Up to Specific issues in real coupled models
Hi, Could you confirm the required position of gradient terms in the argument list for calls to OASIS_PUT with OASIS3-MCT vn4.0? The user guide implies that for 2nd order conservative regridding the 2nd term should be the gradient with respect to longitude and the 3rd term should be the gradient with respect to latitude. (similarly for bicubic, the 4th term should be w.r.t. longitude and the 5th should be w.r.t. latitude). e.g. section 4.3 of the user guide says "..the second and third terms...these terms, respectively the gradient of the field with respect to the longitude..and the gradient of the field with respect to the latitude.. This seems to contradict the contents of the SCRIP routine remap_conserv.F90 which says: "..the second and fourth are second-order gradient weights." and "..the third and fifth weights are for the second-order phi gradient component..." I appreciate the weights field indices produced by SCRIP may not necessarily match the order in which gradients are applied via the OASIS_PUT (but I'd be slightly surprised if they don't). Thanks,Richard
Hi, After verification in the SCRIP library, for Second Order remapping, also for LR grid, you will have 3 weights stored in the remapping file calculated by OASIS, that will apply to each target point. The first will be apply to the field (fld1 in oasis_put), the second will be applied to the gradient of the field with respect to the latitude (fld2 in oasis_put) and the third will be applied to the gradient of the field with respect to the longitude (fld3 in oasis_put). And again fld1, fld2 and fld3 are calculated in the model. The User Guide, https://oasis.cerfacs.fr/wp-content/uploads/sites/114/2021/02/GLOBC_TR_oasis3mct_UserGuide_4.0.pdf section 2.2.7, has been updated. Best regards, Laure
