Commissioning the Landsat Data Continuity Mission
May 08, 2013 • On the beautifully clear Monday morning of Feb. 11, 2013, the Landsat Data Continuity Mission made its way from California’s Vandenberg Air Force Base to low Earth orbit. The more than 1,000 Landsat scientists, engineers, data users, and fans gathered to watch the launch breathed a collective sigh of relief as the bright orange light of the rocket’s booster faded into the distance and a successful launch—one of Vandenberg’s smoothest according to Launch Control—was declared.
But after years of formulation, planning, fabrication, and testing, the satellite’s work had just begun. Like a massive seafaring ship, a satellite is not ready for active service until a commissioning process takes place. You don’t launch a satellite on a Monday and get science data back on Tuesday.
For the Landsat Data Continuity Mission (LDCM) an approximately 100-day commissioning, or on-orbit checkout period, occurs before the satellite can be declared operational. The commissioning process is a highly coordinated effort that proceeds at a swift pace involving over a hundred people directly and hundreds more tangentially.
The major players during on-orbit checkout are the Mission Operations Team, the Ground Systems Team, and the Calibration and Validation Teams. These teams closely coordinate to make sure the spacecraft and instruments are operating correctly and that measurements made by LDCM’s two science instruments are accurate and that they are successfully transmitted back to the ground, processed, and made readily available to the public.
“Right after launch and for the first week or so, day-to-day operations are very busy,” says Vicki Dulski, the LDCM Ground System and Operations Manager. “The team must execute a series of procedures to send commands from the ground to the observatory in order to power on and activate all of the spacecraft subsystems and instruments.” To ensure that the Mission Operations Team can flawlessly follow the strict timeline of commands needed to verify that the satellite is working properly after launch, the sequenced procedures are carefully rehearsed prior to launch.
These important first steps initiate the on-orbit checkout. The commissioning effort can be broken down into five major components: (1) subsystems and communication checkout, (2) instrument initiation, (3) data processing, (4) calibration activities, and (5) orbit placement. During the 100-day checkout period, the commands, testing, and data collection necessary for these commissioning components overlap in a carefully choreographed manner.
The Commissioning Process In Five Acts
Act I – Wake up & talk to me
After being placed in its initial orbit by the Centaur launch vehicle, LDCM successfully deployed its solar array—an essential step because the satellite will rely on the solar panel as its primary power source for the life of the mission. Impressively, just over an hour after launch (19:20 GMT), the solar array was power-positive, meaning that the electrical current provided by the solar array was greater than the spacecraft’s demand. A team of engineers watched from the Mission Operations Center at NASA's Goddard Space Flight Center.
Part of the larger Ground Systems Team, the Mission Operations Team plays a pivotal role in directing the commissioning process. Engineers had been staffing the Mission Operations Center (MOC) for 24 hours prior to the launch, and 24/7 staffing would continue through the early phase of the mission. Their job, explains Dulski, is to monitor the spacecraft for any unexpected signs of trouble.
During the commissioning process, the core team of 20-25 engineers (called the Flight Operations Team) is supplemented with a surge team of as many as 50 people consisting of operation engineers from Goddard plus spacecraft and instrument experts from the vendor groups who built the satellite’s components: Orbital Sciences Corp.—builder of the spacecraft; Ball Aerospace—builder of the Operational Land Imager (OLI); and Goddard’s Thermal Infrared Sensor (TIRS) team. The surge team ramps down as on-orbit checkout proceeds. By the time LDCM reached its operational altitude on Apr. 12, the surge team was in a sustaining engineering role, meaning they are only called in if some sort of on-orbit anomaly arises.
On February 11, the first post-launch order of business for Mission Ops was checking the telemetry communication at each of the Landsat Ground Network stations. Telemetry refers to the wireless signals received from the spacecraft that provide status reports of the spacecraft health—location, battery charge, etc. “Even though a significant amount of testing occurs prior to launch, the extreme environments of launch and outer space could alter the performance in some way that could degrade the mission.
So it’s important to continually observe telemetry to make sure the observatory is performing nominally,” Dulski says. Each station attempted to collect telemetry data as LDCM flew within range. The ground station in Svalbard, Norway initially received a small amount of telemetry information, but the Gilmore Creek, Alaska station did not because of problems locking onto the signal. LDCM telemetry uses S-band communications, i.e., a radio frequency in the 2–4 GHz range. During the next communication opportunity via the Tracking Data Relay Satellite System (TDRSS), telemetry was successfully received and the spacecraft’s subsystems reported normal performance. After getting confirmation that LDCM was healthy, engineers were ready to command the satellite to power on heaters and supplementary communication systems.
