Case IA Modelling for London: the Tyndall Centre UIAF (Urban Integrated Assessment Facility) Jonathan Köhler and Richard Dawson

Transcription

1 Case IA Modelling for London: the Tyndall Centre UIAF (Urban Integrated Assessment Facility) Jonathan Köhler and Richard Dawson 1. Scope of study This study developed an Integrated Assessment Modelling structure for London i, combining sectoral economic analysis with transport demand analysis and environmental assessment for emissions and flood risk. a. What are the questions? Adaptation and mitigation in cities requires integrated thinking that encompasses a whole range of urban functions, at spatial scales from individual buildings to whole cities. Meanwhile, mitigation targets, which are now being adopted by cities, imply major reconfiguration of urban energy systems, transport and the built environment. If conflicts between the objectives of adaptation and mitigation ii,iii and, more generally, between urban growth and sustainable development, are to be avoided as far as possible then an integrated systems view of cities is required iv. Our systems approach seeks to represent the interactions between different urban functions and objectives by linking climate change issues to broader policies such as spatial planning. b. What is modelled? The following areas are addressed by the UIAF for London: Economics: Employment and gross value added for economic sectors in the London region. Land Use change: Land use patterns for London and the Thames gateway to CO2 Emissions: CO2 emissions from energy use, passenger transport and freight transport Heat Waves: Use of the Hadley Centre Regional Climate Model to represent the urban heat island effect. Influence of spacial patterns of development on the risk from heat waves. Droughts: The UKCP09 rainfall scenarios for the Thames and Lee catchments were combined with catchment hydrology models and simulation of the water resource management system. Potential lack of water supply in the long run identified. Flooding: A model of flooding in the tidal Thames floodplain, which is protected by the Thames Barrier and a system of flood defences, has been used to simulate the effects of sea level rise and changing flows in the river Thames. This has been combined with our simulations of land use changes to develop simulations of flooding risk for the Thames river basin and the Thames Gateway. 1

2 2. Methodology/disciplines or academic fields a. Regional economics using Input-Output model; b. Geographical modelling of land use patterns; c. Modelling of spacial distribution of employment and transport accessibility for approximately 600 wards covering Greater London; d. Transport activity and emissions e. Regional Climate model to develop scenarios of ambient temperature distributions f. Water catchment hydrology models for water supply assessment g. Flooding model for the Thames catchment area and Thames gateway h. Scenarios Long run scenario of sectoral economic development and employment Scenarios of land use change and urban area expansion UKCP2009 Climate scenarios for the UK, as a generator of temperature and rainfall scenarios and sea level rise. The UIAF (Figure 1) provides the flexibility to test a very wide range of mitigation and adaption policies by incorporating diverse evidence and representing a number of urban processes and interactions. To manage this potentially overwhelming complexity, development and application of the UIAF was implemented iteratively in six main stages in consultation with the city government, utility owners, NGOs and other stakeholders in London. (1) Define the questions that the assessment is seeking to address: our focus has been upon meeting London s emissions targetsfehler! Textmarke nicht definiert. and London s priority climate risksfehler! Textmarke nicht definiert. of heat, drought and flooding. (2) Identify the drivers of long term change within the urban area: we dealt primarily with socio-economic and climate drivers. (3) Identify the key processes of interaction that determine long term urban change: the main processes of interaction we have addressed have been via the economy, land use and transport. (4) Develop a representative set of scenarios that spans the range of possible futures: Economic growth was defined using a range of GDP scenarios for the UK. Population projections were from Office of National Statistics subnational population projections v. Climate change scenarios used the 2009 UK Climate Projections vi. 2

