I had lots of free time today, which I spent going through dozens of dense NASA reports about the Shuttle TPS, so forgive me if I'm a little vague on which factoid came from where...

It's my understanding that Columbia, as the first and testbed vehicle, started out with an extensive network of sensors, including thermal sensors embedded all over the exterior. A lot of the sensors were later removed to save weight. I came across a report at Dryden (THERMAL RESPONSE OF SPACE SHUTTLE WING DURING REENTRY HEATING), that talked about tests at Edwards using an F-104 equipped with thermal insulation to calibrate an extensive set of TPS temperature readings from STS-5...Columbia. I'd guess that Columbia as originally outfitted could have given the kind of resolution you're talking about.

I'm starting to think that Columbia was different from other shuttles in other ways. In that 1990 paper on launch debris hits ("Safety of the Thermal Protection System of the Space Shuttle Orbiter: Quantitative Analysis and Organizational Factors"), I found a table with hit statistics for the first 33 Shuttle flights. I punched the table into a spreadsheet and crunched the numbers. It was interesting: Where Challenger, Discovery, and Atlantis, had an average of 143-173 hits per launch, Columbia had an average of about 231.

Trying to find a correlation is like the tail wagging the dog, because the ETs were different each mission--why would they all shed more on Columbia? Well, I'm into pure speculation now, but Columbia was heavier, and so maybe it gave the ETs a rougher ride? Did all the additional work done to its tiles (including sensor removal) give it rougher flow characteristics, and did the rework contribute to the final failure? Did astronauts who flew more than one Shuttle describe Columbia as a rougher ride? So many questions!

Regardless of cause, it looks like (at least in its first decade) that Columbia got a lot rougher treatment at launch than other Shuttles.

--Robert