Presentation on theme: "Impact of Climate Change on Road Infrastructure"— Presentation transcript:
1Impact of Climate Change on Road Infrastructure Mark Harvey
2Austroads project published in 2004 Austroads AP-R243/04 : Impact of Climate change on road infrastuctureDownloadable for free from Austroads website, with volume of appendices AP-R243/04AAlso downloadable for free from BITRE website (main volume only)
3Project aimsassess likely local effects of climate change for Australia for the next 100 years, based on the best scientific assessment currently availableassess the likely impacts on patterns of demography and industry, and hence on the demand for road infrastructureidentify the likely effects on existing road infrastructure and potential adaptation measures in road construction and maintenance, andreport on policy implications arising from the findings.Note: The project was not concerned with impacts of transport emissions on climate change.
4Project structureBITRE coordinated the project and prepared the Executive Summary, Introduction and the Policy Implications chapters.The CSIRO Division of Atmospheric Research ran its global climate change models to produce forecasts of climate on a grid of about 50 kilometres up to 2100.The resultant data was passed on to three consultants to assess its implications.
5Project structure Emissions A2 scenario developed from population, energy and economic modelsConcentrationsCO2, methane, sulphates, etc. from carbon cycle and chemistry modelsGlobal climate changeTemperature, rainfall, etc.(200 – 400 km horizontal resolution) from Atmospheric-Ocean Global Climate Change Model Mark 2Regional climate changeMountain & coastal effects, islands, extreme weather, surface properties, etc. (50 km horizontal resolution) from Conformal-Cubic modelImpactsPopulation, industry, road pavements, salinity (various consultants)IPCCCSIROARRB, Monash University, ABARE, BITRE
6Project structure continued The Monash University Centre for Population and Urban Research investigated the likely effects on population settlement patterns and demographics.ARRB Group used these population projections to forecast changes in road transport demand.calculated changes to an index of climate from the CSIRO dataroad demand and climatic indexes were together used in pavement deterioration models to predict the implications for pavement deterioration and maintenance expenditure needs.
7Project structure continued Australian Bureau of Agricultural and Resource Economics (ABARE) employed its hydrological–economic model of the Murray-Darling basin to forecast implications of climate change for salinity and agricultural production in the region, and related this to road infrastructure.Multi-disciplinary project involving experts from a range of fields.
8Project structure continued CSIRO: Regional climate change forecasts to 2100Monash University Centre for Population and Urban Research:Impacts on population projectionsABARE: Impacts on salinity and agriculture in Murray-Darling basinARRB Group: Impacts on demand for roadsARRB Group: Impacts on road pavementsBITRE: report and summary
9Not covered in the study Local flooding implicationsrequires a catchment hydrological model to predict flooding heights, durations and water velocities, andan area topology model to relate flood heights to local road infrastructure.Salinity and impacts of agricultural industries outside the Murray-Darling BasinThe CSIRO models do not forecast sea level rises or the likelihood of changes in storm activity.
10Emissions forecastsInternational Panel on Climate Change (IPCC) emissions scenario selected‘A2’ scenariohigh scenario chosen to provide strong contrast with current conditionspredicated on global population of 15 billion in 2100rate of CO2 release grows, increasing to nearly fourfold by 2100.
11IPCC emission scenarios A2 scenario (red line) used in this study.
12CSIRO Atmospheric-Ocean Global Climate Change Model global circulation model with atmospheric, oceanic, sea-ice and biospheric submodelsglobe divided up into a grid comprised of 300 km squares9 layers of atmosphere, each block having parameters such as temperature, air pressure, wind velocity, water vapour content12 layers of oceantime step of 30 minutesrun from 1870 to 2100suite of properties (temperature, moisture) saved for 6-hourly intervals for the 230 yearstook three months on a supercomputer
13CSIRO Conformal-Cubic General Circulation Model Results from global model used to ‘nudge’ more detailed model for Australiawind speeds outside Australia adjusted to make consistent between modelsgrid of about 50 km squares for Australia and lower resolution for rest of the world (up to about 800 km for the other side of the globe)
14Method of deriving detailed forecasts outputs: monthly means of average, maximum and minimum temperatures, precipitation, solar radiation, potential and actual evaporation for each grid pointconverted tolocal temperature change per degree of global warming for temperaturepercent for rainfall, radiation evaporation change per degree of global warmingused to derive forecast for any IPCC scenario for any grid point over the next 100 years.
