Modelling equable climates of the Late Cretaceous: Can new boundary conditions resolve data-model discrepancies?

S. J. Hunter*, A. M. Haywood, P. J. Valdes, J. E. Francis, M. J. Pound

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

23 Citations (Scopus)


Late Cretaceous (Maastrichtian) climate and vegetation is modelled using the HadCM3L fully-coupled atmosphere-ocean model and the TRIFFID dynamic vegetation model. We investigate data-model inconsistencies in the high-latitudes and continental interiors by exploring the sensitivity of modelled terrestrial climate to vegetation treatment, changing atmospheric pCO2 levels and the representation of Arctic seaway connections. We expand on previous work by using millennial-scale GCM runs with dynamic vegetation to allow for improved representations of ocean heat transport and terrestrial boundary conditions.Incorporating realistic vegetation drives high-latitude warming particularly during summer through reductions in surface albedo and induced atmosphere-ocean feedbacks. Resulting regional warming can exceed 10°C. As pCO2 rises some regions cool as deciduous to evergreen change increases albedo. Incorporating enhanced Arctic connectivity, reconfigured ocean heat transport drives widespread terrestrial warming of ~3°C and >5°C regionally. Applying sensitivities in combination significant palaeobotanical data-model inconsistencies in the northern high-latitudes and continental interiors remain. Further work is required to resolve climate and vegetation model deficiencies and improve the interpretation and geographic distribution of quantitative climate-sensitive geological proxies.

Original languageEnglish
Pages (from-to)41-51
Number of pages11
JournalPalaeogeography, Palaeoclimatology, Palaeoecology
Publication statusPublished - 15 Dec 2013


  • Equable climates
  • HadCM3
  • Late Cretaceous
  • Maastrichtian


Dive into the research topics of 'Modelling equable climates of the Late Cretaceous: Can new boundary conditions resolve data-model discrepancies?'. Together they form a unique fingerprint.

Cite this