The first few attempts to upload commands to the satellite were unsuccessful, but three hours after launch (20:55 GMT) a command was successfully uploaded to the satellite telling it to turn on onboard heaters and activate the satellite's GPS. Just before 1 a.m. the morning after launch, Mission Ops reported, “The spacecraft is performing beautifully.” LDCM is a land surveying science mission with the goal of repeatedly collecting information about the Earth’s dynamic landmasses. The satellite’s two science instruments make their measurements from the cold harsh vantage point of space where the spacecraft and its subsystems serve as the life support system for the science instruments, so it is essential that all of the components of the spacecraft work correctly.
After enabling the onboard heaters and communication assets, Mission Ops began activating the myriad spacecraft subsystems. A meeting is held each morning by the Flight Support Team to review and plan the upload commands, “the load,” that gives the satellite operational instructions. “After the mission planners build the load, the engineering teams from the vendors and within the Flight Operations Team review the loads each day to ensure there are no errors. If errors are found, the load must be corrected, re-built, re-reviewed, and then approved for uplink. This load planning, review, and approval cycle happens every day,” Dulski explains.
Over the next six days all other spacecraft sub-components were activated including the X-band communication link used for science data transmission and the Attitude Control System used for pointing the satellite. During each successive communication opportunity during the six-day period, systems were activated in a carefully planned sequence.
The last subsystem powered up was the spacecraft’s propulsion system, which performed an engineering test burn on Feb. 17. Mission Ops uses the term “burn” when referring to utilization of the propulsion system to move the satellite. “Just six days from our monumental LDCM launch, the LDCM Mission Operations team has done an amazing and outstanding job getting the spacecraft and instruments activated,” the LDCM Project Manger, Ken Schwer, wrote on Feb. 17. From Feb. 20 to March 3 the Mission Ops team executed a series of Attitude Control System maneuvers that pointed the satellite straight down at Earth (nadir), +/- 15 degrees off-nadir, towards the sun, and toward the moon (a lunar slew).
A few days later after a large piece of the thermal instrument (its Earth shield) was deployed as planned shifting the satellite’s center of gravity, additional Attitude Control System maneuvers were executed to compensate for the movement.
On March 15, less than a month after the propulsion system had been powered on, collision avoidance analysis showed that a piece of space junk (object 36907) might require LDCM to perform a burn to avoid a collision. The Project Manager wrote, “No reason for alarm, just part of our job.” Fortunately, the following day updated analysis indicated that a conjunction was no longer a concern and no burn was needed.
Act II – Turning on the instruments
Once the spacecraft subsystems had been successfully activated, Mission Ops focused on powering up the science instruments—the Operational Land Imager (OLI) built by Ball Aerospace and the Thermal Infrared Sensor (TIRS) built by engineers at NASA Goddard Space Flight Center. On Feb. 16, OLI and TIRS were powered on. The next day the instruments started generating test pattern data for transmission to the ground. Then, both instruments began a two-week dry-out period, i.e., procedures to ready them for use in space.
On Feb. 24, the cryocooler, which is responsible for cooling TIRS to its operating temperature of 43 Kelvin (-382° F), was uncaged from its special locked launch position and powered on.
On Mar. 4, TIRS’ large Earth shield was deployed. The Earth shield is a large panel that prevents heat from Earth from contaminating TIRS’ thermal radiators, thereby increasing the efficiency of those radiators. By March 7, TIRS was ready to begin test imaging, despite a snow closure of the NASA campus where Mission Ops is located.
A Flight Operation Team engineer tweeted, “NASA Goddard closed for snow, but all cryo hands on deck to send commands to Landsat to activate [TIRS] hardware.” A pixel map for OLI was uploaded on Feb. 26. The pixel map optimizes OLI’s performance by selecting the instrument’s best detector elements to use. On March 18, the first OLI image was collected in conjunction with TIRS data.
By March 27, two days before a scheduled under-flight of Landsat 7, OLI and TIRS reached a collection rate of 400 scenes per day, the operational mandate. Dennis Reuter, the TIRS project scientist, commented at an all-hands meeting, “Ops since launch have been near-flawless.”
Act III – Data processing
A fluidly operating data processing segment is essential. Getting quality Landsat data into users’ hands is the ultimate goal of the LDCM mission. “During the commission period, a lot of work goes into making sure that all of the various observatory and ground system components are functioning properly, with the goal of ensuring we successfully capture and process the LDCM images,” Dulski explains.