3 (5) Define the policy options that are intended to be analysed and the metrics of assessment: Our analysis includes policy options for: land use planning, energy policy, transport infrastructure and fuel efficiency, heat wave risk management, water resource management, flood risk management. (6) Develop and apply the integrated assessment model to quantify the performance of policy options in the context of a range of different scenarios. Socio-economic and climate scenarios provide the context for the analysis. A process of down-scaling generates climate, economic and demographic projections for the urban area. The MDM-E3 vii multi-sectoral regional economic model was used to generate scenarios of economic activity and employment for 42 industrial sectors in London and UK national energy demand by sector. Figure 1 Overview of the Urban Integrated Assessment Facility. Socio-economic and climate scenarios provide the context for the analysis. A process of down-scaling generates climate scenarios at the city scale as well as economic and demographic scenarios for the urban area which provide the boundary conditions for the city scale analysis, in this case London. A spatial interaction module provides high resolution spatial scenarios of population and land use that form the basis for analysis of carbon dioxide emissions and vulnerability to climate impacts under a wide range of scenarios of climate, socio-economic and technological change. The UIAF can be used to test the effectiveness of a wide range of mitigation and adaption policies, including land use planning, modifications to the transport systems, changing energy technologies and measures to reduce climate risks under different scenarios. 3

4 The regional climate model HadRM3 was parameterised using a tiled MOSES2 surface scheme to allow for sub-grid scale variations at the land surface to capture the urban heat island effect (Mcarthy et al., pub online 2009). For three (a low, medium and high) emissions scenarios a stochastic weather generator viii was used to downscale one hundred realisations of the HadRM3 model to provide, for each realisation, a 100 year hourly timeseries of rainfall, temperature and evapotranspiration onto a 5x5km grid over London. Aggregate projections of employment and the economy from MDM-E3 were used to drive a land use transport model (Dawson et al., 2008) that outputs spatially explicit, at the census ward scale, estimates of change in population and employment. Transport accessibility was measured in terms of Generalized Travel Cost, which accounts for time, monetary and perceived (e.g. overcrowding, safety) costs associated with travel. A transport emissions inventory for personal and freight travel calculates carbon emissions per trip within the urban area ix using Great Britain National Travel Survey x (NTS), freight data xi, and carbon emissions factors (the emissions per journey kilometre) for different modes xii,xiii. The land use model provides population changes for calculating future demand which were tested against changes in behaviour, technology and transport policy xiv,xv. A GHG emissions inventory for energy measures direct emissions, energy and heat consumed and waste produced within the urban boundary xvi. Changes in population and employment from the land use model and projections of UK energy demand from the MDM-E3 model are used to calculate future emissions, which were tested against changes in behaviour, technology and energy policy. Climate risks were calculated by integrating a climate hazard layer, derived from the 5x5km weather generator outputs, with a vulnerability layer derived from the spatially explicit projections of population and employment from the land use model. The benefit of adaptation measures were tested against these risks. Heat risk is calculated as the expected number of vulnerable people (defined as people aged younger than 4 and older than 65) to experience a heat wave event (defined in London as two consecutive days where T max >32C and the night in between T min >18C xvii ) multiplied by the number of most vulnerable people (0-4 and >65 year olds) exposed to the event. The anthropogenic contribution from land use and energy to urban heat is obtained from parametric functions established from a number of HadRM3 simulations (Mcarthy et al., 2009). Drought risk is measured both in terms of the safety margin of potential water resource and the frequency with which emergency drought orders (i.e. requiring deployment of standpipes, rota cuts and/or water tanks) would be necessary. Current and future river flows were calculated by inputting rainfall and evapotranspiration data from the weather generator into the CATCHMOD xviii rainfall runoff model over the Lower Thames catchment. Daily flow data was used as an input into a dynamic water resource system model (see Supplementary Information) that simulates water demand, reservoir operating rules, abstraction, supply and storage capacity for London and the Lower Thames region. Flood risk is measured in terms of expected annual damages xix and calculated by integrating the extreme value distribution of water levels with information on flood defence levels and 4

5 reliability and functions that relate flood depth and duration, to economic damage (in 2005 prices xx ). The sensitivity of impacts to flood depth and the variability in floodplain topography within individual census wards obligated downscaling of our land use model to 100 x 100m grid cells. Figure 2 Assessment of damages from Flood risk + 47m + 89m Baseline 2100 Eastern axis m Centralisation m Sub-urbanisation m Figure 3(a) Land use scenarios. 5

