Overview¶
Model simulations (also called free-running simulations) are outputs from physics-based Earth system models that integrate forward in time without continuous observational constraint. They are the principal tool for studying climate variability, climate sensitivity, and future projections — and they make up a large fraction of the data hosted on NCAR’s GDEX.
This section demonstrates how to access simulation output and use it for scientific analysis.
Characteristics¶
Forward integration — once initialized, the model evolves under its own physics
No observational constraint during runtime — the model can drift from observed climate because of physics biases
Large ensembles — many simulations are run as ensembles (different initial conditions or parameter perturbations) to characterize internal variability and uncertainty
Multiple scenarios — projections often span several forcing pathways (e.g., SSP1-2.6, SSP5-8.5)
Physically consistent — variables are coupled through the model equations, so derived quantities (energy and water budgets) close
Datasets in this section¶
CESM2 Large Ensemble (LENS) — 100-member ensemble of the Community Earth System Model; used here for GMST anomaly, ocean heat content, and GeoCAT-based visualization examples.
NA-CORDEX — North American CORDEX downscaled regional climate simulations.
Where to start¶
If you are new to CESM or large-ensemble analysis, start with the CESM GMST notebook, which walks through computing the ensemble-mean global-mean surface-temperature anomaly from LENS.
Related sections¶
See Reanalysis for the observationally constrained counterpart to free-running simulations.
See Data Fusion for examples that combine simulation output with observations.