The data rate from the Mars Reconnaissance Orbiter (MRO) – in orbit around Mars on the other side of the Sun from our perspective here on the Earth – is set to increase soon, marking the end of solar conjunction and the beginning of MRO’s Extended Science Phase (ESP). HiRISE will capture the first image of our next phase in Mars exploration this weekend. Like all first images after the HiRISE camera has been idle for awhile, we will look especially closely for, but do not expect, any issues.
In these days leading up to the beginning of ESP the downlink team is checking our processing software and tools to make sure they are ready for the observation ID prefix change from PSP_XXXXXX_XXXX to ESP_XXXXXX_XXXX. The automated processing pipelines are ready to go. My own validation and reporting Perl scripts make use of modules that are mission phase aware and pattern match file extensions instead of observation ID prefixes, so I think I am set. Any minor tools we miss can be easily updated as necessary.
With improvements to our tools and new procedures, we can sometimes recover image data previously stuck in the original raw data files. During this quiet period, I had an opportunity to go back to old observations, some from early in the Primary Science Phase (PSP), and recover errant channels that failed our processing software at the time. This resulted in a few new channels of image data that we will include in a later reprocessing of our images. For example, PSP_001746_1515 was originally processed without the RED1_0 channel. You can see this channel gap if you click on the observation’s “Full image (grayscale, non-map projected)” link. An improved mosaic will include most of RED1_0, albeit with a small image data gap near one end. Why did we not create this channel before? Sometimes a data gap occurs between channels, obliterating the second channel’s science header. The software at Jet Propulsion Laboratory’s RSDS that splits our image data into separate raw data channel files cannot figure out where to split the image data correctly within this gap and so image data from two or more channels are trapped in one raw data file. When this jumbo raw data file hits our EDRgen processing pipeline here in Tucson, Arizona, the tool called HiRISE_Observation became confused when it discovered image data from another channel near the end of the file, and the error it flags halts the EDRgen processing pipeline.
We now have improved tools and procedures for dealing with this problem. By running HiRISE_Observation outside of the EDRgen pipeline, it will successfully produce an EDR for the first channel before it complains about the second channel’s image data. While we would love to get at that second (or third, or later!) channel’s image data, right now we recoverthe data that is easiest to reach. Someday, we might have a tool that will recognize a missing science channel header, reconstruct this header, and then fit the header in correctly between the first channel and the additional channels stuck in one raw data file.
When will you see the new PSP_001746_1515 products, newly improved with the recovered RED1_0 channel? Hopefully in a few months. While I have recovered this EDR (as well as EDRs from other observations), we would like to wait to reprocess the entire observation until we have even better image calibration built into our processing pipelines. It makes little sense for me to reprocess this observation now when we are just going to be reprocessing it again sometime early next year. Once all of our improvements are in place in the upgraded processing pipelines, we will reprocess ALL of our observations, a huge undertaking.
Next week we will be back to our normal downlink activities: making sure the processing pipelines behave, validating new observations, providing daily reports to the rest of the team, and processing color and RDR mosaics. We will also add in routine creation and validation of anaglyphs and preparation for our next Planetary Data System (PDS) release in March. The quiet time during solar conjuction was nice, but we are anxious to see new images from Mars!
haha, just kidding. We’re actually trying to simplify things. Even though we’ve gotten pretty good at this, it’s still an incredibly complex process. We would like to continue doing it for a long time and still remain sane. Also, we expect that our funding will be cut in the future, and we want to make sure we can still do everything even if we have fewer people. Plus, things just wouldn’t feel right if we weren’t constantly changing something! 
