Model = HyCAM.ftorch.FAMIPC5.f19.1990-2005.MJO. Generated by run_alldiag.sh (variability + climatology-vs-obs).
Wheeler-Kiladis spectra, MJO index & χ200, lag cross-correlation, combined & spatial EOF, composite life cycle, seasonal wavenumber-frequency — vs ERA/TRMM.
AMWG package vs bundled obs (NCEP/ERAI/ERA40/GPCP/TRMM/CERES...) plus an ERA5/IMERG comparison (maps + zonal-mean cross-sections, with RMSE).
Global-mean total energy & precipitable water over the full 16-yr AMIP integration (1990-2005). Bold = the two AMIP runs (HyCAM vs the stock CAM5 AMIP control, same observed forcing — apples-to-apples); thin = SPCAM / CAM5 / HyCAM F2000 references (perpetual-2000, different compset, magnitude/drift context only). HyCAM TE drifts up ~3.32→3.38 GJ/m²; CAM5 AMIP sits lower & flatter.
Side-by-side Wheeler-Kiladis spectrum, lag-propagation, composite life cycle & wavenumber-frequency vs observations — same compset/grid/period/tapes, only the physics differs. Shows stock CAM5’s weak non-propagating MJO vs HyCAM’s recovered eastward MJO.
Generated by run_alldiag.sh.