Uncertainty in model simulations arises due to the construction of the model (structural uncertainty), the representation of sub-grid scale processes (parameter uncertainty) or the input of model boundary conditions. Perturbed physics ensembles (PPEs) produce an ensemble of simulations using a single climate model. A PPE produces different representations of climate by altering the tuning of parameterisations representing processes occurring on sub-grid scales, such as clouds and radiation. A PPE has been produced to investigate model parameter and boundary condition uncertainty for the mid-Pliocene Warm Period (3.264 to 3.025 Ma BP). Through the use of a PPE, 14 versions (13 perturbed members and the Standard version) of the UK Met Office atmosphere-ocean general circulation model HadCM3 were created. The full ensemble was re-run to assess the impact of simultaneously changing physical boundary conditions for orography, ice sheets and vegetation in combination with perturbed physics. Finally the effect of the potential range in reconstructed mid-Pliocene CO2 was investigated through a sub-ensemble of the PPE. Using data-model comparisons (DMCs), the ensemble members with higher than the Standard values of Charney sensitivity were better able to simulate the magnitude of high latitude mid-Pliocene warming. The strongest performing ensemble members for the DMCs displayed Charney sensitivities of 4.54°C, 4.62°C and 5.40°C, above the upper bound of the IPCC likely range (1.5 to 4.5°C). However, these warmer members with higher Charney sensitivities weakened the data-model comparison in the tropics. Ensemble members with lower than Standard values of Charney sensitivity, close to the lower bound of the IPCC likely range, better resolved temperature reconstructions in the tropics, but were unable to resolve high latitude warming. It is evident that the PPE is able to achieve the magnitude of mPWP warming but not the spatial distribution of the warming. The investigation into boundary condition uncertainty using the PPE reveals that the PRISM3D physical boundary conditions lead to improved simulations of the mPWP climate than the PRISM2 boundary conditions. For the range of atmospheric CO2 concentrations, the results from the sub-ensemble indicate that lower values of CO2 lead to reduced performance of the PPE members compared to the palaeo-data. The conclusion is that concentrations of CO2 below 350 ppmv for the mPWP would make simulating high latitude climates very difficult for climate models.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:674982 |
| Date | January 2015 |
| Creators | Pope, James Owen |
| Contributors | Haywood, Alan ; Dowsett, Harry ; Collins, Matthew ; Lunt, Daniel |
| Publisher | University of Leeds |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://etheses.whiterose.ac.uk/10443/ |
Page generated in 0.0018 seconds