NERC proposal 08

Proposal for July08 NERC standard grant round

Project Partners

UEA (Lenton / Goswami) Tyndall (Warren / Santos) Manchester (Riley / Ford) [Bristol (Williams)]

Case for Support

Work Packages

WP1 - Coupling new models to GENIE

GENIE team will need to write this. Question from Rachel - do you still want to include Hankin - uncertainty in GENIE - here? Or has it been decided this not the way to go?

WP2 - Coupling new models to CIAS

The “Community Integrated Assessment System” (CIAS) is a unique multi-institutional, modular, and flexible integrated assessment system for modelling climate change. At the moment, CIAS is fundamentally a deterministic simulation model and thus its main application is currently to compare different future scenarios of the world. Uniquely, CIAS is supported by a framework SoftIAM which allows various combinations of component computer codes, known as modules, to be connected together into alternative integrated assessment models as follows. This has been made possible by rapid advances in software engineering (see NERC strategy, 8.4). Modules in CIAS include E3MG, a global energy-environment-economy module, including a representation of induced technological change, from the University of Cambridge; an emissions scenario converter, E3MG_ESM; (or alternatively IPCC_ESM, not shown, which provides emissions scenarios used in IPCC assessments); a global simple climate module (SCM), MAGICC; a downscaling module (DSM), CLIMGEN (both University of East Anglia); a global climate impacts module for biome shifts (Potsdam Institute for Climate Research, Germany); a hydrological module MacPDM (Walker Institute); and the coastal module DIVA (consortium led by University of Southampton). In particular, the combination of the SCM and DSM together provide geographical patterns of projected changes in monthly temperature and precipitation from several coupled general circulation models at half degree resolution and are currently being upgraded to match the latest IPCC science. Users are able determine the particular CIAS coupled model they wish to use through a web portal. The SoftIAM portal includes the facility to implement Latin hypercube experimental design (Hankin 2005) facilitating formal uncertainty analysis. The portal allows the user to choose the (marginal)

Figure 1.

distribution for any subset of the parameters present, and to specify the total number of model runs performed. The user may specify any of a wide range of statistical distributions for the uncertainty parameters, including normal, lognormal, uniform, triangular, beta, or Davies (Hankin and Lee 2006) distributions, the latter being specifically designed for risk assessment. Figure 1 illustrates the CIAS coupled model. Illustrative preliminary results have been published (Warren et al. 2008.) for stabilization scenarios considering CO2 only. Component codes are now being upgraded, and under existing funding additional modules will be connected including a water stress module and an agricultural model. An alternative climate model GENIE, in executable form, through the NERC project GENIEfY on which I am co-investigator, will also be connected. Finally, a set of initial multi-gas scenario simulations will be produced, in particular a set of ‘standard stabilization scenarios’ which result in stabilization of greenhouse gases in the atmosphere at 450, 500 and 550 ppm CO2 by 2100 or shortly thereafter. Warren et al. (in press) describes the model and its components in more detail. The CIAS system already links together an integrated research community across the UK and the Potsdam Institute of Climate Impacts Research in Germany. Links are currently being forged with institutes in South Africa and the USA, to further build on a world-leading network in environmental science, and hence directly satisfies the implementation goal mentioned in NERC draft strategy 1.19.

OBJECTIVES

The first aim of this work is to provide within CIAS a tool to study the interactions between climate change policies and land use change policies: in particular avoided deforestation and bio-fuel policies. Avoided deforestation is a key policy component for potential post-Kyoto climate regimes, whilst plans for mitigation involving biofuel cropping have major land use, water, poverty and biodiversity implications. The second aim of the work is to apply the system to study the effects that particular land use policies have in relation to the stabilization (and ultimate further reduction) of greenhouse gas concentrations in the atmosphere, taking into account issues of food security.

