A River Runs Through It
Managing Chandra's Data Flow

Picture of The Data Systems Group
The Data Systems Group ponders a problem in the data stream caused by a solar flare. Left to right: Dave Plummer (back), Eric Schlegel, Lisa Paton, Ian Evans & Dale Graessle.
September 26, 2000 :: Chandra's images are fascinating, informative, and beautiful, but they arrive at the Chandra X-ray Center as a stream of 0's and 1's that only a computer jock could love. It takes at lot of technical know-how and TLC to convert them into something that a scientist can use, and normal people (not to imply that scientists aren't also normal people) can admire. This difficult job falls on the Data Systems group under the direction of Giuseppina (Pepi) Fabbiano, with major help from other groups, especially the Science Operations Team.

"One of our biggest challenges is to get the data in the building," said Joy Nichols, lead scientist of data systems operations, speaking from her office at the Chandra Operations Center.

DSN
Deep Space Network
The challenges start soon after it leaves Chandra and begins its 40,000 mile (more or less, depending on Chandra's position in orbit) journey through space to one of NASA's Deep Space Network antennas in Australia, Spain, or California. From here the data are sent electronically to JPL, where it goes into a pool with all the other data collected by the Deep Space Network.

"The data go AWOL at this point from time to time," Nichols said. "We hardly ever lose data, but it can take 2 hours or 2 days to show up."

Once it arrives at the Operations center, the river of data, which includes spacecraft environmental information as well as science data, is separated into various streams for examination by experts on the instruments, the aspect camera, and other parts of the spacecraft.

This so-called Quick Look data are given to the Science Operations team and the team that built the instrument for evaluation of the telescope performance and acquisition of the target.

Picture of Group
Arnold Rots (left) discusses Chandra archive issues. To his right are Dong-Woo-Kim & Janet DePonte.
Meanwhile, Data Systems scientists and computer analysts begin crunching the data in either the Automatic or Custom processing modes. This involves figuring out exactly where the telescope was pointing at each moment of the observation by tracking fiducial lights on the spacecraft and the positions of well-known stars. The spacecraft motion is then removed using a computer program. Finally, features peculiar to each instrument are taken into account.

The data are then distributed to the scientist who proposed the observation, and placed in the Chandra data archive. The archive, which is the Fort Knox for Chandra's golden data, already contains "well over a million files" according to Arnold Rots, the scientist in charge of the Chandra archive. (These are multiple files of the approximately 500 targets that Chandra has observed.) At present the archive has a storage capacity of 450 gigabytes of compressed data.

"The biggest challenge is to keep it all together," said Rots. "There are so many little operations, so many programs written by so many people. It is a big challenge to make sure that it all comes together."

Ian Evans, the End-to-End scientist for Data Systems, agreed. "If you don't plan very carefully, a change in one part of the system can affect another part of the system downstream, or it could back up the flow and affect the flow upstream somewhere. Our responsibility is to see that it flows consistently and makes sense."

Picture of Group
Janet DePonte (foreground) suggests a software solution. Left to right: Ian Evans, Dave Plummer, Pepi Fabbiano, Arnold Rots, Dong-Woo Kim.
"Accomplishing this requires constant vigilance in normal times," said Fabbiano. "And with any space observatory, things are often not normal."

There can be errors in the data due to faults in the transmission to Earth, the data can get lost temporarily, there are the inevitable bugs in the myriad computer programs, and there are targets of opportunity, in a which a nova or supernova, or a comet flares up or appears suddenly. These targets require changing the observing plan immediately (See SN1999EM and SS Cygni articles), and in some cases even tracking a moving target as in the case of Comet 1999 S4 LINEAR.

"Hectic, but satisfying," said Rots.

"Often the scientists need to see the data immediately so they can plan their next observation," said Nichols, "so we have to sit on the data processing all night, and do it fast and do it right."

"We're always Very Happy to see that it all works," said Janet DePonte, head of the data systems software group.

Solar Flare Diagram
A magnetic cloud of plasma from a solar flare can affect Earth and damage satellites such as Chandra. The white lines represent the solar wind; the purple line is the bow shock produced by the interaction of the solar wind with the Earth's protective magnetosphere (blue lines). (Image not to scale.) Credit: NASA/SOHO
Recent Headlines
There is also the weather -- space weather -- to worry about. Solar flares are unpredictable and dangerous to Chandra's sensitive instruments, especially ones that send a blizzard of high-energy particles Chandra's way. The observatory must be shut down until the solar storm is over, which means that the whole process of sending instructions to and receiving data from Chandra has to be interrupted. This can cause headaches, as it did a few weeks ago. The following excerpts from the operator shift report of Jennifer Lauer give a flavor of what this process is like.

"Turns out that, after our little radiation incident, they didn't change the obsid! (The Observation Identification used to specify each set of commands.) They just continued...(after performing several other operations to shut down the observatory)...Still the same obsid. Oh, lord, what a mess. Catting pieces together now (even though we still don't have all of it) to see if I can figure out when the good bit starts. Hmmm..."

Then:

"Word from the GOT that there are DSN issues. Goldstone had difficulties (apparently its equipment is at the Olympics, too) that they didn't tell us about ahead of time, and the data for the second overnight pass (the one that ended at about 7:30 AM today) is just trickling in to JPL. Woohoo! Joy just reported that the Goldstone data has been found; it was in CA. I hope it had a good time...."

The long and complex river of data running from the Chandra X-ray Observatory to the scientist requires many teams of skilled and dedicated people who carefully lay plans for a smooth trip, but who are also ready, willing and able to ride out the rapids.


Return to Chandra Chronicles

Highlights and Feature Stories from the Chandra Project
Archived reports of past Chandra Chronicles
Receive the Chandra Digest
Updated: March 27, 2008