What is it about?
To evaluate the realism of climate models, we need to see if they can detect actual climate change signals from background ‘noise.’ While models were originally designed for impact studies rather than detecting these signals, we can use specific methods to identify when climate signals become distinguishable. We introduce the ‘Time of Detection’ (ToD), which uses a statistical test to identify these signals earlier and with more accuracy than the traditional ‘Time of Emergence’ (ToE) method. We applied ToD and ToE to assess sea surface temperature (SST) data from both observational records and climate models (CMIP5/6). Findings show that SST warming signals are already detectable across much of the tropics. However, warming signals in the Pacific are less consistent between observations and models, likely due to stronger warming patterns in models. Relative SST signals, which are generally weaker than SST, are detected in specific regions but do not fully emerge. In the western Indian Ocean, modeled warming was detected in fewer models due to data limitations, while warming in the southeast Pacific is more robust and aligns with the expansion of atmospheric patterns like the Southern Hemisphere Hadley cell. By detecting subtle signals, the ToD method helps us better evaluate the accuracy of climate model warming patterns.
Featured Image
Photo by Markus Spiske on Unsplash
Why is it important?
Recent studies highlight the importance of understanding climate change warming patterns, especially the differences between model projections and real observations in the equatorial Pacific. We propose that our new ‘Time of Detection’ (ToD) approach is more effective than the traditional ‘Time of Emergence’ (ToE) for confidently comparing these patterns. Consistent with previous research, our analysis reveals a model-observation mismatch in the equatorial Pacific. However, using ToD allows us to detect a weaker warming in the southeast Pacific, confirming that this pattern in climate models
Perspectives
This publication brings valuable clarity to how we assess climate model accuracy by introducing the ‘Time of Detection’ (ToD) method. The ToD approach seems to provide a more nuanced way to evaluate climate models by focusing on when warming signals become detectable rather than merely when they emerge against background noise. This can be particularly valuable in regions like the equatorial and southeast Pacific, where model-observation discrepancies have posed challenges in the past. By addressing these subtleties, the authors contribute meaningfully to our understanding of model realism, especially in the southeast Pacific, where their work shows climate models may realistically capture subdued warming patterns. This distinction is crucial for improving confidence in future climate projections and may help refine climate policy and adaptation strategies. Overall, this study is a thoughtful advancement in climate modeling, promoting more precise comparisons between modeled and observed warming patterns.
Gopika Suresh
CSIR-National Institute of Oceanography
Read the Original
This page is a summary of: Assessing CMIP models’ ability to detect observed surface warming signals related to climate change, Journal of Climate, August 2024, American Meteorological Society,
DOI: 10.1175/jcli-d-24-0102.1.
You can read the full text:
Contributors
The following have contributed to this page
