That corrects data by changing the way it's collected. That's OK if it's properly documented of course, because you're improving the accuracy of your measuremen

That corrects data by changing the way it's collected. That's OK if it's properly documented of course, because you're improving the accuracy of your measurements. You may then have difficulties comparing the two periods but there are techniques to deal with that. Where climatologists go wrong is when they say "and now we'll rewrite the prior observations to what we think they would have been, had the weather station been there all along". That's not scientific. Consider how easy it is to make mistakes that way, or have biases in which only adjustments beneficial for your career are included. Especially when your central claim is about long term trends over time, it should be obvious why editing the past isn't allowed.