Read on Twitter
HadCRUT5 is out. You can find the paper and data here.
https://www.metoffice.gov.uk/hadobs/hadcrut5/
Sevenish reasons why you - a discerning scientist - might want to use HadCRUT5. THREAD
https://www.metoffice.gov.uk/hadobs/hadcrut5/
Sevenish reasons why you - a discerning scientist - might want to use HadCRUT5. THREAD
1: HadCRUT5 is based on updated datasets of land surface air temperature (CRUTEM5 @TimOsbornClim ) and sea surface temperature ICOADS release 3 https://rmets.onlinelibrary.wiley.com/doi/full/10.1002/joc.4775 supplemented by drifter data from @CMEMS_EU
2: HadCRUT5 includes up-to-date adjustments for changes in how sea-surface temperatures were measured. Including corrections for ship-buoy differences and shifting biases in ship-based measurements. This was published in 2019 as the HadSST4 data set: https://www.metoffice.gov.uk/hadobs/hadsst4
3: HadCRUT5 uses a Gaussian-process based statistical method to provide an improved analysis and interpolate gaps in the data set including the infamous "hole at the pole". The statistical analysis takes into account the estimated uncertainties in the underlying data.
4: We don't interpolate into areas where the data provide only a weak constraint on the statistical analysis. There are areas of the world, particularly early on, which we don't reconstruct. Global and hemispheric time series include an uncertainty term to account for this.
5: If you *like* your dataset with all the gaps in, we still love you: you can get all the wonderful gappy data too.
6: Uncertainties in HadCRUT5 are presented by an easy-to-use ensemble. A set of samples are drawn from the posterior distribution of the stat model so it integrates uncertainty from all sources: measurement error, gridbox sampling, bias uncertainty and limited coverage.
7: HadCRUT5 is provided under an Open Government Licence (so you know where you stand when using the data) and the final journal articles will be open access. Till then, read the papers anytime on our website https://www.metoffice.gov.uk/hadobs/hadcrut5/
https://www.metoffice.gov.uk/hadobs/hadsst4/
https://www.metoffice.gov.uk/hadobs/crutem5/
https://www.metoffice.gov.uk/hadobs/hadsst4/
https://www.metoffice.gov.uk/hadobs/crutem5/
We've been listening to the feedback we received on earlier versions of the data set and aimed to make a more comprehensive, better and easier to use data set.
While HadCRUT5 contains lots of nice things, it's best as part of a balanced diet. Using a range of different global temperature data sets - NOAAGlobalTemp, GISTEMP, Berkeley Earth, one of the reanalyses maybe - will give a broader view on structural uncertainty.
And with that....