15Key findings: temperatures average annual temperatures increase by 2⁰ to 6⁰C by 2100Tasmania coastal zones least affected, inland areas most affectedmore extremely hot days and fewer cold days, for exampleaverage number of summer days over 35⁰C in Melbourne to increase from 8 at present to by 2070average number of winter days below 0⁰C in Canberra to drop from 44 at present to 6-38 by 2070
16Average annual temperature: base (2000) and 2100 climate 5151719212324252627282930323440Base (2000) climate2100 climate
17Temperature changes: year 2100 relative to base climate
18Key findings: rainfall and evaporation general reduction in rainfall except for the far north where there will be significant increaseswhere average rainfall decreases, more droughtswhere average rainfall increases, more extremely wet yearsin the north, more intense tropical cyclones, more severe oceanic storm surges, more frequent and heavier downpoursevaporation to increase over most of the country adding to moisture stress on plants and drought
19Average annual rainfall: base (2000) and 2100 climate 152025303540506070100125200>500Base (2000) climate2100 climate
20% change in average annual rainfall 2000-2010 134-9-12-14-25
21Sea level rise Not predicted in CSIRO modelling. IPCC projects rise of 9 to 88 cm by 21000.8 to 8.0 cm per decade
22Impact on population and settlement patterns: methodology undertaken by Monash University Centre for Population and Urban Research (Dr Bob Birrell)population projections developed for Australia as a whole, States and major metropolises (based on ABS mid-range projections supplemented by ANU demographic projection software).adjustments made to the projections for the eight major metropolitan regions for climate changeusing expert judgement supported by a comfort index (function of temperature and humidity). A comfortable climate is a major driver of internal migration.
23Population base case: without climate change total fertility rate will fall to 1.6 and net overseas migration 90,000 per year over the 21st centurytotal population 19.1m in 2000 to 27.3m in 2100greater concentration in four major metropolises: Sydney, Melbourne, Brisbane and Perthother growth outside the four cities is in non-metropolitan Queensland and WA.significant increase in share of population in Queenslandfor planning purposes, need to take base case, then adjust for climate change
24Population: climate change impacts of the eight metropolitan regions assessed, only Darwin and Melbourne gain population from climate changeeven though hotter, wetter Darwin less attractive, higher rainfall should promote agricultural productionbut note the contrary view from the recent Northern Australia Land and Water Taskforce report: lack of suitable soils; high evaporation and lack of dam sites limits water storageLosers: Adelaide (water supply), Cairns (less attractive climate) and Perth (water and climate)coastal areas of NSW and Victoria more attractive climatehotter, drier climate in inland areas will may have adverse impact on agriculture
252100 population without and with climate effects Selected Statistical Division% of 2000 population without climate changeAdjustment factor with climate changeClimate change factors driving population changeSydney159%1.00temps higher but not expected to affect population growthMelbourne125%1.15temperatures higher resulting in more attractive climateBrisbane211%0.96temperatures higher resulting in less attractive climateMoreton305%0.98Adelaide63%0.79restricted water supply, especially in springPerth195%0.88less attractive climate; restricted water supplyDarwin275%1.34temps high but heavy rainfall drives increased agricultureACT93%Cairns279%0.83
26Note: Climate change is not the most important influence on population patterns. range of projections for 2100 compared with 2000without climate change: -63% Adelaide to 305% Darwinwith climate change: -50% Adelaide to 369% Darwin
27Impact on road demand: Methodology passenger and freight tasks considered separatelybase-case forecasts developedcars a function of population, per capita car ownershipfreight a function of population, per capita freight, average payload (trend to larger vehicles)converted to equivalent standard axel loads for pavement impactsARRB used a gravity model to estimate impacts of climate change on traffic.If population at A increases by 100a% and population at B by 100b% due to climate change, then traffic between them increases by 100[(1+a)(1+b)-1]%.
