.. _release_note_573: MNH-V5-7-3 ============================================================================ Release date : 04/05/2026 .. note:: We fully encourage all users of 5-7-2 to move to 5-7-3. As this release includes only bug fixes, the results should be exactly the same as 5-7-2 (except for possible corrections), but with more stability and reliability. This is true in the same environment (especially with the same compiler version and options). Aircraft ********************************** * ini_aircraft: ensure that first trajectory time is 0. (time since launch) * treat case of aircraft initially outside of physical domain * ini_aircraft: set NMODEL=1 automatically if only 1 domain (does not need to set it explicitely in namelist anymore) IBM ********************************** * fix for IBM_GOTO_MODEL * IBM_INTERPOS: bugfix: use XXHTAM and XYHATM instead of MXF(XXHAT) and MYF(XYHAT) that introduce bad values near boundaries * rewrite of IBM_LOCATCORN: fix for vertically non-regular grids + optimizations + safeguards SURFEX ********************************** * READ_SURFX2COV_MNH / READ_SURFX2COV_1COV_MNH: bugfix: force GCOVER_PACKED to FALSE if not found in file (instead of having an unpredictible value) * WRITE_SURFX2_MNH: do not write emission fields (no real meaning in backup files + problem with last dimension which is the number of instants and not the vertical levels * rewrite temporal_dists to handle dates < 1990 + simplify the algo using the 1st March deviation * fix when using TEB and sea coupling Outputs & Backups ********************************** * minor fix for outputs: XOUT_TIME_FREQ_FIRST(s,m)=0. now allows output at start (instead of being ignored) * budgets: correct condition to write vertical coordinates if MASK * reduce NBULISTMAXLEN to 128 to workaround limitation of gfortran compiler for size of character strings in namelists Others ********************************** * fix when using export VER_CDF=CDFPERSO * fix for LFILTERING in NAM_HURR_CONF * increase compilation timelimit for compilation to 1h30 on HPC * set_perturb: remove MPPDB_CHECK of XTHM (no more meaning + crash in some cases) * fix: pass XSVT and not XSVT(:,:,:,1) to MEAN_FIELD to prevent crash in DEBUG mode if NSV=1 * fix when using coupling with oceanic currents + refactoring * fix: XACPRR used before allocation (when using old electricity scheme, fix issue #52)