to become available. We can download all [Terra instrument] data to a TDRS [spacecraft] and the TDRS can send it down when it is able to transmit the data to a ground station. This enables us to get our data from the satellite instruments pretty quickly, and means that we don’t have to be data-limited on the individual instruments. In this respect, we haven’t had to change how we handle our data downloads nor have we had to change the amount of data we’ve been collecting over the last 20 years.

You mentioned real-time and near real-time data. What is the significance of having near real-time data from Terra instruments available through NASA’s Land, Atmosphere Near real-time Capability for EOS (LANCE)?

It has been phenomenal, and has allowed the Terra instrument teams to invigorate their science since it brings them closer to their users. To make an image available in a few hours you have to trust that your instruments are calibrated properly, you have to trust that your algorithms are going to process those data properly, and you have to be willing to give the image to someone right away who knows that the image was processed more quickly [than a standard data product]. You have to be in constant contact with your community to know what they’re using these images for and know what corners you can cut so that they are not taking a loss in data quality from this rapid processing.

At the same time, [the process of creating near real-time data] creates lessons learned that you can apply to the next generation of instruments. For example, the Visible Infrared Imaging Radiometer Suite [VIIRS] instrument [aboard the joint NASA/NOAA Suomi National Polar-orbiting Partnership (Suomi NPP) and Joint Polar-orbiting Satellite System (JPSS) satellites] uses a lot of the lessons-learned from producing MODIS data in near real-time and direct broadcast funneled into what they wanted to use VIIRS for, which was operational weather forecasting.

While the ground systems acquiring Terra data and the algorithms used to process Terra data have been improved continuously over 20 years, the instruments aboard the satellite are still collecting data using late-20th century technology. What challenges does this create?

Having these systems operating for 20 years actually helps, because you can see how things trend. These are really well-built instruments that, for the most part, are behaving really nicely, so it’s much easier to identify these small changes over time.

Overall, there have not been many things that have happened over the past 20 years that have caused problems with how we handle the data, the way it comes down, or the way it gets ingested. Our data processing is still being done in the same way. It’s just been massive improvements to make the data more widely available, more quickly available, and easier to use.

Some of the biggest changes have been in the way data ordering happens. The interfaces with the users have continued to evolve. When the first Ethernet connections became available and became trusted, this was a huge improvement in not only the amount of data we could deliver, but the speed in which we could deliver these data.

We understand our user community very well and we are constantly striving—“we” as in EOSDIS and others—to make sure that the information is there for the users to let us know what we can do to make the data better or easier to use. In the end, better data makes my job as Terra Project Scientist a bit easier.

You oversee the work of five different instrument teams, each of which has their own agenda and needs. How do you balance the needs of these teams?

Mission logo showing the Terra satellite at the top and the words

There’s a Terra team mentality where everyone realizes that if one of the sensors on Terra benefits, it benefits the entire mission. For example, we changed the way we do our lunar roll maneuvers for MODIS because the MISR team noted that they were losing a bit of data near the Poles that were important to their processing scheme. All it took was someone on the MISR team saying that this was happening and we immediately all worked as a group to find the most equitable solution that would benefit all teams. We found a solution so that neither MODIS nor MISR took a hit. Once we knew there was an issue, they spoke up, we heard it, and we fixed it.

Ultimately, I am the one who makes the final decision, but it is a group agreement that comes into play into how we accomplish these decisions. So far, we have not had an issue on Terra where any instrument group felt that something was done that was bad for their group.

Terra’s ASTER instrument is an international partnership with Japan, and MOPITT is a joint venture with Canada. What challenges does this present?

When you have teams comprising different countries, you have not only cultural differences, but also differences in how governments work. For example, fiscal year boundaries can get really tricky when your fiscal year ends on one month and another country’s fiscal year ends on a different month and this impacts funding allocations for instrument support. In the case of our Japanese colleagues on the ASTER science team, you have to make sure you have an interpreter who not only can translate English and Japanese, but who also understands the science and the instrument.

The time difference also is a barrier. No matter what we do, if it involves multiple instruments, someone is going to lose sleep. Just the time zone differences will cause someone to be up overnight.

Once Terra data are downloaded, we send ASTER data to Japan and MOPITT data to Canada. Our Japanese colleagues process the exact same processing stream that we do in the U.S., just with different processors and with slightly different code. We make sure everything is the same. They distribute on their side and we distribute on our side, so there is a complete copy in Japan. For MOPITT it’s a little different. We process all of the MOPITT data at NCAR [the National Center for Atmospheric Research in Boulder, CO] and archive the data at NASA’s Langley Research Center, so it’s pretty straightforward.

Any disagreements we have are between scientists and engineers, not between Japanese or Canadian or American scientists and engineers. Scientists and engineers will always be scientists and engineers, regardless of country.

How much longer do you anticipate being able to get science quality data from Terra? What are your science and instrument teams doing to prolong the acquisition of science data?

We’re still on primary hardware and all indications are that the Terra platform and sensors can continue to operate and provide the high-quality data the science community has come to expect for the foreseeable future.

