* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Development of a Methodology and a Tool for the Assessment of Vulnerability of Roadways to Flood Induced Damage
Global warming controversy wikipedia , lookup
Heaven and Earth (book) wikipedia , lookup
ExxonMobil climate change controversy wikipedia , lookup
Fred Singer wikipedia , lookup
Michael E. Mann wikipedia , lookup
Soon and Baliunas controversy wikipedia , lookup
Climate resilience wikipedia , lookup
Atmospheric model wikipedia , lookup
Politics of global warming wikipedia , lookup
Climatic Research Unit email controversy wikipedia , lookup
Global warming hiatus wikipedia , lookup
Climate change denial wikipedia , lookup
Climate engineering wikipedia , lookup
Climate governance wikipedia , lookup
Citizens' Climate Lobby wikipedia , lookup
Global warming wikipedia , lookup
Climate change adaptation wikipedia , lookup
Economics of global warming wikipedia , lookup
Climatic Research Unit documents wikipedia , lookup
Climate change feedback wikipedia , lookup
Climate sensitivity wikipedia , lookup
Media coverage of global warming wikipedia , lookup
Effects of global warming on human health wikipedia , lookup
Solar radiation management wikipedia , lookup
Public opinion on global warming wikipedia , lookup
Climate change in Saskatchewan wikipedia , lookup
Attribution of recent climate change wikipedia , lookup
Scientific opinion on climate change wikipedia , lookup
Physical impacts of climate change wikipedia , lookup
Global Energy and Water Cycle Experiment wikipedia , lookup
Climate change and agriculture wikipedia , lookup
Effects of global warming wikipedia , lookup
General circulation model wikipedia , lookup
Climate change in the United States wikipedia , lookup
Surveys of scientists' views on climate change wikipedia , lookup
Climate change in Tuvalu wikipedia , lookup
Years of Living Dangerously wikipedia , lookup
Instrumental temperature record wikipedia , lookup
Climate change and poverty wikipedia , lookup
Effects of global warming on humans wikipedia , lookup
Use of System Dynamics to Understand the Long Term Impact of Climate Change on Pavement Performance and Maintenance Cost Rajib B. Mallick, PhD, PE Professor Civil and Environmental Engineering Department Worcester Polytechnic Institute (WPI) Worcester, MA 01609 Phone: (508) 831 5289 Fax: (508) 831 808 Email: [email protected] Michael J. Radzicki, Ph.D. Associate Professor of Economics Department of Social Science & Policy Studies Worcester Polytechnic Institute Worcester, MA 01609-2283 Phone: (508) 831-5767 Fax: (508) 831 5896 Email: [email protected] Jo Sias Daniel, Ph.D., P.E. Professor W171 Kingsbury Hall 33 Academic Way Department of Civil Engineering University of New Hampshire Durham, NH 03824 Phone: (603) 862-3277 Fax: (603) 862-2364 Email: [email protected] Jennifer Jacobs Professor Department of Civil Engineering University of New Hampshire Durham, NH 03824 Phone: (603) 862 0625 Fax: (603) 862 3957 Email: [email protected] Paper prepared for consideration for presentation at the 2014 Annual Meeting of the Transportation Research Board (TRB) and Publication in the TRB Research Record: Journal of Transportation Research Board Total number of words: 6202+ 1 Table+7 Figures = 8,202 Use of System Dynamics to Understand the Long Term Impact of Climate Change on Pavement Performance and Maintenance Cost ABSTRACT An increase in the vulnerability of the nation’s roadway network to changes in climatic conditions has become an issue of significant concern. Heavy rainfall and high temperatures are just a few examples of climatic factors that can affect pavement performance and conditions in a very detrimental way. Some of these effects, specifically on costs associated with constructing and maintaining roads, may not be significant in the short term, but could become very significant in the long term. The objective of this study was to present a framework to utilize system dynamics to understand the long term impact of climate change on pavement performance and maintenance activity. A system dynamics model was created with available pavement performance and climate change data, and the effects on various key factors such as pavement life and maintenance cost were evaluated through simulation. Preliminary results show that the long term effects of changes in air temperature, rainfall, sea water level rise and number of hurricanes on pavement performance are significant, and that costs are expected to increase very significantly (>160% in 100 years) and non-linearly, in the future. Based on these findings, recommendations are made for obtaining more accurate and reliable data on relevant climate change factors and for using a system dynamics approach to integrate the multidisciplinary topics of climate change, pavement design and performance, and economics into comprehensive studies. 