28Impact on road demand 2100: conclusions 60% additional traffic (total vehicles passengers and freight)dramatic increase in Queensland, moderate in Syd-Mel corridor, decline around Adelaide, increase in Perth urban only slight rise in Perth intercapital trafficproportion heavy freight vehicles will rise from 12.1 to 13.9%total road freight to rise by 112% from 2000 to 2100average payload to increase by 25%, most in next decadeequivalent standard axles per articulated truck to double due to higher mass limitstotal ESA-kms on National Highway to rise by 230%due to freight growth, higher mass limits and payloads
29Impact on pavement performance: methodology climate represented by ‘Thornthwaite moisture index’a function of precipitation, temperature and potential evapo-transpiration. Index varies from +100 on Cape Yorke Peninsula to -50 in central Australia.used a National Highway System road databasepavement models estimate present value of life-cycle road agency costs (maintenance and rehabilitation) and road-user costs (travel time and vehicle operation)select treatment options and timings to minimise present value of costs subject to specified constraints on maximum roughness and annual agency budgets.
30Pavement Life Cycle Costing (PLCC) model 60 year analysis period, 7% real discount rateNational Highway System split into 60 sections with similar climate characteristics, traffic levels, vehicle mixes and pavement characteristicspavement deterioration a function of pavement age, cumulative equivalent standard axle loads, Thornthwaite index, and average annual maintenance expenditure.
31HDM4 modelmuch more detailed pavement deterioration algorithm covering roughness, rutting, cracking, potholing, ravelling, strength etc and consequently much more detailed data requirementscase studies of 8 road segments analysed in detailone segment from each state and territory located in or near a metropolitan areadata inputs that vary with climate are site-specific changes in Thornthwaite index, traffic levels and per cent heavy vehicles
32Other points to note Note: only trucks cause pavement wear, not cars but cars impact on the models because increased roughness adds to road user costs for carsLimitationseffects of floods, severe storms and sea-level rise not taken into accountno allowance for expansion of lane-kilometresdesign pavement strengths assumed to remain unchangedroad agencies may not minimise present value of costs due to budget constraints causing maintenance to be deferred and higher than economically warranted maintenance standards in some areas for social and equity reasons
33Thornthwaite moisture index: base (2000) and 2100 climates -45-30-1520406080>100Base (2000) climate2100 climate
34Changes in Thornthwaite moisture index: 2000 to 2100
35Changes to Thornthwaite index tendency to a drier climate overall (negative change in Thornthwaite Index)central area of Australia relatively unchangedlocalised areas where the changes are greatest includesouth-west of Western Australianorth-east Victoria and southern NSWsouth-west Tasmaniatop-end Queensland.
37Comparing optimal road agency costs comparison is not with and without climate changebut 2000 traffic volumes and climatewith 2100 climate-adjusted traffic volumes and climateNorthern Territory and Queensland experience large increasesprimarily due to population growth but wetter climate contributes.South Australia declines due to smaller population and drier climate.
38Maintenance: rehabilitation funding split maintenance = routine and periodic maintenance (pothole patching, kerb and channel cleaning, surface correction, resealing)rehabilitation = chipseal resheeting, asphalt overlays, stabilisation, pavement reconstructionfor Australia as a whole, no predicted change in 35:65 splitrehabilitation proportion to rise (maintenance proportion to fall) significantly for Tasmaniaconverse for WAreflects differences in pavement age distributions and life times
39HDM4 results: road agency costs Base climate2100 traffic changes only2100 climate & traffic changesCost ($'000)Change (3/1)Change (5/1)Col number12345ACT97.80%NSW46.272.757%72.858%NT176.6176.9177.1QLD83.6103.123%106.127%SA9996.8-2%TAS140.2159.514%159.4VIC128.7175.536%177.338%WA205.9244.519%244.1undiscounted total costs per kilometre over 20-year periodVirtually all the changes are from population growth leading to traffic increases, not climate change.