Other factors are more programmatic, such as funding and being a good neighbor to other satellites. Terra’s orbit will start to change after we do our last orbit maintenance burn in the spring [of 2020]. We’ll still have fuel to maintain altitude and for collision avoidance, but Terra’s crossing time will start to drift to earlier in the morning. The operations group has done a great job of keeping a constant crossing time for the past 18 years to a very narrow +/- two minutes, which is phenomenal. That has helped the data look more similar from day-to-day and from year-to-year. The operations group that is responsible for this is incredible and the effort they use to optimize fuel and maintain this has been just unbelievable.

Still, our fuel for that type of maintenance has run out, and we are keeping a reserve that allows us to do collision avoidance and eventually to lower our orbit when directed by NASA Headquarters so that we stay out of the way of other satellites. My hope is that we can continue operating Terra until our hardware limits what we can do, and the job of the Terra team is to make sure that the science justifies doing that.

Speaking of fuel, Terra launched with approximately 87.5 gallons of hydrazine fuel, and currently has about nine gallons of fuel remaining. What strategies do you and your team employ to maximize this dwindling resource?

The main issue [for maintaining the Terra spacecraft in orbit] is fuel, and we need fuel to do two things. We need to have the fuel reserve required to safely bring the spacecraft lower into the atmosphere and then de-orbit. The second item we need fuel for is maneuvers. There are two maneuvers we do. The first is debris avoidance and risk-mitigation maneuvers, both of which take very little fuel. The other is maneuvers to maintain our equator crossing time.

When our fuel started to run low, the question was, OK, what are we going to do? We realized that we don’t have to lower Terra’s orbit as much as we thought in order to keep other satellites safe; we just have to come down only a few kilometers, which means we now have extra fuel to hold our orbit longer than anticipated. NASA Headquarters decided that as long as we could lower safely then we could use the rest of the fuel for maneuvers to hold our crossing time. This gave us an extra three years at the current crossing time.

We’ll break our equator +/- two-minute crossing-time window in the fall of 2020. Then we’ll slowly start drifting in time and eventually get to a point where our crossing time occurs early enough that [NASA Headquarters] will ask us to lower our orbit. We can keep collecting data as the orbit lowers and the data will still be usable—we can make time corrections for a lot of the data we collect and we can update the algorithms, among other things. We can also learn new things by having an earlier crossing time and there are advantages to the science team to being able to do this.

The biggest problem when you start going to an earlier crossing time is that the Sun angle is lower and lower in the sky. Measurements that rely on sunlight start to struggle, and for instruments that rely on sunlight this is a problem. But because much or our data are collected in spectral bands where the human eye doesn’t operate or need reflected sunlight, like radiated energy, we can still collect data. We obviously collect a great deal of data in the dark with ASTER, MODIS, CERES, and MOPITT. There’s a lot of things that can still happen. We just need to make sure that we’re still doing good science.

What is the overall benefit to the science community to having remotely-sensed data from Earth observing satellites?

The big picture is the big advantage—you get data over larger spatial areas over very short periods of time. ASTER has a 60 km swath. To have an aircraft or a balloon do this would entail multiple flight lines, a lot of organization, and much more time. There are some advantages [to using an aircraft or balloon]: you can pick the time of day or a specific period within a growth cycle when there are no clouds, and you can get higher spatial resolution in some cases. But now you have to splice together maybe 100 lines of data where you flew out and flew back and flew out and flew back, and repeated this over and over. It might take two hours or more to do this. ASTER can acquire the same scene in a few seconds and cover everything you need in one 60-by-60-kilometer swath. Plus, you get another scene right after this and another over this same location 16 days later. You can even point ASTER to get to another location. With MODIS, we get the entire globe in one-to-two days; in some areas more than once a day.

This cannot be emphasized enough. There are areas you just can’t get to with ground or air measurements. This is the standard argument for all on-orbit assets that we have. The spatial coverage you can get, the temporal coverage you can get just can’t be matched with any other types of sensors.

Will there ever be another mission with the same impact as Terra?

I think it is good that we can say “no.” The next mission that is like Terra will have a different impact because it will have different instruments and a different objective and different scientists—hopefully a lot of young scientists that have a lot of new ideas about how to maximize the instruments and the data from the instruments.

It’s the younger scientists, the new scientists, that can come with all these great ideas about how we can do things better and more quickly. They can go straight to “I have good ideas and I want to implement them.” It’s a three-legged stool: you have the engineering, you have the data, and you have the scientists. None of them survive without the other.

Terra has had a fantastic impact on science. If we do our job right and make sure that the Terra data are archived properly and easily available and, more importantly, easily usable, then our impact will keep changing as new generations of scientists use these data and combine these data with other missions. This will ensure that we continue to get a new impact from these Terra data and that future missions will have an impact of their own.

Read More:

Twenty Years of Terra in Our Lives

Terra: The Hardest Working Satellite in Earth Orbit

Terra: Five Instruments—One Monumental Data Record

Details

Last Updated

Data Center/Project

Atmospheric Science Data Center (ASDC)
Level 1 and Atmosphere Archive and Distribution System DAAC (LAADS DAAC)
Land Processes DAAC (LP DAAC)
National Snow and Ice Data Center DAAC (NSIDC DAAC)
Ocean Biology DAAC (OB.DAAC)