Last glacial maximum constraints on the Earth System model HadGEM2-ES

Peter O Hopcroft,Paul J Valdes

doi:10.1007/s00382-014-2421-0

Peter O Hopcroft, Paul J Valdes

Open Access

https://doi.org/10.1007/s00382-014-2421-0

Copy DOI

Journal: Climate dynamics	Publication Date: Dec 2, 2014
Citations: 27	License type: CC BY 4.0

Affiliation: University of Bristol

Abstract

We investigate the response of the atmospheric and land surface components of the CMIP5/AR5 Earth System model HadGEM2-ES to pre-industrial (PI: AD 1860) and last glacial maximum (LGM: 21 kyr) boundary conditions. HadGEM2-ES comprises atmosphere, ocean and sea-ice components which are interactively coupled to representations of the carbon cycle, aerosols including mineral dust and tropospheric chemistry. In this study, we focus on the atmosphere-only model HadGEM2-A coupled to terrestrial carbon cycle and aerosol models. This configuration is forced with monthly sea surface temperature and sea-ice fields from equivalent coupled simulations with an older version of the Hadley Centre model, HadCM3. HadGEM2-A simulates extreme cooling over northern continents and nearly complete die back of vegetation in Asia, giving a poor representation of the LGM environment compared with reconstructions of surface temperatures and biome distributions. The model also performs significantly worse for the LGM in comparison with its precursor AR4 model HadCM3M2. Detailed analysis shows that the major factor behind the vegetation die off in HadGEM2-A is a subtle change to the temperature dependence of leaf mortality within the phenology model of HadGEM2. This impacts on both snow-vegetation albedo and vegetation dynamics. A new set of parameters is tested for both the pre-industrial and LGM, showing much improved coverage of vegetation in both time periods, including an improved representation of the needle-leaf forest coverage in Siberia for the pre-industrial. The new parameters and the resulting changes in global vegetation distribution strongly impact the simulated loading of mineral dust, an important aerosol for the LGM. The climate response in an abrupt 4× pre-industrial CO2 simulation is also analysed and shows modest regional impacts on surface temperatures across the Boreal zone.

Highlights

During the last glacial maximum (LGM) the global mean temperature was 3–5 °C cooler than during the preindustrial era (Jansen et al 2007; Braconnot et al 2012)
Using an atmosphere-only version of HadGEM2 we investigated the climatic response to glacial climate forcings
HadGEM2 responds strongly to LGM boundary conditions and the version of the TRIFFID dynamic vegetation model coupled interactively within HadGEM2 simulates a massive expansion of bare soil at the LGM which is not supported by palaeo-vegetation compilations (Prentice et al 2000) or other modelling studies (Crucifix and Hewitt 2005b; Harrison and Prentice 2003)

Summary

Introduction

During the last glacial maximum (LGM) the global mean temperature was 3–5 °C cooler than during the preindustrial era (Jansen et al 2007; Braconnot et al 2012). Equal contributions to this cooling can be attributed to the large ice-sheets and reduced levels of major greenhouse gases that characterised this time period (Petit et al 1999; Peltier 2004; Braconnot et al 2012). The LGM has been investigated in detail as it is thought to provide a good test for climate model responses to changes in greenhouse gas-induced radiative forcing (Braconnot et al 2012). Coordinated ensembles of climate model simulations of this period under standard experimental protocols have been established and analysed in some detail (Braconnot et al 2007a, b; Izumi et al 2013; Schmidt et al 2014).

Methods

Results

Conclusion