Climate models
How well CMIP6 captures North American storm pulses
Open access · cc by · source: Europe PMC
An Eulerian three-parameter storm metric shows CMIP6 GCMs share cyclone biases versus ERA5, while 12 km CRCM6 is usually closest.
Study at a glance
- Design
- Computational / modelling — Eulerian intense-cyclone metrics comparing ERA5, 27 CMIP6 GCMs, and CRCM6
- N
- N=27 · 27 CMIP6 simulations plus 5 CRCM6-GEM5 runs evaluated against ERA5 (1980–2014)
- Population
- Intense low-pressure systems over North America in reanalysis and climate models
- Outcome
- Biases in cyclone depth/tendency and skill of 12 km CRCM6 vs GCMs
Structured fields used in claim comparison tables when every cited study has a complete layer.
Key findings
GCMs underestimate oceanic cyclone depth and overestimate it over land. Resolution helps most over stormy oceans. 12 km CRCM6 has the smallest domain errors even when driven by different CMIP6 models, so the regional model—not the driver—sets much of the skill.
Methodology
Authors identified the strongest yearly sea-level-pressure anomalies at each North American grid point in ERA5, 27 CMIP6 models, and CRCM6-GEM5 (1980–2014), then fit 3-day pressure and humidity traces with peak, tendency, and asymmetry.
Limitations
“Errors” are differences from imperfect ERA5, not truth; the linear 3-day fit is poor for tropical cyclones; there is no Lagrangian track or future-scenario analysis.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
An Eulerian three-parameter storm metric shows CMIP6 GCMs share cyclone biases versus ERA5, while 12 km CRCM6 is usually closest.
Evidence for the claim as stated.
CMIP is also a catalogue of process fields, not only of warming. Across 26 Indian monsoon extremes, the dynamic (vertical-velocity) moisture-budget term contributed more than 90% on average; six CMIP6 models with omega/humidity agreed that dynamics lead thermodynamics. Twenty-seven CMIP6 models underestimate North American oceanic cyclone depth and overestimate it over land versus ERA5; 12 km CRCM6 has the smallest domain errors even when driven by different CMIP6 models.
Evidence for the claim as stated.
Reanalysis is the yardstick for 'does this GCM get storms and monsoon rain right?' CMIP6 models underestimated oceanic cyclone depth and overestimated it over land versus ERA5; 12 km CRCM6 had the smallest domain errors even when driven by different GCMs. High-resolution HighResMIP versions reduced dry bias versus MSWEP (basin mean 6.33 mm/day) and beat low-resolution twins on South Asian monsoon timing, while ERA5 itself had a 2.37 mm/day wet bias versus MSWEP.
Evidence for the claim as stated.
Which reanalysis you treat as 'observed' changes the bias you assign to models. HighResMIP skill versus MSWEP is better than versus ERA5 because ERA5 is 2.37 mm/day too wet in that basin; North American cyclone 'errors' are differences from ERA5, which the authors refuse to call truth. Two papers can disagree about a model without disagreeing about the physics, if they picked different reference products.
Evidence for the claim as stated.
Multi-model ensembles are also used as a catalogue of sensitivity. CMIP5 versus CMIP6 regional extreme sensitivity (TXx, TNn, Rx1day) at +1.5, +2 and +4 °C is very similar in the multimodel mean; at +1.5 °C, regional-sensitivity uncertainty often exceeds global-sensitivity uncertainty, especially in CMIP6. Twenty-seven CMIP6 models plus CRCM6 were the ensemble for North American cyclone pulses.
Evidence for the claim as stated.
Idealised or nested experiments use small ensembles for a different purpose: tracing a drought to a basin of SST, or asking whether a regional model or its GCM driver sets storm skill. Those spreads are not interchangeable with a 100-member 21st-century probability ratio.
Evidence for the claim as stated.
Twenty-seven CMIP6 GCMs underestimate North American oceanic cyclone depth and overestimate it over land versus ERA5 (1980–2014). Resolution helps most over stormy oceans. Nested 12 km CRCM6-GEM5 has the smallest domain errors even when driven by different CMIP6 models, so the regional model — not the GCM driver — sets much of the Eulerian storm-pulse skill. Those ‘errors’ are differences from imperfect ERA5, not truth.
Evidence for the claim as stated.
Open questions
Tensions this paper is part of
From concept pages' “where studies disagree.” Disagreement means the same question; scope means different assays, populations, or outcomes.
Which reanalysis you treat as 'observed' changes the bias you assign to models. HighResMIP skill versus MSWEP is better than versus ERA5 because ERA5 is 2.37 mm/day too wet in that basin; North American cyclone 'errors' are differences from ERA5, which the authors refuse to call truth. Two papers can disagree about a model without disagreeing about the physics, if they picked different reference products.
Idealised or nested experiments use small ensembles for a different purpose: tracing a drought to a basin of SST, or asking whether a regional model or its GCM driver sets storm skill. Those spreads are not interchangeable with a 100-member 21st-century probability ratio.
Related papers in this topic
Same topic cluster — not a recommendation engine.