We will link together the JULES land use change modelling system provided by our partners at the Hadley Centre, and further link in agricultural models through colleagues at the University of Reading. The agricultural models are already being linked to JULES through the QUEST program (led by the Walker Institute). Initially JULES will be linked into CIAS via the IMOGEN system, but we will also explore the use of Weather Generators to link JULES directly to the outputs of other climate models in CIAS (partner Rob Wilby). IMOGEN contains a simple weather generator, which produces climate outputs on a daily timestep as required by JULES. Upgrading of the climate models in CIAS will also be necessary (partner Sarah Raper) to make them consistent with outputs of the IPCC Fourth Assessment.

Add description of JULES (Richard Betts/Jason) Add description of MAGICC6 (Sarah Raper)

NOTE TO ALL: I HAVE ALREADY HELD DISCUSSIONS WITH RICHARD BETTS AND JASON LOWE AT HADLEY ABOUT THIS: THEY WILL HAVE A PERSON AT HADLEY TO WORK WITH ME ON THIS AND WE HAVE AGREED THIS WOULD WORK BEST IF I ALSO HAD A PERSON AT THIS END. HENCE THIS PIECE OF FUNDING IS MEANT TO FUND THE LAND USE PERSON AT THIS END. I HAVE ALSO SPOKEN TO ROB WILBY AND HE WOULD LIKE TO BE COSTED IN TO THE PROPOSAL FOR AN AMOUNT OF TIME TO BE DISCUSSED TO PROVIDE ADVICE ON THE POSSIBLE USE OR ASSIST WITH THE ACTUAL INCORPORATION AND USE OF WEATHER GENERATORS IN CIAS. SARAH RAPER ALSO NEEDS A SMALL AMOUNT OF FUNDING TO PAY FOR HER CONTINUED HELP AND ADVICE IN CIAS. HENCE ROB WILBY, RICHARD, JASON AND SARAH NEED TO BE CC'ED AND INVOLVED IN THE FINALISATION OF THE PROPOSAL.

ANOTHER NOTE TO ALL: I AM NOT INCLUDING THE LINKS TO THE ECONOMICS MODEL IN THIS PROPOSAL. I HAVE STILL NOT RECIEVED THE FINAL VERSION AND IN ANY CASE THAT COUPLING IS ALREADY FUNDED AND I FEEL IT IS BETTER TO KEEP THE PIECES OF WORK SEPARATE. SURE THERE ARE RELEVANT ECONOMIC ASPECTS BUT THIS IS THE NERC PROPOSAL SUBMISSION.

WP3 - Coupling GENIE and CIAS with BFG2

I suggest we start with text concerning the technical side of the coupling ....and how to improve that. I'll leave that to Graham and Rupert, to be agreed with Santi. Santi is able to make the meeting on the 19th so this can be discussed there.

NOTES TO ALL: We have already had funding to couple CIAS and GENIE, so I am not sure that we can really include here the funding needed to complete this. BUT

(A) Currently Mike Salmon is working on some finalisation of GENIE I understand, after which he will do the multigas conversion. However it may be that multigas conversion is not in the previous GENIEfY proposal (?) - if not then here is some suggested text below which would follow the technical part, on that ...also it may be that we could apply for funding for a more sophisticated long term approach to the work Mike Salmon has agreed to do (although Mike should still do his work and we'd use that initially).

(B) The other question is - did we actually agree to couple in the GENIE executable or the modularised GENIE in the GENIEfY proposal? If we did not specify, we could argue that in the GENIEfY proposal we only coupled in the executable, and that we now want to couple in the modularised version - again suggested text below. However I have a feeling that there was extensive text in the GENIEfY proposal about the intention to couple in the modularised system.

Three scientific coupling tasks are relevant here (A) Multigas conversion of GENIE for use within CIAS (B) Coupling of modularised GENIE to CIAS

(A) MULTIGAS CONVERSION OF GENIE FOR STABILIZATION SCENARIO ANALYSIS WITHIN CIAS Under existing funding the GENIE model has been connected to the CIAS model, to enable coupled climate-economy experiments to be carried out which study the levels of carbon taxes necessary to stabilize greenhouse gas concentrations in the atmosphere by 2100 at various levels. [Using emissions scenarios only rather than an economics model, CIAS can also explore the implications of trajectories beyond 2100, but the economic model cannot project beyond 2100]. The current version of GENIE contains only CO2 concentration simulations, and since stabilization scenario analysis required consideration also of CH4, N2O, CFCs, HFCs, PFCs and SF6, the model requires modification so that the concentrations of these gases can be simulated given anthropogenic emission trajectories from CIAS over the 21st century. (GENIE group may wish to add detail; or not if we just use this to get extra funding for Mike Salmon).