1 INTRODUCTION Research has clearly shown, and continues to show, that the nation is becoming increasingly vulnerable to devastating damage to the infrastructure as a result of changing global climatic patterns – for example floods associated with hurricanes of increasing frequency and rise in sea water and ground water level (1, 2, 3, 4, 5, 6). The Intergovernmental Panel on Climate Change (IPCC) has stated that the frequencies of heavy precipitation, as well as rainfall from tropical cyclones, are likely to continue to increase in this century (7). Evidence from three recent hurricanes, Katrina, Irene and Sandy, in the US supports the research findings very strongly; the resulting floods would affect much of the coast and millions of citizens. As projected by the IPCC, rising sea levels is virtually certain (>99% probability of occurrence) to continue, and increases in intense precipitation events are highly likely (>90% probability of occurrence) to become more frequent in widespread areas of the United States (8). As a consequence, along coastal areas and low-lying river areas flooding will be expected to occur more frequently, such as the Midwest flooding along the Mississippi river (9). Pavements constitute the most widely used part of the nations’ infrastructure for the transportation of people and goods. An increase in pavement temperature as a result of rising air temperature, reduction in subgrade and hot mix asphalt (HMA) modulus due to high rainfall (saturation), and flooding due to rise in sea water level (in coastal areas) and hurricanes can inflict significant damage to pavements, making them unserviceable – causing a major negative impact on the nation’s economy. The ingress of moisture in pavements due to flooding can also accelerate the onset of other types of non-water related distress such as rutting and cracking. The impact of change in climate on pavement performance has been investigated by a few researchers in the past (10, 11, 12) and most recently, Meagher et al. (13) presented a methodology to assess the impact of future climate change on pavement deterioration. Instead of using traditional historic (stationary) data sets, they conducted a mechanistic-empirical analysis of pavement structures (designed for four sites across New England) with predicted future temperatures, and evaluated their effects on the rutting performance of the pavements. The effect on the rutting potential of the HMA layer was found to be significant. The authors demonstrated and proposed the concept of downscaling regional climatic model (RCM) temperature datasets to create modified hourly climatic files in the mechanistic empirical pavement design guide/software (MEPDG/S). Increasing damage in pavements will definitely lead to more rapid deterioration of pavements, reduction in pavement lives, and consequent increases in maintenance cost. The U.S. spends nearly $200,000,000 per day building and rebuilding roads and bridges right now; more rapid deterioration as a result of climate change will increase this number. Already overburdened pavement maintenance and repair budgets will be stressed even further and the changing climate will likely require new approaches to pavement design, maintenance and management. This paper presents a framework for a methodology that can be used to answer one of the most relevant questions for the development of new ways to manage and maintain pavements: what will be the impact of future climate on pavement life and maintenance costs? 2 System Dynamics System dynamics is an approach that helps develop a strategic view of a “system”, which could be an industry, society or a nation, by modelling the different parts and simulating the dynamics of interaction among the different parts. This helps to determine the changes over time, and hence develop a view which could not be obtained from “spot” studies conducted with either a few of the critical elements of the system or within the confines of a specific time (second, hour, day or year) time. One of the most powerful elements of this approach is the ability to “link” elements and model the interdependencies of the various elements across disciplines (for example climatic science and civil engineering). Known as causal (or feedback) loops, these links help to understand the dynamic nature of a problem, and simulate the systems over time. The important aspect of simulating “over time” is that whereas impacts (such as that of unsustained growth in population or construction) may appear to be linear over a short time period (say a span of five years), in reality, they may be of exponential nature over a decade or a few decades. Having the