Known Bugs and Issues

Release v7.1

CDS API and ERA5 data

See ticket #230 on for information.

Installation problems with GATEWAY and DESTINATION parameters

See ticket #263 on for information. Fixed in v7.1.1.

Installation problems with the Fortran program calc_etadot

See ticket #264 on for information. Fixed in v7.1.1.

ECCODES Error Code 250

If you get an error message from ECCODES with code 250, looking like this:

ECCODES ERROR   :   wrong size (184) for pv it contains 276 values
ECCODES ERROR   :  get: pv Passed array is too small
... ERROR CODE: 250
      Command '['<path-to-flex_extract>/flex_extract_v7.1/Source/Fortran/calc_etadot']' returned non-zero exit status 250.

then you have set a wrong maximum level in the CONTROL file! It is important to properly select the maximum level depending on the data set you would like to retrieve. Only the following values of the number of model levels available are valid: 16, 19, 31, 40, 50, 60, 62, 91, 137.

The ERA-Interim data set uses 62 model levels and ERA5 as well as CERA-20C uses 137.

The operational data sets use different numbers, depending on the date. For example, from 25/06/2013 on, 137 model levels were used in the operational system. Therefore, every time you extract data from a date later than 25/06/2013 you have to select LEVEL=137 in the CONTROL file, or you have to define LEVELLIST=1/to/137. Of course, you can stop before the top of the atmosphere, such as 60/to/137, but you have to include the maximum level number (i. e., the lowest level, see note below!) number as the last one. Table 2 in the scientific model description paper gives an overview of the level lists and the corresponding date they were introduced.


Be aware that the ECMWF model counts the levels from the top of the atmosphere downward to the surface. Level 1 is the topmost level, and e.g. 137 would be the level closest to the ground.