(B) COUPLING OF MODULARISED GENIE TO CIAS Under existing funding the version of the GENIE model coupled to CIAS is an executable version. Additional benefits can be obtained through coupling in the modularised system, which allows a more flexible approach to the integration of climate and land use, and specific exploration of feedbacks in the system.

WP4 - Doing some science

SCIENCE FROM THE NEW MODULES COUPLED TO CIAS (WP2) When the land use-agricutlural modelling system is operational within CIAS we will use it to study in a spatially specific manner the following: (i) how inclusion of avoided deforestation or biofuel cropping in various regions on various timescales lowers the levels of fossil fuel emission reductions needed to achieve various stabilization scenarios for greenhouse gases in the atmosphere, and thus enhances the probability of constraining global temperature rise to within, for example, a 2ºC threshold. If the CIAS economic model is operational we would also be able to provide estimates of how carbon taxes needed to remain below the 2C threshold might be reduced. (ii) under which biofuel cropping scenarios more land might be required to avoid competition with non-fuel agriculture (iii) in cases where more land might be required, we would assess where that might be and whether it would conflict with the need to preserve large areas of natural ecosystems to assist their adaptation to climate change (in relation to NERC strategy goal 3.2),and avoid deforestation (iv) where the water requirements of biofuels, ecosystems and agriculture might conflict, rendering envisaged efficient land use policies infeasible due to water constraints. CIAS already has the potential to project drought frequency and drought length during the 21st century under a wide range of climate change scenarios, and this capability will be used to complement this study.

We will seek advice from Carlos Peres at UEA on avoided deforestation scenarios, whilst for biofuels scenarios we will draw on the advice of the UEA biofuels network led by Alastair Grant and of Andrew Lovett who has worked on biofuel/land use competition issues at the UK scale at UEA.

Outputs: Deforestation and biofuel cropping scenarios to 2100 detailing spatial use of land (protected forest, biofuels, agricultural land); estimates and maps of the sources and sinks of carbon in these scenarios with a particular focus on the uptake by the forests in regions where deforestation is avoided or the releases due to biofuel cropping in formerly forested areas; benefits for reduced implications of reductions in fossil fuel emissions required to reach particular climate goals; and estimates of global cereal production. Journal papers using these scenarios to illustrate policy relevant aspects of the interaction between land use change policy and climate policy.

SCIENCE FROM THE COMBINED CIAS_GENIE SYSTEM (WP3)

We will use the coupled CIAS-GENIE system to study: (i) The carbon taxes and emission scenarios necessary to achieve stabilisation of greenhouse gases in the atmosphere by 2100, at various levels, with and without 'overshoot' of these concentrations in intermediate periods. We will include in our studies the particular overshoot scenarios which are to be studied by the QUEST Overshoot Scenarios workshop to be held June 16-18th. Currently in CIAS such studies are carried out only with the MAGICC model used in the Third Assesment Report. A comparison of the MAGICC-based and GENIE-based solutions to the problem of stabilization will provide interesting insight into the functioning of the simple climate model set to emulate the behaviour of various GCMs, versus a more sophisticated EMIC approach. (ii) The uncertainties in the above analysis, using the techniques developed by Hankin (Robin to add how will do uncertainty analysis of GENIE and GENIE-E3MG coupling, if he thinks this can be done)

Work Plan

Year 1

• WP1 - Manchester and UEA work on coupling new models of physics and land surface and also on multigas implementation (in reality this bit will hopefully be done before this proposal is funded)

• WP 2 - Tyndall to integrate latest economic / land use etc models and

• WP 3 - ongoing effort to keep up-to-date GENIE on CIAS server