6 The shading shows the location of population in 2005 disaggregated by census ward (darker shade shows a higher population). For the 33 London boroughs and 9 local authorities in the Thames Gateway, the bar graphs (scale in legend) show (i) the 2005 population and change in population between for the four land use scenarios (ii) Current development trends, (iii) Eastern axis, (iv) Centralization and (v) Suburbanization the growth in population, aggregated from ward scale land use model outputs. Figure 3 (b) Current and future flood risk. The 2005 flood risk which totals 29m, expressed in terms of expected annual damages, is shown in shades of red at a census ward scale (darker shades denote a higher risk). The proportion of risk in 2100 in each Borough or Local Authority for the current trends land use policy from (i) Retrofit: Existing development if retrofit, (ii) Future resilience: Future development if constructed to the highest flood resilience standard, (iii) Current standards: Additional contribution to risk if there is no retrofitting of the existing building stock and no imposing flood resilient construction standards on new buildings, and (iv) Climate: additional contribution from the UKCP09 medium climate change scenario from sea level rise and changes to the frequency of extreme flows. The pie charts are scaled according to the natural logarithm of the total risk in the borough or local authority (where the size 10 shown in the legend corresponds to 22k). 6

7 Figure 4 Heat risk assessment. (a) Example spatial weather generator output showing the mean Tmin for the summer months (June, July and August) in (b) The additional heatwave risk (defined in terms of the expected number of vulnerable people exposed to heatwave events as a result of 50% increase in anthropogenic heat emissions and the additional 133km 2 of developed land in Central London associated with the Centralization land use policy. a b 7

8 Figure 5 Management of drought risk. The risk of droughts is expected to increase over the 21 st century as a function of both climate and socio-economic drivers. These can be managed most effectively through a portfolio of adaptation measures. Measures considered here include: (i) management of current and future demand by retrofitting old building stock with water efficiency devices, imposing high building standards for water efficiency on developers, water use education programmes, greywater recycling, metering and rainwater harvesting [Measures equate to reducing total demand year on year to 34% by 2100 to 110l/person/day corresponding to mid-range guidance for sustainable homes xxi ], (ii) improving the efficiency of water distribution by patching and replacing pipes to reduce leakage [Reducing leakage year on year up to 40% by 2100], (iii) augmenting the storage capacity of the system with reservoirs [Total storage volume: 300,000Ml] and (iv) injecting new capacity into the water resource system through measures such as desalination, effluent reuse, inter-basin water transfer and lorry or ship transfer [Total potential capacity: 300Ml/day]. Other measures not considered here include the alteration of reservoir control rules and abstraction licenses, introduction of a water market, changed agriculture policies, groundwater recharge and new technologies. 3. Data, data sources The following data sources were used: ONS regional economic and demographic statistics UKCP2009 climate scenarios Ward level data for London on employment distribution and transport accessibility DTI Energy inventory DECC Regional energy statistics 8

9 4. Links with policy makers/urban planners A series of workshops and meetings were held with stakeholders in London, including the Greater London Authority, Transport for London, the Environment Agency and Thames Water. The objective of this process was to understand the problems facing decision makers in London and demonstrate how the UIAF can help to analyse solutions. 5. Results of the case studies No single policy will enable cities to grow whilst reducing emissions and vulnerability to climate change impacts a portfolio of measures is required. Due to long lead times, immediate and in some instances radical action to reduce fossil fuel dependence in the energy, building and transport sectors is required if an 80% cut in emissions is to be achieved by Measures to reduce demand (in use of energy, transport, water etc.) tend to be more cost effective and less likely to have adverse impacts in other sectors than measures taken to increase supply. However, both supply and demand side measures will be required to respond adequately to the climate and socio-economic changes. Economic drivers of long term change: A multisectoral regional economic model has been used to generate long term projections of employment and Gross Value Added in London. Our base line simulation shows employment in London growing by about 800,000 by 2030, driven by demographic changes and changing working practices. Business and financial services, along with science-based services are expected to grow most rapidly, with heavy industry diminishing. Land use change: Future patterns of land use between now and 2100 have been simulated for all of London and the Thames Gateway. The new land use model simulates the effects of changes in employment, the transport network and land use planning policy. We have simulated four alternative land use futures for London: (i) a baseline case, which applied current policies and trends in to the future (ii) Eastern axis in which employment opportunities, transport infrastructure development and a preference for lower density living stimulate substantial population growth in east London and the Thames Gateway (iii) Centralisation in which employment and population growth is concentrated in central London, with a corresponding increase in density (iv) Suburbanisation in which employment remains strong in central London, but expands into the suburbs, focused on existing hubs (e.g. Croydon). To steer land use change away from the baseline towards alternative futures requires major shifts in land use planning, transport connectivity and capacity, and employment opportunities. Carbon dioxide emissions: Various scenarios of carbon dioxide emissions from the energy use, personal transport and freight transport have been analysed. Growth in population, economic activity and mobility are potentially strong upward drivers of emissions. We have analysed portfolios of emissions reduction policies that are currently under consideration, but find that more radical policies are required in order to meet the GLA s target for 60% emissions reductions by Their success depends upon the availability of 9

