[southern_cross]

At the time I went back later to look at the data again, almost none of the various links on the relevant web pages still worked. Some working links claimed to link to both raw and adjusted data but then didn't actually have the raw data. Some still had things like precipitation data but the temperature data itself was now missing. Some claimed to have "raw" data but it wasn't the data I was actually looking for and had accessed in the past. And so on. I know that I saw at least one statement which seemed to indicate that the data I was looking for was no longer available for legal reasons.

"Legal reasons" was the cover they used to not make this data available to begin with, but some authority or another finally forced their hand here. It wouldn't surprise me to find that they still don't want it out there, and so it may only be available on an on-again, off-again basis, maybe only after somebody complains about not being able to get it. Nor would I be surprised to find that the data itself mysteriously changes over time, as many have claimed. I believe that they have even admitted to such - that they routinely "correct" various data sets, meaning that what you see today may be different from what you see tomorrow.

In any case, I didn't expend a lot of effort the second time around, and the first time around I was mostly interested in what the "official", final data sets were claiming about temperatures (at least at first), because that was what was being put out to the public. And after finding issues with local temps, I ultimately went out and got local data from local sources, completely bypassing the official sources. And one of the first things I discovered was that there was an awful lot of data - both old and new - that was ultimately marked as "missing" in the official records but was sitting there plain as day in the local records. I never did quite figure out what was going on there either, but I have my suspicions.

The whole process just left me with a bad taste in my mouth, and so I don't generally trust anything other than the original station data from the original sources. The HadCRUT4 stuff may be a bit better; I haven't really looked into it but others have and they say it can't really be trusted either.