Influence of Mean State Biases on Projections of the Tropical Warming Pattern

The rise in anthropogenic greenhouse gases since the 20th century has led to regionally varying warming rates. Observations over recent decades reveal subdued warming—or even cooling—in the eastern and southeastern tropical Pacific, linked to intensified equatorial trade winds. Understanding this warming pattern is crucial due to its wide-reaching impacts: it alters atmospheric stability, driving rainfall changes through the warmer-get-wetter mechanism; affects El Niño variability and tropical cyclone intensity; and influences cloud cover, planetary albedo, and transient climate sensitivity.

While  models from the Coupled Model Intercomparison Project (CMIP), including IPSL-CM6A-LR, generally capture the subdued warming in the Southeast Pacific, they project enhanced warming in the eastern equatorial Pacific (El Niño-like response) that contradicts observations. A major factor behind this discrepancy could be the persistent cold and dry equatorial Pacific bias in these models, particularly pronounced in IPSL-CM6A-LR

To test this hypothesis, we analyze coupled flux-corrected simulations designed to reduce mean state biases. Corrections to momentum and heat fluxes mitigate cold tongue and western Pacific dry biases, as well as easterly wind errors. However, the double Intertropical Convergence Zone (ITCZ) bias remains substantial. We examine how these biases influence the tropical Pacific warming pattern during the historical period and under 21st-century climate projections, while addressing the limitations of stationary flux correction. Finally, we outline planned sensitivity experiments to explore the key physical processes driving the tropical warming pattern in response to climate change.