Diagnostic Dashboard

IFS-NEMO demonstrates the highest fidelity in reproducing global climate variability, whereas ICON exhibits a systematic bias characterized by excessive continental volatility and suppressed oceanic convective variance.

High-resolution IFS-coupled models demonstrate exceptional global fidelity and successfully resolve CMIP6-era topographic and coastal upwelling biases, whereas the ICON model is compromised by severe, systemic circulation and radiative errors, though all models persistently struggle with parameterized convection and marine stratocumulus clouds.

High-resolution DestinE models fail to capture the observed 1990–2014 intensification of the Pacific Walker Circulation and underestimate Arctic amplification trends, confirming that resolution alone does not correct the phase mismatch of internal decadal variability inherent to free-running coupled simulations.

DestinE high-resolution models significantly narrow the inter-model spread of CMIP6 and reverse long-standing shortwave radiation biases, but they introduce an overly active hydrological cycle and systematic cold biases driven by an overly reflective planetary albedo.

High-resolution (~5 km) DestinE models demonstrate marked improvements in downward surface shortwave radiation over the CMIP6 multi-model mean, though they continue to exhibit overly active hydrological cycles and distinct systemic biases, with IFS-FESOM outperforming ICON in representing the baseline climate state.

High-resolution models improve the representation of upwelling zones and topographic features but suffer from a systematic global cold bias and persistent North Atlantic circulation errors, with IFS-NEMO outperforming ICON's extreme ocean-cooling/land-warming dipole.

IFS-NEMO outperforms other high-resolution models by minimizing the double-ITCZ bias, though all simulations exhibit a systematic global wet bias, realistic grid-scale storm intensity overestimations, and a persistent drying deficit over the Amazon.

The high-resolution DestinE models successfully correct the CMIP6 global Earth Energy Imbalance bias and reverse Southern Ocean shortwave errors, yet heavily rely on spatial error compensation to mask persistent structural biases in marine stratocumulus and surface longwave cooling.

While the ~5 km DestinE models capture global SST temporal trends, they exhibit strongly diverging mean states and persistent classical boundary current biases—with IFS-NEMO performing best overall while ICON suffers from extreme regional and seasonal errors.

While IFS-NEMO accurately simulates sea ice area and extent, all high-resolution models struggle with profound volume and thickness biases—with IFS models massively overestimating Arctic ice and ICON nearly losing Antarctic summer ice—highlighting that high resolution alone cannot overcome uncalibrated polar energy budgets.

Despite ~5 km resolution, all models suffer from severe initialization shocks, large intrinsic drifts, and persistent classic coupled biases—such as Western Boundary Current errors and a pervasive >3 PSU Arctic summer salty bias—indicating that increased resolution does not inherently close global energy and freshwater balances.