ability to view this change over a long time period is essential for developing policies for an industry or society or a nation and to make sure that the far reaching consequences of adoption of these policies are indeed beneficial in the long run. This ability can help differentiate good policies from bad policies – policies may appear to be “good” in the short term, but in the long term may have disastrous consequences, and only a proper system dynamics model can capture it. Another key aspect of the system dynamics approach is the ability to show the root cause (or causes) of a problem. This is because a good system dynamics model can include all of the essential elements and their interdependencies and hence, the relational dynamics. The steps in the modelling of system dynamics consist of identifying and defining a problem, developing a dynamic hypothesis, modelling, simulating, and developing policies and evaluating them. The components of a typical system dynamics model are stock, flow and connectors. Stock represents a commodity that can increase or decrease in value or number or quantity. Flow is the rate of increase or decrease in the stock, which can be influenced by “connectors”. A simple example is that if the amount of available natural aggregate can be considered to be a stock, the amount of use of aggregate in tons per km of a road can be considered to be a connector, that dictates the rate of flow, which can be the same as the rate of use of aggregates or can also be influenced particularly by the amount of recycled aggregates. System dynamics has been utilized since the nineteen sixties in modeling and simulation of a wide range of problems, including social, industrial, sustainability, food, population growth and natural resources. Forrester (1971, 14) introduced the subject through his sustainability model on world dynamics, and was followed by Meadows (1972, 15) who presented the limits to growth model. This study has been updated twice since then (16, 17). A model to illustrate the unsustainability of our civilization was presented by Sterman (2012, 18), which includes growth, carrying capacity and technology, as well as the delays in regeneration of resource, development of technologies and adoption of policies. The carrying capacity of the earth system and detrimental effect of unsustainable growth on the environment has also been investigated with system dynamics modeled by Radzicki (1995, 19), Bueno (2011, 20) and Bockernmann et al (2005, 21). System dynamics has been used to investigate the relationship between quality of life and sustainability by Beck (2011, 22, 2012, 23), whereas Grosskurth (2007, 24) has looked at a 3 case study on regional sustainability in Netherlands. Predictive models for the forecast of fossil fuel predictions have been developed by Maggio and Cacciola (2012, 25), whereas the link between resource productivity, consumption and sustainability has been investigated by Schmidt-Bleek (2008, 26). Objective The objective of this study is to develop a framework or methodology to utilize system dynamics to understand the long term impact of climate change on pavement performance and maintenance activity. MODELING AND SIMULATION The overall pavement analysis approach and its consideration in the model are shown in Figure 1. A system dynamics model linking climatic changes to pavement maintenance and costs was developed, as shown in Figure 2. The “connections”, in the form of equations, and the parameters are explained in Table 1. The increase in cost of maintenance was determined on the basis of increase in thickness of HMA that is required to maintain the deteriorated roads. Since the “paving for maintenance” depended on the fraction of the “roads needing maintenance”, as the latter increases, the cost should also increase. However, note that here the term maintenance has been applied for both maintenance and rehabilitation, since here maintenance is assumed to be applied for roads that have exhausted their structural performance “lives” as predicted by the MEPDG/S. There is scope of improvement in this analysis (beyond the scope of this paper) by differentiating between surface distress related maintenance and structural rehabilitation. In this case, the rate of use of HMA per two lane-km has been calculated on the basis of an average of 75 mm (between 50 mm of maintenance overlay and 100 mm of rehabilitation layer) thick HMA, with a density of 2,355 kg/m3. This paper uses a single pavement cross section with rutting as the primary distress to illustrate the power of using a system dynamics approach. A two-lane highway, which lies along the coast of New Hampshire, was selected as the model road. To represent changes in climate, four factors were considered: 