For the ground system segment, the major steps during commissioning are establishing ground station communications, verifying data throughput and processing, and incorporating data calibration measures. Jim Nelson, the LDCM Ground System Manager at USGS, likened establishing communication to tuning in a radio. “You set parameters at the ground station to match the spectrum and frequency of the spacecraft and verify that communication between the ground station and satellite is clean.”
In addition to the Mission Operation Center in Maryland, the Ground Systems Team manages three receiving stations that make up the Landsat Ground Network, located in Sioux Falls, South Dakota; Gilmore Creek, Alaska; and Svalbard, Norway. These receiving stations are responsible for downlinking the satellite telemetry and science data that feeds the USGS Landsat data archive.
The stations also have the capability of up-linking commands to the satellite. Additionally, LDCM has a network of International Ground Stations that tested their LDCM communication interfaces during the commissioning process. The telemetry downlinks were rapidly established after launch for each Landsat Ground Network station, but upload commands from Sioux Falls remained unsuccessful for 10 days until a faulty cable that was inverting the signal was found and replaced. Minor glitches during the first few days were extremely stressful for the team, because they didn’t know if the issue was on the spacecraft or ground end until the problem was identified and resolved.
After establishing the lower-rate S-band communications used for telemetry, higher-rate X-band communications, used for transmitting the larger science data signal, were tested. “S-band” and “X-band” both describe portions of the microwave radio region used for communication. The X-band is used to transmit both real-time and recorded LDCM science data.
When recorded, or “playback,” data is downlinked, “the ground system must close the loop with the [onboard] recorder, checking to see that expected files have been received and letting the recorder know that it's okay to delete those files that have successfully been received,” explains Landsat System Engineer, Terry Arvidson. Using test pattern data both real-time and playback data were downlinked to each station starting on Sunday, Feb. 17.
Additionally, “hot-handover” scenarios were tested in which two stations with overlapping receiving circles hand-off communications so data can be continuously downloaded. “Blind acquisitions” were also tested to make sure the ground stations could handle any unforeseen anomalies.
A blind acquisition tests Mission Ops ability to communicate with the satellite in the event that no communication is scheduled, but that telemetry and commanding is necessary, triggering the need for the satellite transmitter to be turned on via ground command. In addition to the three main Landsat Ground Network stations and the TDRSS communications link, a number of International Ground Stations have the ability to downlink real-time Landsat data when the satellite is within range.
Typically, the International Ground Stations are not involved during the commissioning process, but many stations are keen to get LDCM data as soon as possible, so six international stations successfully downlinked test data on March 25. Data downlinked to the Landsat Ground Network are transferred to the Data Processing & Archive System where they are ingested, processed, and archived.
All of the special conditions entailed with an orbiting satellite such as crossing the equator, poles, prime meridian, and international date line, must be tested to make sure the associated shifts in latitude, longitude, and date are correctly captured. On the calibration-side all of the lunar, stellar, solar, and onboard calibration data that is collected is downlinked, processed, and added to the archive so it can be analyzed and trended by the calibration team.
Starting April 25, operational daily imaging schedule inputs began being uploaded to the satellite, another milestone towards becoming operational. A demonstration of the 16-day imaging and calibration cycle that the satellite will follow for the life of the mission is underway. “All of the pre-launch testing paid off. Data processing is going incredibly well,” Nelson reported.
Part of the Landsat program’s legacy has become a tradition of rigorously calibrated data—i.e., information that accurately reflects physical conditions of the ground. A team of calibration scientists from NASA, USGS, partner vendors, and universities work together to monitor the performance of LDCM’s two science instruments.
This is accomplished by assiduously tracking the data collected by the satellite’s instruments, specifically, measurements of reflected light from sources that scientists already know well. These include onboard OLI lamp data (i.e. a source with a known radiance), OLI observations of a solar diffuser of known reflectance, TIRS observations of an onboard blackbody and deep space, stellar measurements (for instrument pointing accuracy), and lunar measurements (i.e. a target with static land cover free of atmospheric effects).
When collecting lunar calibration data, the satellite swings itself into position to point its instruments at the Moon; this is the satellite’s most complex maneuver. Augmenting the spacecraft calibration measurements, lots of ground images collected by the two science instruments are examined. Hours upon hours of careful planning and scheduling are done to make sure relevant, well-characterized calibration sites around the world are acquired for this purpose.
This copious amount of data is analyzed by the OLI and TIRS instrument teams in order to assess the performance of the instruments and fine-tune their radiometry, i.e., the quality of their measurements of Earth’s radiance. To establish that a given image pixel is correctly geolocated on Earth, an extensive amount of geometric calibration is directed and conducted by USGS’ Jim Storey.
