FIXED: more problems found in ISIMIP2a/3a/3b climate input data

Posted by Martin Park on June 29, 2020

Dear ISIMIP community,

This e-mail is to inform you about the progress we made on two kinds of problems recently found in our climate input data.

  1. The problems described in the e-mail sent to all ISIMIP modelers on June 16 have now mostly been fixed by re-doing the bias adjustment of GCM data for ISIMIP3b and the harmonisation of GSWP3 to EWEMBI and W5E5 for ISIMIP2a/3a using ISIMIP3BASD v2.4.1. The update affects all variables except ps, rlds, and tas. Files at PIK, DKRZ, and ESGF have already been replaced for
  • GSWP3-W5E5 obsclim
  • the primary ISIMIP3b GCMs

Not yet done are

  • GSWP3-W5E5 counterclim and spinclim (expected to be done by the end of next week)
  • the secondary ISIMIP3b GCMs (expected to be done by the end of this week)

For more details about this data update see the caveat descriptions for the 3a datasets and the 3b datasets.

  1. Another issue was raised by Sebastian Ostberg (PIK) and has now been fixed too: There used to be negative hurs and huss values in the GSWP3 obsclim dataset. These were now replaced by zeros. The negative values also resulted in missing hurs and huss values in the counterclim and spinclim versions of GSWP3. These were now replaced by zeros too.

Best wishes,
the ISIMIP team at PIK