1. Rise in air temperatures, causing a rise in pavement temperature, and hence decreasing modulus of HMA; 2. Increase in average annual rainfall, which affects duration for which the subgrade soil can be expected to be at or close to saturation, and hence affects the modulus of the subgrade soil; 3. Increase in sea water level and; 4. Increase in number of Category 3 hurricanes, both of which are expected to increase the number of inundation (flooding) occurrences, and hence cause a lowering of the modulus of HMA and subgrade. The single pavement structure was analyzed for a range of temperatures and subgrade moduli values, to evaluate the effect of these climate changes on its rutting life (that is, years to failure by rutting). 4 Change in Sea Water Level Change in Maximum Air Temperature Change in Maximum Pavement Temperature Change in Number of Category 3 Hurricanes Change in Average Annual Rainfall Change in Number of Inundations Change in Months with 100% or Close to 100% Saturation of Subgrade Soil Temperature Adjusted Minimum HMA Modulus Effective Inundation plus Temperature adjusted Minimum Modulus Effective Subgrade Resilient Modulus MEPDG/S Analysis Rutting Life of a Pavement=f(Subgrade modulus and HMA Modulus) Deterioration Rate Considered in the System Dynamics Model Figure 1. Pavement analysis approach. Note: Data in filled text boxes are from climatic data analysis, rest are from pavement related literature and/or MEPDG/S analysis 5 Legend Figure 2. System Dynamics Model 6 Table 1. Equations Used In The System Dynamics Model (explanations within parentheses) ______________________________________________________________________________ Hot_Mix_Asphalt(t) = Hot_Mix_Asphalt(t - dt) + (-Hot_Mix_Asphalt_Use_Rate) * dt INIT Hot_Mix_Asphalt = 1e+20 tons (this number has been chosen arbitrarily) OUTFLOWS: Hot_Mix_Asphalt_Use_Rate = Annual_Km_Paved*Hot_Mix_Asphalt_Use_Per_Km Number_of_Inundations_per_Year(t) = Number_of_Inundations_per_Year(t - dt) + (Increase_in_Number_of_Inundations_per_Year) * dt INIT Number_of_Inundations_per_Year = 0.267+1+0.150 INFLOWS: Increase_in_Number_of_Inundations_per_Year = 10.695*Rate_of_Change_in_Sea_Water_Level+1*Rate_of_Increase_in_Number_of_Flood_Cau sing_Hurricanes Roads_Needing_Maintenance(t) = Roads_Needing_Maintenance(t - dt) + (Deterioration Paving_for_Maintenance) * dt INIT Roads_Needing_Maintenance = 50 (this number has been chosen arbitrarily) INFLOWS: Deterioration = New_Roads/Average_Pavement_Life_Adjustor OUTFLOWS: Paving_for_Maintenance = Roads_Needing_Maintenance*Fraction_of_Roads_Maintained Maximum_Pavement_Temperature(t) = Maximum_Pavement_Temperature(t - dt) + (Temperature_Increase) * dt INIT Maximum_Pavement_Temperature = 58 (maximum temperature at project location) INFLOWS: Temperature_Increase = 0.78*Rate_of_Change_in_Maximum_AirTemperature Months_with_100%_Saturation(t) = Months_with_100%_Saturation(t - dt) + (Increase_in_100%_Saturation_Months) * dt INIT Months_with_100%_Saturation = 2 INFLOWS: Increase_in_100%_Saturation_Months = (0.0071*Rate_of_Change_in_Average_Annual_Rainfall) New_Roads(t) = New_Roads(t - dt) + (Paving_of_Rehabilitated_Road + Paving_for_Maintenance - Deterioration) * dt INIT New_Roads = 0 INFLOWS: Paving_of_Rehabilitated_Road = Authorized_Rehabilitations_Kms_Per_Year Paving_for_Maintenance = Roads_Needing_Maintenance*Fraction_of_Roads_Maintained OUTFLOWS: Deterioration = New_Roads/Average_Pavement_Life_Adjustor Annual_Km_Paved = Paving_of_Rehabilitated_Road+Paving_for_Maintenance Authorized_Rehabilitations_Kms_Per_Year = 2 ______________________________________________________________________________ Note: Each “Stock” has an inflow and in some cases, an outflow as well. These inflows and outflows are controlled by “Flows”, and both Stock and Flows can be affected by/can affect “Converters”; INIT – Initial; Climatic Factors are explained in the next section 7 Table 1. Equations Used In The System Dynamics Model (explanations within parentheses) (Continued) ______________________________________________________________________________ Average_Pavement_Life = 25.574+0.000551*Effective_Resilient_Modulus_of_Subgrade_Soil+0.0000452*Effective_Inundation_pl us_Temperature_Adjusted_Modulus_of_HMA (based on regression equations) Average_Pavement_Life_Adjustor = If (Average_Pavement_Life>1) THEN (Average_Pavement_Life) ELSE (1) Effective_Inundation_plus_Temperature_Adjusted_Modulus_of_HMA = Temperature_Adjusted_Minimum_HMA_Modulus0.03*Number_of_Inundations_per_Year*INIT(Temperature_Adjusted_Minimum_HMA_Modulus) Effective_Resilient_Modulus_of_Subgrade_Soil = 26,500*((12Months_with_100%_Saturation_Adjustor_for_Maximum_of_8_months)/12)+8,690*(Months_with_100 %_Saturation_Adjustor_for_Maximum_of_8_months/12) Fraction_of_Roads_Maintained = 0.1 