10 carbon neutral electricity supply and upon progressive physical changes to urban form and function. Heat waves: A new land surface scheme has been introduced into the Hadley Centre s Regional Climate Model in order to represent the urban heat island effect. Using a weather generator adapted from UKCP09 we found that by the 2050s, one third of London s summer may exceed the current Met Office heat wave temperature threshold. We have analysed the potential for different spatial patterns of development to reduce the risk from heat waves. Droughts: The UKCP09 rainfall scenarios for the Thames and Lee catchments were combined with catchment hydrology models and simulation of the water resource management system. London is very vulnerable to changes in the surface water regime, which will be increasingly stressed by climate change and population growth. Although new storage facilities can maximise exploitation of the surface water resource, on their own they are insufficient in the long term and will need to be accompanied by vigorous demand management and provision of new resources from desalination or inter-basin transfers. Flooding: A model of flooding in the tidal Thames floodplain, which is protected by the Thames Barrier and a system of flood defences, has been used to simulate the effects of sea level rise and changing flows in the river Thames. This has been combined with our simulations of land use changes, which have a profound effect on the magnitude of increase in flood risk in the future. The Eastern Axis land use scenario leads to a fourfold increase in flood risk by 2100, whilst the risk doubles for the Suburbanisation scenario. We have tested the effectiveness of various options to improve flood defences and enhance resilience to flooding when it occurs. In summary, urbanisation and dynamics of urban form were found to be a driver of heat island effects, water resource pressures and flood risk. CO2 emissions were determined by transportation and population distribution. There is a need to consider the climate impacts, water demand and flooding risk of developments in the Thames Gateway. The UIAF enabled resilience assessments to be undertaken and plans to reduce flood risk were proposed. Important connections between disciplines were found that determined the results. CO2 emissions were determined by transport and population distribution. Water demand was determined by water resource availability and urban spacial development. Flood risk was determined by spacial pattern of development. Climate scenarios were determinants of long run temperature change and flood risk. The heat island effect combines with temperature change due to climate change to determine the risk of heat waves. 6. Conclusions: Added value of IA for urban planning We have quantified the synergies and conflicts between adaptation to climate change and mitigation of carbon dioxide emissions, for example by examining the contribution that urban energy use makes to the urban heat island. We have used the UIAF to begin to understand 10

