The HadCM3 GCM used in this study was developed at the Hadley Centre for Climate Prediction and Research, which is a part of the UK Meteorological Office. This GCM consists of a linked atmospheric model, ocean model,and sea ice model. However, for the present study we use only the atmospheric component of the model (HadAM3).

The horizontal resolution of the model is 2.50 degrees of latitude by 3.75 degrees of longitude. This gives a grid spacing in the mid-latitudes of 278 km in the North-South direction and 417 km East-West, comparable to a T42 spectral resolution model. The HadAM3 atmospheric model consists of 19 layers.

Pope et al. (2000) describes the impact of the new physical parametrizations used in the HadAM3. The land-sea mask and orographic details used do not differ from HadAM2 (Johns et al. 1997; Johns 1996). The model has a timestep of 30 minutes and a new radiation scheme that can represent the effects of minor trace gases (Edwards and Slingo 1996). Parametrizations of simple background aerosol climatology discussed by Cusack et al. (1998) are also included. The model convection scheme has also been improved according to Gregory et al. (1997). A new land-surface scheme includes the representation of the freezing and melting of soil moisture. The representation of evaporation now includes the dependence of stomatal resistance on temperature, vapour pressure, and CO₂ concentration (Cox et al. 1999). The above changes have increased the ability of the HadAM3 model to perform a realistic simulation of surface heat flux.

For this modelling study the required boundary conditions for the HadAM3 model were supplied by the United States Geological Survey’s (USGS) Pliocene Research Interpretations and Synoptic Mapping (PRISM) Groups, PRISM2 enhanced 2º x 2º digital data set, or they were specified as current values. Carbon dioxide was set at 315 ppm, a value in line with previous modelling studies for the middle Pliocene (i.e., Chandler et al. 1994). In practice, using prescribed SST means that the model is relatively insensitive to the precise choice of CO₂ concentration. Prescribed boundary conditions integrated into the model that are specific to the middle Pliocene include: (1) continental configuration, modified by a 25 m increase in global sea level, (2) modified present-day elevations (elevation of the western cordillera of North America reduced by 50 per cent; 950 m), (3) reduced ice sheet size and height for Greenland (50%) and Antarctica (33%), (4) Pliocene vegetation distributions, and (5) Pliocene SST and sea ice distributions (see Haywood et al. 2000a and Dowsett et al. 1999 for more detailed descriptions of PRISM2 boundary conditions and their integration into the UKMO GCM).

The PRISM2 data set is the latest in a line of data sets generated by the USGS that reconstruct the palaeoenvironment of the middle Pliocene. The PRISM2 reconstruction consists of a series of 28 global scale data sets on a 2× latitude by 2× longitude grid, and as such, represents the most complete and detailed global reconstruction of climate and environmental conditions older than the last glacial (Dowsett et al. 1999). PRISM2 evolved from numerous studies that summarised conditions at a large number of marine and terrestrial localities (e.g., Cronin and Dowsett 1990; Poore and Sloan 1996). A detailed description of the new PRISM2 data set, that fully documents the construction of PRISM2 and how it differs from earlier PRISM data sets, is now available as a USGS Open File Report (see Dowsett et al. 1999).