Hot_Mix_Asphalt_Use_Per_Km = 636 tons (for average of 50 mm maintenance and 100 mm rehabilitation layers) Increase_in_Cost_of_Road_Construction_and_Maintenance_% = 100*(Hot_Mix_Asphalt_Use_RateINIT(Hot_Mix_Asphalt_Use_Rate))/INIT(Hot_Mix_Asphalt_Use_Rate) Months_with_100%_Saturation_Adjustor_for_Maximum_of_8_months = IF(Months_with_100%_Saturation<=8)THEN(Months_with_100%_Saturation)ELSE(8) Rate_of_Change_in_Average_Annual_Rainfall = 2.8189 mm Rate_of_Change_in_Maximum_AirTemperature = 0.024C Rate_of_Change_in_Sea_Water_Level = 0.0093 m Rate_of_Increase_in_Number_of_Flood_Causing_Hurricanes = 0.000556 Temperature_Adjusted_Minimum_HMA_Modulus = 158,800316012*Maximum_Pavement_Temperature ______________________________________________________________________________ Note: Each “Stock” has an inflow and in some cases, an outflow as well. These inflows and outflows are controlled by “Flows”, and both Stock and Flows can be affected by/can affect “Converters”; INIT – Initial; Climatic Factors are explained in the next section 8 Climatic factors The climatic data used in this model were obtained from the USGS (27) hosted at the http://regclim.coas.oregonstate.edu (accessed on 7/14/2013). The climate data include two 30year periods that were analyzed, 1970 to 1999 and 2040 to 2069. The GFDL CM2.0 general circulation model from the U.S. Department of Commerce/National Oceanic and Atmospheric (NOAA) Geophysical Fluid Dynamics Laboratory (GFDL) USA was dynamically downscaled using the RegCM3 regional climate model. RegCM3 is a high-resolution atmospheric model coupled with a physically based model of surface processes. The atmospheric composition in the RegCM3 simulations match the 20th century and A2 scenario time series that were developed for the IPCC AR4 (Intergovernmental Panel on Climate Change, 2007) and used in the driving GCMs. A grid cell located in Massachusetts was used for the analysis (Lat: 42.12946 N, long: 72.10891 W). Note that in the model, two parameters were used for the variables that are dependent on the climatic factors: an initial value and a rate of change per unit change in the specific climatic factor. Temperature: The daily average air temperature record (average and maximum air temperature) for two 30-year periods were analyzed, 1970 to 1999 and 2040 to 2069. The 30-yr average value for each day of year was determined for both periods. Average daily temperature differences were calculated for each day of year then averaged for the entire year resulting in an average temperature change projected to occur in 70 years. A linear increase over time was assumed. On the basis of these data an increase in maximum air temperature of 0.024oC per year was predicted and utilized in the maximum pavement temperature versus maximum air temperatures model (LTTP model) to estimate a coefficient of change in the maximum pavement temperature per degree change in the maximum air temperature. In this initial systems analysis, the projected climate is highly simplified. A single location with one climate model, one emissions scenario, and a constant rate of change was used to provide estimates of climate data. Extending beyond a single model and scenario is strongly recommended because there is considerable uncertainty in the climate factors. Sources of uncertainty result from unknown relationships between climate and atmospheric concentrations, seasonal differences, difference among model physics and parameterization, and future carbon emission which depend on human activities. Relative to recent regional studies that have evaluated ensembles of models under different emission scenarios, the values used in this study are not unreasonable. The temperature increases are quite similar to the ensemble average values but the precipitation values are somewhat higher than those found by Hayhoe et al. (28). Hayhoe et al. provide insight to the uncertainty that might be expected in the northeastern U.S. They found approximately a 20% increase in temperature for the higher A1F1 scenario and a 15% decrease in temperature for the lower B1 emission scenario. The annual temperature change differed across the nine models reviewed by up to 75%. Furthermore, projected temperature and precipitation changes differ by location. For example, the CMIP 3 projected temperature increases are more modest across the southeastern United States and western coast as compared to other regions (29). Precipitation is projected to increase across the northern U.S., but decrease in the Southwest. Rainfall: An increase in average annual rainfall of 2.8 mm per year was predicted using the same approach as above. Using these rainfall data, soil moisture was modeled using a onedimensional bucket style model (30) to predict the change in the number