There are three stages to validation: quick look validation, in-depth validation, and geometric validation. Most of the validation is done by undergraduate students in a variety of departments from the University of Arizona. If they notice something odd, they flag it and notify the full-time operations team, who do a more detailed analysis. Currently there are two students who work solely with validation, and me. I volunteer when their work load becomes heavy (normally I work as a programmer for different software needed at HiROC).

The first stage is to let the staff at HiROC know quickly if there is any problem in commanding, or if there is haze in the field of view. Low resolution “browse” images that have come down in the last 24 hours are examined by someone (usually either myself or RichardLeis) to see if there is anything obviously wrong. If there is suspicion, a flag is raised and then other people will take a look at the images. This quick look helps prevent any commanding issues from continuing, and also helps us avoid taking more pictures in areas with some kind of atmospheric distortions (”haze”). This could include dust storms, clouds, melting ice caps in the polar regions, etc.

The second stage of more in-depth analysis involves a tool called HiVali. This tool allows one to quickly take a look at an image in more detail, and see if there are any kinds of problems with it. It reports statistics of pixels to see if there is saturation or low contrast. It checks to see if there are any gaps in the image, and other kinds of routine image checks. One part of this process, the part that takes the most time, is the visual validation, where a human physically looks at every inch of the picture in high resolution to see if there is anything odd. These are usually things which can be fixed in the calibration processing.

The third stage is also a quick look, which is done after the image has been geometrically projected. If something is found that is strange in this stage, then the geometric processing is re-done to correct the error. If the image is in good condition, then many of the intermediary files are marked for deletion. Note that the EDRs and the final products are always kept, but there are several stages to the pipeline, as readers of this blog will know.

Sometime during this process, science team members also look at the image, to find if there is anything of special interest in the image. If there is, a caption is written, and it is prepared for the next batch of images to be released. If not, the image may be released without a special caption. All images will be released within 6 months of being taken. Once everything is perfected at HiROC, this release time will be reduced greatly, perhaps even to a few weeks. Currently, though, we are still working on the process of getting the images released.

The uplink team is constantly looking where to point the camera next. There is a program which is in beta testing now called HiWeb which allows scientists and other people to input suggestions. The Uplink team reviews the suggestions in the database, assigns a priority to each of these suggestions, and then finds when we can point the camera at the part. They also make sure a certain percentage of the upcoming pictures are assigned to look for a Phoenix landing spot, as this is a high priority item at the moment. They are still learning exactly how to best command the camera, and are constantly sharpening their skills.

The downlink team is making sure operations run smoothly at HiROC. They are verifying that the processing has taken place, make sure that the images have been calibrated correctly, that there are no image processing artifacts on the images we are about to release. If there is any artifacts created from processing the image, the source of the problem is identified and fixed, and then the image is reprocessed. While previously we have sent images to the public that had some small processing artifacts during the post-MOI and Transition imaging, we currently are waiting until the images have been completely validated. The downlink team is also taking a quick look at each image that comes down, and making sure there isn’t something unexpected, for example, haze at Mars, lots of saturated pixels, etc. If any such problems are found, they notify the uplink team, to ensure that we don’t have continuing problems. These problems are very rare, but on occasion happen, due to the changing nature of Mars.

During and after the validation process, the images are reviewed by several of the science team members of HiRISE. Things of special nature are noted, and these images receive captions. The others are slated for a more general release. Due to the large size of the HiRISE images, it is almost impossible to search every square inch of the pictures by any one or even a small group of people. I’m sure many of you have noticed this with just the images which have been released, there are many more which are still being validated which have yet to be released.

The Systems team is responsible for making sure that the HiROC computers are all working in top shape. They are quick to find problems when they arise and fix them so that it does not affect the flow of data here. They are preparing servers for two upcoming services that HiROC will provide, HiWeb, which was mentioned previously, and a JPIP server, which will allow for the effective distribution of JPEG 2000 images.

The software team is writing software that will make people’s lives easier. Some are working with the HiPlan suite of tools, which is used to plan upcoming images, to make it even easier to use for the uplink team. Some are working on HiVali, the validation software, which is used to make it easier to verify that an image is ready to release to the public, quickly finding problems with the image. Some are working on HiView, a program which will allow distribution of images over the JPIP protocol to the general public. Still others are working on getting HiWeb ready for public release.

Let me also talk a bit about a few upcoming products mentioned in this entry. HiView will allow you to download only the parts of a HiRISE image that you find most interesting. It will work great, even for those who have slow internet connections. I personally have tested this with a connection rate of 1kBytes/sec, and it works reasonably well even at that slow speed. It will allow the user to save the parts of the image they find the most interesting to their hard drive for future study (HiView will require a constant internet connection to download the image)

Another upcoming product is HiWeb. HiWeb will allow any user (Yes, that’s you!) to suggest future targets to image with the HiRISE camera. Preference is given to targets of scientific interest. The suggestions are given a priority, and placed in a database to be targeted depending on the orbit of MRO and the allocated bandwidth.

So, that’s what’s happening at HiROC these days. In short, we are all very busy, but very much enjoying our work. I personally can’t remember a time that I’ve had as much fun working as these last few months have been. And surely the best is yet to come!

That’s GuyMac on the left, HiCommander (me!) in the middle, and HiKu on the right.

HiKu is part of the operations staff on the uplink side of things. He’s on the team that does the targeting and the planning. What was he doing on a Saturday? Same thing he does every day: targeting and planning.

GuyMac and I are part of the software development team—we write the programs the ops team uses to do their job. GuyMac works primarily for the downlink group. He spent the day on a program called HiVali, which will be used by the downlink ops team to make sure a given HiRISE image accomplished its goal.

I work for the uplink group. I spent the day working on a program with the second greatest HiRISE software name: the HOGG. That’s the HiRISE Observation Generation GUI. The “the” is an important part of its name, by they way. You don’t use HOGG to generate HiRISE camera parameters. You use the HOGG.

A lot of our HiRISE tools have funny names; the most common way to name a piece of software here is to get a one-word description of its function and then add “Hi” to it. The planning software? HiPlan. The commanding software? HiCommand. The validation software? HiVali. The camera temperature modeler? HiTemp.

I like “the HOGG” for three reasons. One, who doesn’t like hearing their peers use such a ridiculous word to talk about something serious? Two, it breaks the HiRISE naming convention. And three? I get to correct people and point out that the “the” is part of the name. Still, it’s only my second favorite HiRISE software name. My favorite?

That’d be the HiRISE photometry predictor.

HIPHOP.

I’ll write more about these tools in the coming weeks.