40Impact on salinity in the Murray-Darling Basin: methodology ABARE Salinity and Land-use Simulation Analysis (SALSA) modelnetwork of land management units linked through overland and ground water flowshydrological: rainfall, evapo-transpiration, surface water runoff, irrigation, ground water recharge/discharge rates, salt accumulation in streams and soilclimate projections incorporated by changing rainfall and evapo-transpirationrate of flow of groundwater depends on ‘hydrolic gradients’very flat in lower parts of the catchment
41Impact on salinity in the Murray-Darling Basin: methodology continued land-use allocated to maximise economic return from use of agricultural land and irrigation waterrelationship between yield loss and salinity for each agricultural activityland-use can shift with changes in salinity and water availabilitycosts of salinity measured as reduction in economic returns
42Catchments in the Murray Darling Basin covered by the SALSA model
43Impact on salinity in the Murray-Darling Basin: Key findings UnitsBase scenarioWithout climate changeWith climate changeYear20002100Net production revenue$m, npv382737183400Area of high water tables‘000 ha113753414404SALT CONCENTRATIONDarling – below the Macquariemg/L152277483Murray – below the Murrumbidgee141181198Murray – below the Darling226301343Murray – at Morgan313445548SURFACE WATER FLOWSDarling – below the Macquarie confluenceGL734577846060Murray – below the Murrumbidgee confluence812890405259678977204435
44Impact on salinity in the Murray-Darling Basin: comparisons area affected by high water tablesbase case rise: from 1.1m hectares in 2000 to 5.3m hectares in 2100climate change: rise to 4.4m hectares in 2100climate change mitigates salinity problems but nowhere near sufficient to reverse the rising trendnet production revenuebase case: falls by 3% due to high water tables, shift from pasture to croppingwith climate change: falls by 11% due to reduced surface water flows, switching from irrigated to dryland activitiesless demand for road transport
45Impact on salinity in the Murray-Darling Basin: comparisons higher water tables are bad for road pavements, but this is a problem in both the base and climate change scenariosslightly less with climate changereduced surface water flows make salt concentrations higher in rivers which reduces yields from irrigated productionand rusts steel reinforcing in concrete structures in riverine environments such as bridges and culverts.
46Summing up: uncertainty high level of uncertainty aboutIPCC emissions forecastsCSIRO estimates of climate impactsconsultants’ forecastsuncertainties built upon uncertaintiesnumerical results are broad indicators that tell a story
47Summing up: demand for roads Higher car and truck traffic from population growth is the main driver of investment and maintenance needs for roads.Large changes are forecast without climate change.strong growth for SE Queensland, Cairns, Darwin, Brisbane, Sydney, Melbourne, Perthdecline for Adelaide and inland areasClimate change adds to forecasts for Darwin and Melbourne and reduces forecasts for Adelaide, Perth and Cairns.
48Summing up: road design and maintenance less rainfall should slow pavement deteriorationbut effects so small as to have negligible impact on costsexception for far northern parts of Australia, which are forecast to become wetter. Capacities of culverts and waterways may prove inadequate.sea-level rise a concern for low-lying roads in coastal areaschanged and frequencies of floods in some areasrequires modelling of individual catchments to forecast impacts
49Overall conclusionChanges affecting road infrastructure will occur regardless of climate change.Climate change is just another factor in the mix, and usually not the most important.The main impacts on road infrastructure may come from changes in flood heights and frequencies, and sea-level rise with storm surges, which were not addressed in detail in the project.impacts vary greatly between locations
50Subsequent research: ARRB: Climate change framework for Queensland Department of Main Roads in 2008 report not published, but a summary is available in a conference paper by Evans, Tsolakis and Naude (ATRF Conference 2009)comprehensive list of potential impacts on road infrastructure and operationsdetailed review of (short- and long-term) climate change forecasts for Queenslandframework to assess risks, and to assist in the planning of climate change mitigation and adaptation responses.
51ARRB Framework Four impacts relevant to Queensland temperature changes (increases in very hot days)rainfall changes (reductions and increases) and floodingrising sea levels with storm surgesincrease in cyclone frequency and intensity.Phases of framework to identify investment prioritiesidentify climate change effectsgeographic scale, certainty, timeframedetermine impacts on transportadaptation strategiesplanning and project evaluation
52Other research underway Austroads project: ‘Impact of climate change on road performance’, undertaken by ARRBsoftware to provide climate information from 1960 to 2099 by GPS coordinates based on CSIRO modellingminimum and maximum daily temperatures, rainfall, Thornthwaite moisture indexpre-2007 based on historical meteorological dataClimate Futures Tasmania Infrastructure projectWorld Road Association (PIARC) Technical Committees: C.3 (natural disasters), D.2 (road pavements), D.3 (bridges), D.4 (geotechnics and unpaved roads)all have working groups on adaptation to climate change