jenkfgdas_eupd.ecf (GFS v17) — Deep-Dive Analysis

Path: ecf/scripts/enkfgdas/analysis/create/jenkfgdas_atmos_ens_update.ecf | Date: 2026-06-11 Note: Renamed from eupd to atmos_ens_update in v17.

Summary

EnKF update (EUPD) — performs the Ensemble Kalman Filter update to generate ensemble member analysis increments from ensemble forecasts and observations. CRITICAL — this is the core EnKF algorithm.

Resources: 35 nodes, 9 mpi x 14 OMP = 126 cpus/node, exclusive, 55-min wall

Modules: gsi | J-Job: JGLOBAL_ENKF_UPDATE

v17-specific notes

• Heavy resources: 35 nodes x 126 cpus = 4,410 cores. Largest EnKF task.
• Hybrid threading: 9 MPI x 14 OMP per node.
• 55-min walltime for 80-member EnKF update.
• CRITICAL on critical path: produces ensemble analysis that feeds recentering and next cycle.