Earth system model ARTS-ESM-L: basic properties, simulation of present climate and historical climate changes
Abstract
Abstract The paper presents a new Earth system model ARTS-ESM-L, developed on the basis of INMCM48 with updates to its main components. CMIP7 DECK experiments are used to analyse the processes determining the model’s climate sensitivity and present-day climatology biases, and to evaluate simulated histor-ical climate changes. Equilibrium climate sensitivity increased from 1.8 to 2.9 K and transient climate response from 1.3 to 1.8 K, moving the model from the lowest CMIP6 values to near the ensemble mean. A radiative feedback decompo-sition attributes the weakening of the net feedback mainly to the cloud feedback, which notably increased after the switch to a prognostic cloud parameteriza-tion. In ARTS-ESM-L, mixing of cloud air with unsaturated environmental air reduces both cloud water and cloud fraction in a warmer climate, whereas in INMCM48 cloud water diagnosed from temperature increased with warming. The new scheme also brings global mean cloud radiative forcing close to observations and reduces Arctic and Southern Ocean surface temperature biases. The Atlantic meridional overturning circulation at 26.5° N is stronger than observed despite realistic deep mixed layer volume in the Labrador and Nordic seas, because the main deep convection occurs west of Ireland. The ARTS-ESM-L reproduces the observed historical changes in global mean near-surface and lower-stratospheric air temperatures and land snow cover extent. Driven by historical anthropogenic carbon emissions and land use changes, the model simulates atmospheric CO2 and land and ocean carbon uptake reasonably well. The ARTS-ESM-L is now ready for the CMIP7 Assessment Fast Track experiments and palaeoclimate simulations.
Fields of Study & Subjects
Climate variability and models
Meteorological Phenomena and Simulations
Atmospheric aerosols and clouds
Publication details
- Journal / source
- Research Square
- Publication date
- 12 March 2026
- Language
- EN
- Document type
- Preprint