cc: pmcox@meto.gov.uk, p.j.valdes@reading.ac.uk, jfbmitchell@meto.gov.uk, sfbtett@meto.gov.uk date: Fri, 28 Aug 1998 17:29:45 +0100 from: Peter Cox subject: Re: writing Thematic proposal to: Keith Briffa Keith, I attach a draft of the model description/development section of the NERC Thematic proposal which I produced by editing Simon's contribution and adding some bits on the carbon cycle. Let me know if you need anything else. Cheers Peter Believable predictions of climate change require models which both reproduce the current climate and its variability, and also respond realistically to previous climatic forcings (e.g. well documented changes in CO2 and solar insolation). The model development programme at the Hadley Centre mirrors these goals, in aiming to improve the representation of the processes already included in GCM's, and also to incorporate new processes and related feedbacks which may be relevant to future (and indeed past) climatic changes. On both counts, the third generation Hadley Centre coupled climate model, HadCM3, must be considered a significant advance over its predecessor. HadCM3 (Gordon et al,?) was successfully developed to remove the need for ocean flux-corrections and multi-century integrations have already been completed which show minimal climate drift. A suite of improvements were made to both the ocean and atmosphere components of the GCM. The ocean model resolution was increased to 1.25 degrees by 1.25 degrees (from 3.25 by 2.5 degrees) and this along with new parametrisations of mixing processes (due to Gent and McWilliams,? and Visbeck,?) has resulted in a much improved simulation of ocean currents. The atmospheric component of HadCM3 includes a state-of-the-art radiation code (Edwards and Slingo,?), which allows explicit representation of the major greenhouse gases and aerosols and a parametrisation of momentum transport by convection (Kershaw and Gregory, ?). A new land surface scheme, called ``MOSES'', accounts for the effects of soil water phase changes and the impact of CO2 concentrations on transpiration (Cox et al, 1998). HadCM3 represents the first step in the development of an Earth System Model in that it includes a sulphur-cycle sub-model which converts prescribed emissions into aerosol concentrations, which can influence the Earth's radiation budget both directly and indirectly through their influence on the properties of clouds. Other work at the Hadley Centre is also extending the remit of the GCM by including formerly prescribed elements as interactive components of the model. A variant of HadCM3 is being used as the basis for the first coupled climate-carbon cycle model. "HadCM3LC" will incorporate models of the ocean and land carbon cycle which have been developed at the Hadley Centre in collaboration with partners from elsewhere. "HadOCC" represents chemical uptake of carbon dioxide by the ocean and includes a simple 4-box model of ocean biology which was developed at the Southampton Oceanography Centre. An interactive vegetation model, "TRIFFID", resolves the terrestrial biosphere into five plant functional types which can grow and spread based on the competition between the types and the carbon balance of each. Land surface parameters (such as albedo and roughness) are updated based on the changing vegetation distribution and structural properties, so that physical climate -vegetation feedbacks can be included as well as those associated with CO2. The model developments outlined above will require additional and new methods of validation. Slow processes such as forest migration and carbon uptake cannot be adequately validated by detailed observation of the current climate system and its variability. Here, palaeoclimate data is likely to be especially useful, as it offers the possibility of testing the ability of Earth System models to reproduce past equilibrium states of climate, vegetation and CO2. An effective partnership with the Hadley Centre in this area would provide valuable guidance on future model developments and improvements. With this in mind we would aim to complement the experiments which the Hadley Centre plans to carry out before the end of 1999: 1) A 1000-year simulation with constant forcing. 2) An ensemble of 4 simulations in which the model will be forced with anthropogenic forcings (Greenhouse gases, sulphates and ozone) for the period 1850--2000. 3) An ensemble of 4 simulations in which the model will be forced with changes in solar irradiance (Lean et al) and Volcanic aerosol changes (Sato et al) for the period 1850--2000. 4) An ensembles of 4 simulations in which the model will be forced with changes in well mixed Greenhouse gases for the period 1850--2000. 5) One of the anthropogenic and one of the well mixed greenhouse gas simulations will be continued to 2100. 6) Control (zero CO2 emissions) and anomaly (anthropogenic emissions) coupled climate-carbon cycle simulations for the period 1850--2100.