date: Wed, 3 Oct 2007 15:18:49 -0400 from: "Kate Willett" subject: Re: Scheduling your Nature paper to: "Phil Jones" Thanks Phil, do I need to do anything over here or just be aware that someone might contact me? I've just come accross something interesting in my data - it looks like the land T and dewpoint T data is recorded only to whole numbers prior to 1982 too. I've only looked at a few stations so far but these have been in completely different countries but show the same thing. I don't think/hope this affects any results so far too much. I remember noticing this to some extent in the early stages of dataset building but didn't notice a pattern. I've only noticed it just now because I'm looking at max and min supremums of pentads. In HadCRUH there is no apparent shift comparable to that in the Marine data at least. Did you notice anything like this with your hourly temperature data and if so what did you do about it? I've coming back to the UK for the Met Office interview at the beginning of November so will be around Norfolk for a bit too seeing friends/famliy. I'll pop into CRU at somepoint and say Hello. Kate