 |  |  |
| Aerial photographs of Antarctic dry valleys illustrate the "nesting" technique to be used by the Mars Descent Imager. In the first photo, the locations of the series of nested descent images is shown, each a factor of 2:1 smaller than the previous image. The second photo shows the largest nested image, which is 500 meters across and has a resolution of 5 meters per pixel. The third photo shows the final nested image, which is 160 meters across and has a resolution of 16 centimeters per pixel. The boulder, top right, in this image measures about 6 meters across. PHOTO CREDIT: Malin Space Science Systems | ||
The camera, known as the Mars Descent Imager, or MARDI, will be positioned between the legs of the Mars Polar Lander, with the exhaust of the hydrazine engines in view. It will begin clicking its shutter after the lander vehicle's heat shield has been jettisoned -- about 6.5 kilometers (about 4 miles) above the surface -- and while the craft is still swinging on its parachute harness. The last few images -- perhaps eight -- will be captured after the parachute has been jettisoned at about the 1 kilometer (.62 mile) altitude and as the craft makes a controlled descent, slowed by retro rockets, to the frigid northern edge of the Martian south pole's layered terrain.
"MARDI's images will make all of us much more comfortable in making interpretations of the lander's pictures because they will give us a context," says Peter Thomas, a senior researcher with Cornell University's astronomy department. "For the first time we will have a complete scale of pictures of Mars, from less than a millimeter all the way up to orbiter pictures." The camera has a 70-degree field of view, and the estimated difference in resolution between the first and the last black-and-white images will be a factor of about 800.
Thomas is one of three Cornell researchers on the MARDI team, led by Michael Malin, president of Malin Space Science Systems, San Diego. Also participating in the development of the imaging system, and present at NASA's Jet Propulsion Laboratory (JPL) in Pasadena, Calif., to interpret the images after they are received from the Mars Polar Lander, are Cornell astronomy professors Joseph Veverka, who also is chair of the Cornell astronomy department, and Steven Squyres. Also on the team are M.A. Caplinger of Malin Space Science and M.H. Carr of the U.S. Geological Survey in Flagstaff, Ariz. MARDI was developed under a $3.5 million JPL contract.
At present, the highest-resolution images of the Martian surface, taken from orbit, are made up of pixels (or picture elements) each covering 1 1/2 yards of terrain. That is about to change dramatically to images with each pixel covering a fraction of an inch of the surface.
The descent camera pictures will be used to interpret ground features and will aid in the mission's main purpose, studying the layers of ice and dust covering the polar region. These images will be captured with a "nesting" technique, meaning that each successive image will be nested within the previous picture. As the spacecraft loses altitude, each successive image will cover a smaller area within the previous larger image. The camera has no ability to aim, but simply points where the spacecraft points. "The first image will be several kilometers on one side, but the camera has a fairly wide angle so that even with the spacecraft swinging on a parachute, the images should remain nested within one another," says Thomas.
The nesting technique, he notes, will enable researchers to find a ground feature, such as a boulder, in the image taken closest to the ground, then work back to the largest picture. The spacecraft's electronic memory retains each image, plus details of when the image was taken, which direction the spacecraft was pointing at the time and its altitude. In this way, says Thomas, "you can take pictures and reconstruct them from that geometry."
The number of images returned to JPL will be limited by the storage capacity of the spacecraft's memory. For this reason, the on-board computer has been programmed both to reject some images taken by the camera and to write over others. The computer will be instructing the camera to capture images in different image formats (in terms of pixels) based both on altitude and the number of images already taken. If the computer determines that the altitude has not changed sufficiently, it will not save the image.
"If the memory's storage is full and the camera is still taking images, the computer is programmed to throw out some lower-resolution pictures," says Thomas. "We want to maintain nesting and protect the higher resolution images as we get really close to the surface. Those images closer to the surface are of platinum value."
The "overwhelming purpose" of the descent camera's images, says Thomas, is to tie what will be seen with the lander's camera on the surface of Mars with images taken a few feet from the surface. "We've seen the whole of Mars in 100-meter resolution, but only 1 percent of the surface in three-meter resolution. These images will be filling the gap."