I released the first files for the climate data service today, a day ahead. I’ve already gotten some feedback, so I’ll answer it here.
This is beyond awesome! I’ve kept an up to date version of a handful of these series but nothing like this. I’m teaching time series right now and can’t wait to unleash my students on these data. Quick question: in fields2.csv you note that all anomalies use the baseline period Jan.1900 – Dec.1999. I assume that is true for both files ‹ the global data as well as the US data. True?Wish list:
For the temp data sets I use Cowtan and Way sometimes and Berkeley too. You might could add those ‹ but how many temp data sets do you really need?
And for ice I use the PIOMAS volume data in addition to the NCIDC data.
TSI in addition to sunspots.
An aerosol time series?It would be nice to have a link to the source data in the fields files too I suppose.
What a resource.
Thanks a zillion
The baseline period Jan.1900 – Dec.1999 applies to US state-by-state data only. Other fields are either already anomaly (global temp series) or are baselines chosen logically. I’ll update the documentation to provide that information.
I plan to add both Cowtan & Way, and Berekely Earth, soon. I too wonder “how many global temp time series do you need?” but I sometimes want to use them all. It presently includes the big five (GISS, NOAA, HadCRU, RSS, UAH) but both C&W and Berkeley are prominent enough to include.
I suspect that PIOMAS sea ice volume data has some problems. But I might as well include it — it’s up to users what to do with it.
I thought about TSI (in addition to sunspots) but I don’t know where to get it that’s *up-to-date*. That’s why I use sunspot counts for regression analysis — it’s easy to keep current. I’ll consider adding TSI anyway. Patience, …
Aerosol time series is also tricky. For regression I’ve just infilled recent values with zero because data aren’t very current. I’ll add it to the “coming-but-not-so-soon-list.”
Thanks.
Open Mind 
Thanks for the data Tamino! It is very comprehensive and the format is clear & useful. A great service!
Just a couple of suggestions for the future (since you asked)…
1) Where available, it may useful to include (maybe as separate file) the uncertainties associated with mean metrics. For example I think CRUT provides various upper and lower bounds for their global anomalies?
2) Rather than sending monthly zip files with all data subscribers might possibly want, it may be easier – for you and subscribers – to maintain a website where subscribers can download the data they want, when they want it. Tools like R Shiny make this sort of thing pretty simple to make.
You can find mostly up-to-date (though mid-Feb) PMOD TSI data here: ftp://ftp.pmodwrc.ch/pub/data/irradiance/composite/DataPlots/composite_42_65_1602.dat
A thought: would it be worth your while to offer a ‘one shot’ deal of some sort? I wouldn’t use a regular subscription enough to justify it, but could imagine using a ‘snapshot’ for a specific writing project. Don’t know if that makes business sense from your side, though.
[Response: Hmm… let me think about that.]
What problems, if I may ask?
GHGs would be a great addition too. Monthly time series of CO2 from mauna loa is easy. Other gases take some more work. But given the interest in CH4 these days that would be a good addition too. Thanks!