Ocean surface temperature variability: large model-data differences at decadal and longer periods

Area of Science:

• Paleoclimatology
• Oceanography
• Climate modeling

Background:

• Instrumental sea surface temperature (SST) records are too short for multidecadal variability analysis.
• Proxy records of past SSTs are often noisy, limiting their utility.
• Accurate assessment of long-term SST variability is crucial for climate change studies.

Purpose of the Study:

• To develop and apply a novel noise filtering technique to late Holocene SST proxy data.
• To estimate SST variability across a range of timescales, from annual to millennial.
• To compare proxy-based SST variability estimates with climate model simulations.

Main Methods:

• Application of a new noise filtering technique to a global network of SST proxy data (coral, foraminifer, alkenone).
• Estimation of SST variability from annual to millennial timescales.
• Comparison of filtered proxy-based SST variability with instrumental records and general circulation model (GCM) simulations.

Main Results:

• Filtered proxy-based SST variability estimates show consistency across different proxy types and with instrumental records where frequency bands overlap.
• General circulation models systematically underestimate SST variability compared to both instrumental and proxy-based estimates.
• Discrepancies between models and observations are largest at low latitudes and increase significantly with timescale, reaching two orders of magnitude at millennial timescales for tropical SSTs.

Conclusions:

• The study highlights significant discrepancies between simulated and observed/reconstructed SST variability, particularly at longer timescales and low latitudes.
• These findings suggest potential deficiencies in either current climate models, observational data, or both.
• Implications include the need for improved climate models and a re-evaluation of past climate change attribution and future climate predictions.