CMAT / CVDP Multi-metric Assessments

[jfasullo]

How do the CESM3 development runs perform against observations and against each other?

[jfasullo]

Here is the latest CMATv1 scoring table, sorted by total score, highlighting the high scores for #121 historical (which used raw CMIP6 BB). Still a detectible improvement from many earlier runs (internal variability 1 sigma=1).

[jfasullo]

An improvement in dynamics is particularly notable. Attached are the summary Z500 plots for ENSO teleconnections. 121 scores 0.86 while 118 scores 0.6 (similar to other model versions). These scores are in the top-right panel title and derived from pattern correlations with obs [center left] and CESM2 [center right] with differences below.