Saturday, April 02, 2005

"Sound Science" and Yucca Mountain

Some highlights from e-mails written by federal employees researching the suitability of Yucca Mountain as a repository for nuclear waste:

  • "Science by peer pressure is dangerous but sometime it is necessary."
  • "The QA bullshit grows deeper. I may need to say that I did everything by hand for the data package I am submitting that you and [redacted] reviewed. The program I wrote is not in the system and QA will be all over it like flies on &%#$. All references to [redacted] are being deleted. Here’s my question: When we go to start QA’ing the site-scale modeling work, will I get taken to the cleaners because I am not referencing either a tech procedure or a scientific notebook?"
  • [In response to the above:] "What if you just download the raw files from [redacted] and say you used those? Do they need to know any more than that? You don’t really need to do an analysis just say this is the data I used. Maybe that would work."
  • [Replying:] "Not a bad idea. I am now considering it. Ideally, one would assume that the more information you proved QA, the better the QA. In reality, it seems that the opposite is true. At any rate, it’s a damn shame to be wasting time with this sort of thing."
  • “Model simulations have been in progress but about 3 weeks ago I found a small error in the model input that was generated using the [redacted] data. The error was minor but would have created a QA nightmare so this was fixed and the simulations are being re-done (I’ll send you a summary of the results when I get to this point). The input files are basically re-formatted [redacted] export files with a minor amount of parameter estimation occurring to fill small gaps in the record (even for the high ranking sites, there are gaps all over the place). Here’s the weird news; to get this milestone through QA, I must state that I have arbitrarily selected the analog sites. So for the record, seven analog sites have been arbitrarily (randomly) selected. Hopefully these sites will by coincidence match the sites you have identified. P.S. please destroy this memo."
  • "Dealing with this QA bullshit is really starting to make me sick."
  • "Don’t look at the last 4 lines. Those lines are a mystery that I believe somehow relate to the work [redacted] was doing in entering the 1994 data. These lines are not used by [redacted] (we stop at 9/30/94). I’ve deleted the lines from the 'official' QA version of the files (which do have headers). In the end I keep track of 2 sets of files, the ones that will keep QA happy and the ones that were actually used."
  • "There is of course, no scientific notebook for this work. All work is in the form of electronic files…. They may be expecting to see something that at least looks like a scientific notebook documenting work in progress. I can start making something up but then the [redacted] projects will need to go on hold."
  • "The programs, of course, are all already installed otherwise the [redacted] would not exist. I don’t have a clue when these programs were installed. So I’ve made up the dates and names (see red edits below). This is as good as its going to get. If they need more proof, I will be happy to make up more stuff, as long as its not a video recording of the software being installed."
Matthew Wald has the story.


Post a Comment

<< Home

Fed. R. Evid. 702: If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case.