It's a hard sol's night at Space Sciences, processing those Mars images

By Bill Steele

Behind almost every great historical event, there are people toiling in the trenches.

For the Mars rover mission, that includes the Cornell Calibration Team, a mix of staff, graduate students and undergraduates who perform critical processing steps on the raw images sent back from the planet, in the MarsLab in Cornell's Space Sciences Building.

Images are received at the Jet Propulsion Laboratory (JPL) in Pasadena, the manager of the mission, but immediately transmitted to Cornell for processing and calibration.

Graduate student Jason Soderblom (on computer screen) at the Jet Propulsion Laboratory (JPL) in Pasadena chats, through a high-speed computer link, with grad student Eldar Noe, center, and freshman John Roma Skok in the MarsLab in the Space Sciences Building. Soderblom and others working with Professor Jim Bell at JPL receive raw images from the Mars rover and send them to the Ithaca campus for calibration and final processing. Robert Barker/University Photography

In the lab, computer screens line three walls. Four large wide-screen displays show the images currently being processed, ranging from frames that will make up large, colorful panoramas, to close-ups of rocks and soil. Other computers display data about the status of communication with Mars. A small video camera perched on a computer monitor and running Apple's iChat videoconferencing software maintains a constant two-way connection with Cornell associate professor of astronomy Jim Bell, leader of the rover panoramic camera, or Pancam, team at JPL.

A speakerphone plays the chatter from JPL's mission control. In the background can be heard the "wake-up" tune that NASA engineers play at the start of each Martian day. When the Spirit rover first unfolded its wheels, the engineers played "Get Up, Stand Up." When it rolled off the lander onto Martian soil, they played "Who Let the Dogs Out?" When the robot arm was deployed, it was "Reach Out I'll Be There."

Most of the people who work in the MarsLab (formally, the Cornell University Mars Data Analysis Facility), from staff members to graduate students to undergraduates, are volunteers. They are brought together by the fun of being part of something big. Though, of course, said graduate student Eldar Noe, "It's going to look great on any undergraduate's resumé."

In mid-January many of the students were working with Bell at JPL, and the burden of the work was being shouldered by Noe, research staff members Jonathan Joseph, Rich Chomko and Min Hubbard and Cornell undergraduate John Roma Skok.

Other members of the team include undergraduates Graham Anderson, Chase Million, Alex Shapero, Stephanie Gil, Benjamin Herbert, Dmitry Savransky and Kristen Frazier; research staff members Pam Smith, Mary Mulvanerton and Diane Bollen; and Emily Dean, an undergraduate at Ithaca College who first became involved with the rover project while still a student at Lansing High School.

Noe, Joseph and Chomko are living by the time on Mars, where a day, or "sol," lasts 24 hours, 39 minutes and 35 seconds, the time it takes the planet to rotate once on its axis. A Mars rover can operate only during hours of sunlight, when its solar panels can make electricity, so the MarsLab workday is about 12 hours and 20 minutes long and starts about 40 minutes later each Earth day. Noe has put cardboard over his bedroom windows and has adjusted to sleeping a bit longer every night. But, he said, "My cognitive abilities are way down."

The rover's two Pancams take black-and-white digital images but photograph the same scene several times through different colored filters. This gives scientists valuable information about the makeup of rocks and soil, but also makes it possible to combine three images taken through red, green and blue filters to make a color image.

This is where "calibration" becomes important. The color of the light falling on the Martian landscape can change from minute to minute, as the angle of the sun changes. Swirling dust in the thin atmosphere can change the lighting from day to day. Before the Pancams take each set of pictures, they swivel around and snap a series of pictures of a small calibration target (informally known as the Martian sundial) mounted on the rover's rear solar panels. MarsLab workers compare the known colors of the sundial with what the camera sees. Software written by Joseph and research support specialist Jascha Sohl-Dickstein uses this information to color-correct the Pancam images.

Sometimes an additional correction is necessary to adjust for the fact that the one of the Pancam's "red" filters is actually in the infrared, an image that is used for stereo imaging and which by itself is "more mineralogically interesting," the MarsLab operators say. True colors also are important as an aesthetic representation of what the vistas might look to a person standing on Mars, as well as an indication of the presence of minerals that might help determine if water was once present in the area.

The MarsLab team also examines each image for defects. Sometimes the data stream from Mars can be interrupted, and an image will be fragmented, calling for a retransmission.

Other software written by Joseph gives the team powerful tools to do the work quickly and precisely. Any part of an image can be clicked on to measure its brightness or to present a graphic display of the relative intensity of the three colors in one image. Joseph also has written one of three programs that can be used to stitch several small images together into a larger mosaic, like the panorama released shortly after Spirit landed.

The work schedule soon will intensify. The second rover, Opportunity, is scheduled to land on the opposite side of the planet Jan. 25 (at 12:05 a.m. Eastern time), so its day will correspond to Spirit's night, and from then on the MarsLab will, be open around the clock, with four six-hour shifts. Maybe JPL mission control should then play "Hard Day's Night.