It's internal e-mail traffic concerning a discontinuity in the US temperature data. In the first of
four documents, NASA assumes that they're chasing a bug in their software. I'm still working through the other three: our government prints out e-mails, then scans the print-outs as PDFs ... poorly. Sheesh.
According to
Christopher Horner, the later documents include e-mails detailing significant inconsistencies in the input data. I haven't gotten far enough to put them into context yet.