ISIMIP3: Flux variables on 360_day calendar for monthly and annual data #31

Matthias Büchner opened this caveat on Dec. 9, 2022.

Subscribe

Discussion

Matthias Büchner reported on Dec. 9, 2022.
We became aware of a potential problem with using the 360_day calendar we requested for monthly and annual data. Most models resolve the days of the year internally, using the 365-day or proleptic_gregorian calendar, just like the input data, meaning they account for the length of the months according to a real calendar. This can be problematic when aggregating monthly given fluxes in s-1 or d-1 from the output data. NetCDF tools and APIs would then assume that each month has 30 days and thus miscalculate aggregations for months or years with varying number of days per month in the model simulation.

Before analyzing flux variables from ISIMIP3 already downloaded from the portal please lookup the calendar attribute of daily outputs from the same model and apply the number of days per month accordingly. Files available until today will be republished with adjusted daily time axes (CLASSIC biomes model and all data from the marine fisheries and ecosystems sector). A list of data sets affected can be seen below. Files project participants have collected from our DKRZ storage until today will be rewriten as well before being published.

We ask that groups that have data in their post-processing pipelines shortly before submission to DKRZ to adjust the time axes and units. The protocol and the QC-tool has already been adjusted accordingly. We will assist you if you need to delay submission for an extended period of time due to this issue. Please do not hesitate to contact isimip-data@pik-potsdam.de for quick assistance.
No comments have been submitted, yet. Login to leave a comment.

Details

Affected datasets can still be used for simulations or research.

Severity low

Status resolved

Versions
before 20221208
Data product
Output Data
Time step
annual, monthly
Simulation round
ISIMIP3a, ISIMIP3b

Affected datasets

There are 2253 datasets affected by this caveat. Here we only display the first 100 datasets. You can download a complete list of all the datasets, or files affected by this caveat as JSON. Alternatively, can also download a list the files as a flat file suitable for Wget. You can use the search interface to further restrict your query.