CRU TS + CRU CY BLOGIn date order, newest at top; manually edited
20 April 2023 CRU TS & CRU CY v4.07 released
20 April 2023 CRU CL CLD climatology reverted to v1.0, see Read Me for details
29 June 2022 v4.06.01 CRU CY CLD published, basd on the CRU TS point release (below)
28 June 2022 v4.06.01 CRU TS CLD published, see Read Me for details
16 June 2022 v4.06 CLD withdrawn
30 May 2022 Revised climatology, CRU CL v1.1, released
26 May 2022 CRU TS & CRU CY v4.06 released
11 January 2022 Advance Warning: revised climatologies
17 March 2021 CRU TS & CRU CY v4.05 released
6 June 2020 Obligatory Welcome to the Blog post
6 June 2020 Bug report: affecting multiple variables, minor
3 June 2020 Bug report: affecting VAP, minor
16 June 2022: Withdrawal of CRU TS v4.06 CLDShortly after release, we received reports that the CLD variable was behaving erratically (and unrealistically) in South America.
After investigation, we have withdrawn v4.06 CLD temporarily; a new, improved 'point release' of CLD (v4.06.01) will be released in the near future.
We apologise for the inconvenience.
11 January 2022: Advance Warning: revised climatologiesIn September 2021 we were contacted by Bruno Lemke (NMIT, NZ). Together with his colleagues Matthias Otto (NMIT) and Tord Kjellstrom (HEIT, NZ), he had identified anomalies in the CRU CL v1.0 0.5° climatologies. These fields underly CRU TS, as they are used to convert gridded anomalies back into actuals for publication.
Upon investigation, we found a number of 'unrealistic' areas. There was no obvious pattern to these areas, occurring in different parts of the world and on different months, for several variables. A forthcoming technical note will detail this work in full.
We elected to regenerate the half-degree climatologies from the ten-minute climatologies (CRU CL v2) published a few years later, arriving at new half-degree climatologies that eliminate the issues found. This will affect all CRU TS variables to a greater or lesser extent. It will not alter changes over time, such as trends, because the climatology is applied equally to every January, every Febrary, and so on.
The next version of CRU TS will be v4.06; it was not felt necessary to increment the primary version number for what is essentially a 'bug fix'. However, in some parts of the world there will be noticeable differences from v4.05; as usual, comparison plots will be available.
We thank Bruno Lemke and colleagues for their diligence in locating these anomalies.
6 June 2020: Obligatory Welcome to the Blog postNeedless to say, I've been meaning to run a CRU TS + CRU CY blog for many years. It's been frustrating, not having a route to communicate developments of different kinds (yes, it's a bit of a shame that two of the first three are bug reports; but then, that's what finally drove me to set this up). Apart from bug reports and paper announcments, there will be posts when I want to explain enhancements I'm planning, problems we've hit with the process, or anything else really.
6 June 2020: Bug Report: affecting mutiple variables, minorAn exercise to plot anomaly counts for various variables revealed a cluster of stations near 100°W, at various latitudes. Investigations revealed that the new-format MCDW data includes stations with lat and lon missing; this was unexpected (as previous MCDW updates always had location information) and the exception was handled badly at the two points where it could have been caught. Comparisons between v4.03, v4.04, and a product with that bug corrected, (v4.04.02, unreleased) are available as PDFs: TMP DTR PRE VAP WET CLD TMN TMX VAP WET Note that many of the changes in the corrected version (green) simply return the series to the 4.03 values (thin black). Overall the differences are not considered major enough to warrant a point release. The bug fix will carry forward to the next release of CRU TS. However, if any users would like to make use of any variables from the corrected run (4.04.02, unreleased) then please get in touch. This version includes the VAP correction (3 June 2020).
3 June 2020: Bug Report: affecting VAP, minorAs part of the upgrading of the Google Earth interface to CRU TS, it was discovered that synthetic VAP stations (used to 'top up' when actual VAP observation counts are insufficient) were being allocated on a 'first come, first served' basis, when the nearest should have been selected. All stations used were within the CDD, so the results could not be said to be 'wrong' - but they could have been better. A comparison between 4.04 VAP and the 'corrected' VAP (v4.01.01, unreleased) is here (PDF); the differences are not considered major enough to warrant a point release. The bug fix will carry forward to the next release of CRU TS.
3 April 2020: NEW CRU TS v4 Paper: Announcement