- Jul 28, 2022
-
-
Renaud Person authored
-
Renaud Person authored
-
Renaud Person authored
-
Renaud Person authored
-
-
rebasing with 68-summer-body-2022
-
-
-
-
- Jul 26, 2022
-
-
Sebastien Masson authored
-
-
-
Sebastien Masson authored
-
Sebastien Masson authored
-
- Jul 18, 2022
-
-
- Jun 27, 2022
-
-
- Jun 17, 2022
-
-
- Jun 13, 2022
-
-
- May 16, 2022
-
-
- May 13, 2022
-
-
Sibylle Techene authored
-
Guillaume Samson authored
-
- May 12, 2022
-
-
Guillaume Samson authored
-
Guillaume Samson authored
-
- Apr 22, 2022
-
-
Sebastien Masson authored
-
- Apr 13, 2022
-
-
sibylle authored
-
- Apr 12, 2022
-
-
sibylle authored
-
- Apr 11, 2022
-
-
Sibylle TECHENE authored
-
- Apr 08, 2022
-
-
Jérôme Chanut authored
Allows restarting a MLF run with RK3 and vice versa (provided Euler first step)
-
Jérôme Chanut authored
- Suppress the use ub2_b/vb2_b un_bf/vn_bf with AGRIF and RK3. Consider un_adv/vn_adv in that case, i.e. the available time integrated flux from before to after time step. - From the change above, one can now remove the definition and the allocation of ub2_b/vb2_b un_bf/vn_bf with RK3 (these are purely related to a leapgrog time stepping). - Move definition and allocation of un_adv/vn_adv from dynspg_ts into oce to prevent from circular dependencies. - Restored read/write of barotropic arrays in restart with RK3 in the case where ln_bt_av=.false. (Demange's time stepping with continuous in time barotropic time stepping). Aditionnal arrays need also to be stored in case of refluxing with AGRIF.
-
Jérôme Chanut authored
some of these were producing out of bounds arrays
-
- Apr 05, 2022
-
-
Tomas Lovato authored
-
- Apr 04, 2022
-
-
Tomas Lovato authored
-
- Mar 30, 2022
-
-
Jérôme Chanut authored
This causes seg. fault with RK3, these arrays being undefined
-
Sebastien Masson authored
-
- Mar 29, 2022
-
-
Sebastien MASSON authored
-
- Mar 28, 2022
-
-
sebastien masson authored
-