The calibration work conducted during the on-orbit checkout refines the pre-launch calibration in order to provide the best possible data to users.
Landsat 7 Under-flight
Multiple field campaigns were also conducted when LDCM flew underneath Landsat 7 during its orbital ascent. The under-flight allowed Landsat 7 and LDCM data to be collected and compared when atmospheric and illumination conditions were identical. NASA, South Dakota State University, and the University of Arizona collected radiance, reflectance, and atmospheric field measurements during the three days of under-flights (Mar. 29–31).
Additionally, a NASA airborne sensor, known as G-LiHT, collected hyperspectral data over ground calibration sites and an experimental forest (i.e., well characterized targets) during the under-flight.
NASA and the University of Arizona made measurements in Arizona at Red Lake playa on March 29 and at McLaws Playa on March 31. Both teams made hand-held radiance measurements and initial comparisons of ground data collected showed good consistency. Additionally, hyperspectral and highly accurate (laser-calibrated) measurements were made by the airborne G-LiHT and a tripod-mounted radiometer called SOLARIS.
Despite the under-flight coinciding with Easter weekend, NASA had four people in the field, the University of Arizona had 2 people in the field, and South Dakota State University had three people in the field (at a test site in Brookings, SD). The G-LiHT instrument flew for ten days surrounding the under-flight. The G-LiHT team included the pilot and an instrument operator, Larry Corp—who worked through Easter and Passover to get the data collected.
In addition to the people in the field, members of the Landsat Science Team and scientists associated with some of the intensive study sites (such as Long Valley Experimental Forest in Arizona) shared knowledge and data that contributed to the calibration campaigns. “It really is a community activity,” Bruce Cook, the LDCM Deputy Project Scientist and G-LiHT principal investigator commented. The LDCM spectral bands on OLI and TIRS are slightly different than the spectral bands measured by Landsat 7’s Enhanced Thematic Mapper Plus (ETM+) instrument. The LDCM bands have been narrowed to reduce atmospheric effects. So it is expected that there will be slight differences in the radiance measured by the two satellites.
The beauty of using the G-LiHT and SOLARIS hyperspectral instruments to do calibration ground collects is that their many narrow-band spectral measurements can be aggregated to match the LDCM and Landsat 7 bands for more rigorous cross-comparisons. Joel McCorkel, a Landsat Science Team member and SOLARIS instrument scientist, explained that measurements made by the precise laser-calibrated SOLARIS and G-LiHT are important to compare with the Landsat 7 and LDCM measurements.
“We can link this new calibration technique from our laboratory instruments to the field measurements, and transfer that calibration to both Landsat sensors at the same time,” McCorkel explains. The large coordinated field efforts are undertaken to ensure that the true physical radiance of the ground is correctly captured by LDCM. This so-called vicarious calibration helps lock down what the satellite sees to physical ground measurements.
The field teams shared their measurements after the field campaigns and these values were compared with each other and the satellites’ measurements. “At the end of the day we want to be able to come up with the same answer with several calibration techniques and understand the reasons for any discrepancies,” explains McCorkel. This information will help refine LDCM’s pre-launch calibration and assure that data users get highly accurate measurements from LDCM.
Act V – Heading “home”
The launch vehicle places the satellite in orbit, but the satellite must use its own propulsion system to reach its final orbiting altitude of 705 kilometers. The Flight Dynamics engineers carefully planned a series of four ascent burns—maneuvers using the propulsion system—to boost the satellite into its final orbit. The first ascent burn took place on March 10 and four days later the second ascent burn was successfully performed, putting the satellite on track for the three-day under-flight of Landsat 7.
On April 7, a week after the under-flight, the third and longest ascent burn took place over 56.2 seconds. A 50.5 second ascent burn on April 12 delivered LDCM to its 705 km orbit and two days later, the propulsion system was used to orient the satellite at a 98.2 degree inclination (the angle between the equator and the descending satellite path). After which, the LDCM Project Manager, Ken Schwer, declared, “After many years in the making, LDCM has reached her final home!”
As of the beginning of May all of the major commissioning steps have been taken. Between now and the end of the month, the results of the on-orbit calibration will be incorporated into the data processing work flow and three reviews will take place. Then LDCM will officially become Landsat 8 and beginning on May 30, free, well-calibrated Landsat 8 data will be available for anyone in the world to download and use.
Contributor: Laura E.P. Rocchio