While addressing a sub-optimal performance issue for interpolation from the octahedral reduced Gaussian grid introduced with IFS cycle 41r2 to regular latitude-longitude grids, a problem was discovered with the method used to calculate the longitudinal points in the source grid. In some specific cases, this problem leads to an incorrect computation of interpolation weights and hence to incorrect interpolated values at some points of the output grid. The problem has been addressed in EMOSLIB versions 4.3.x. See Bug fix implemented in EMOSLIB 4.3.x for further details. |
This version of libemos was tested against BUFR tables version 000408, please check also changes of the previous version 000437
[EMOS-252] - fix Cray failures on creating and checking Legendre coefficients file size
[EMOS-260] - incorrect values for interpolation to regular_ll particular sub-areas crossing 0 degrees longitude (particular cases)
[EMOS-262] - incorrect values for interpolation to single points
This version of libemos was tested against BUFR tables version 000407, please check also changes of the previous version 000420
[EMOS-232] - include climate.v014 land-sea masks
[EMOS-220] - fixed interpolation of wave model parameters sub-areas straddling the Greenwich meridian
[EMOS-205] - smaller memory allocation for interpolations of u/v on reduced_gg grids
[DAPP-249] - drifting buoys with new BUFR template (subtype 182)
[EMOS-188] - HIRLAMW memory allocation issue under very specific conditions
[EMOS-216] - HRES corrections and performance improvement for West-most longitudes and high North/South latitudes, including LSM