11 how policies can be devised that yield benefits in relation to a number of objectives and avoid undesirable side-effects. The research has demonstrated the central role of land use planning in guiding and constraining pathways to sustainable urban layout in the long term. Land use profoundly influences carbon dioxide emissions and vulnerability to climate change. It also constrains opportunities for innovations like sustainable urban drainage systems or local heat networks. Land use and infrastructure planning decisions can become locked in because of the way in which infrastructure shapes land use and the built environment, and vice versa. The research has demonstrated scenarios of how these interactions can operate over the 21st century on spatial scales from the whole city and beyond to individual neighbourhoods, providing tools for planners and infrastructure designers to assess the long term sustainability of plans and policies. Can these methods/results be applied to other cities? Yes, if the data is available. All these models rely on extensive statistical data sets. Also, consistent scenarios of climate change are required for the long run vulnerability assessments of temperature change and flood risk from sea level rise. i Dawson R.J., Jim W. Hall, Stuart L. Barr, Michael Batty, Abigail L. Bristow, Sebastian Carney, Athanasios S. Dagoumas, Stephen Evans, Alistair C. Ford, Clare M. Goodess, Colin Harpham, Helen M. Harwatt, Mark P. McCarthy, Jonathan Köhler, Miles R. Tight, Claire L. Walsh, Alberto M. Zanni (2009) A systems approach to integrated assessment of adaptation and mitigation in cities, Tyndall Centre Working Paper, Newcastle University. ii Dawson, R. J. (2011) Potential pitfalls on the transition to more sustainable cities and how they might be avoided, Carbon Management, 2(2): iii OECD (2010), Cities and Climate Change, OECD Publishing. ( iv Rosenzweig, C. Solecki, W. Hammer, S. A. and Mehrotra, S. (2010) Nature 467, (doi: /467909a) v ONS (2008) 2008-based Subnational Population Projections for England, Office for National Statistics ( link verified 18 th August 2011). vi UKCIP (2009) UK Climate Projections science report: Climate change projections, UK Climate Impacts Programme ISBN ( link verified 18 th August 2011) vii Barker, T. and Peterson, W. (eds.) (1987) The Cambridge Multisectoral Dynamic Model of the British Economy, Cambridge University Press, Cambridge. viii Kilsby, C. G. et al. (2007). A daily Weather Generator for use in climate change studies. Environmental Modelling and Software, 22: ix Harwatt, H., Tight, M.R. and Timms, P.M. (2011) Personal transport emissions within London: exploring policy scenarios and carbon reductions up to International Journal of Sustainable Transportation 5(5): x Department for Transport (2006) Transport Statistics for Great Britain 2005, The Stationary Office, London ( link verified 18 th August 2011). xi TfL and University of Westminster (2008) London Freight Data Report, Transport for London. xii Department of Environment, Food and Rural Affairs (2007) Passenger transport emissions factors: Methodology Paper. The Stationery Office, London. xiii Boulter P.G., Barlow R.J., McCrae I.S. and Latham S. (2009) Emisssions factors 2009: Final summary report, Published Project Report PPR361, TRL Limited. 11

12 xiv International Energy Agency (2008) Energy technology perspectives 2008: scenarios and strategies to Organisation of Economic Cooperation and Development. xv King, J. (2008) The King review of low-carbon cars. Part I: the potential for CO2 reduction. The stationary office, London. xvi Carney, S., Green, N., Wood, R. and Read, R. (2009) Greenhouse gas emissions inventories for 18 European regions, University of Manchester, Manchester ( accessed 13 th November 2011). xvii NHS (2011) Heatwave Plan for England, Department for Health. xviii Manning L. et al. (2009) Using probabilistic climate change information from a multimodel ensemble for water resources assessment, Water Resources Research, 45, W11411, doi: /2007wr xix Dawson, R.J. and Hall, J.W. (2006) Adaptive importance sampling for risk analysis of complex infrastructure systems, Proc. R. Soc. A, 462(2075): xx Penning-Rowsell, E.C. et al. (2005) The benefits of flood and coastal defence. Middlesex University Flood Hazard Research Centre, London, UK. xxi DCLG (2006) Code for Sustainable Homes: A step-change in sustainable home building practice, Department for Communities and Local Government. References Dawson, R. J. et al. (2008), Attribution of flood risk in urban areas, J. Hydroinformatics, 10(4): (doi: /hydro ). McCarthy M.P., Harpham C., Goodess C.M. and Jones P.D. (published online), Simulating climate change in UK cities using a regional climate model, HadRM3, Int. J. Climatology (doi: /joc.2402). UKCIP (2009) UK Climate Projections science report: Climate change projections, UK Climate Impacts Programme ISBN ( link verified 18 th August 2011) 12