of times the soil could 9 be expected to be at or close to saturation. The bucket model is a simple model for soil moisture dynamics, which includes pavement wetting, infiltration into the pavement structure, and drainage to field capacity. The model was forced using daily precipitation data for the current and future 30-yr study periods for a 1 m thick subgrade consisting of silty loam soil. Daily soil moisture was used to estimate an increase in the number of days (and hence months) at which the soil will be close to saturation. Sea water level: Sea level rise values were estimated from Kirshen et al.’s (2010,14) study that determined local Stillwater Elevations (SWEL) for Portsmouth, NH. These local SWEL values were based on data obtained from the Fort Point station located in Newcastle, NH and from methodology included in the published work entitled “Coastal flooding in the Northeastern United States due to climate change” (31). An increase in sea water level of 0.0093 m per year was estimated. While inundation will vary by specific location, based on the available data, a site having a current inundation of once every year in the future might reasonably be expected to be inundated approximately 5 times annually (during astronomical high tides) for the 40 year future data period. The change in inundation per m rise in sea water level was determined. Number of category 3 hurricanes: On the basis of currently available historical hurricane data (http://www.aoml.noaa.gov/hrd/hurdat/) the number of Category 3 or higher hurricanes was estimated as 0.15 per year for the current 30 year period. Future 30 year period were estimated to be 0.167 per year, using Done et al.’s (2012, 32) estimated 10 to 15% increase in the frequency of high-impact events in future climates. For each occurrence of the hurricane, an occurrence of inundation was considered for the pavement. Pavement properties Two critical pavement properties were considered for assessing the impact of the changes in climatic factors: resilient modulus of the subgrade soil and HMA. An effective modulus of the subgrade soil was calculated, considering the months at which the soils could be considered to be dry and those at which the soil could be considered to be fully saturated. A 100% modulus was assumed for the soil in the dry months, whereas for the saturated months, a modulus of 50% of that at dry condition was considered, on the basis of data and recommendations given in NCHRP, 2004, Appendix DD-1:Resilient Modulus As Function Of Soil Moisture-Summary of Predictive Models (33). Finally, a weighted average modulus was calculated on the basis of the moduli and the corresponding months of saturation. Using the MEPDG, the minimum surface HMA moduli were determined for cites with different high temperatures. A linear model relating maximum pavement temperature (calculated, as described earlier, from maximum air temperature) to minimum modulus was determined. This modulus was then adjusted for the effect of water through inundation. Note that this may not be an accurate approach, but it has been used here as a practical approach to illustrate the use of a system dynamics framework. In the case of the effect of inundation on the modulus of HMA, it has been concluded (34) that the duration of the inundation period does not seem to have an effect on the damage of the pavements. Therefore, instead of inundation period, the number of cycles of inundation (comparable to number of cycles of conditioning in the laboratory) was assumed to have a significant effect on the modulus of the HMA. For example, a study (35) showed that the tensile strength ratio of a moisture susceptible mix dropped from 0.9 to 0.6 from one moisture conditioning cycle to ten, which is a reduction of 0.03 times the strength per cycle of inundation. 10 This relationship was utilized to represent the change in HMA modulus as a function of inundation cycles. Pavement performance analysis The temperature-pavement life and subgrade modulus-pavement life relationships were developed using the pavement structure for a two lane highway in coastal NH. The MechanisticEmpirical Pavement Design Guide (MEPDG) software version 1.1 was used to analyze the pavement structure under different climates and subgrade modulus values. The pavement structure consisted of 3 inches of asphalt concrete (12.5 mm NMAS, PG 64-28) over 16 inches of gravel, 24 inches of sand, and an A-1-b subgrade. Modulus values and other MEPDG inputs were obtained from NHDOT. Default values were used when specific information was not available. (Note that the thicknesses and moduli values used in this paper are given in US customary units as they have been used in the MEPDG software). For this particular pavement structure and traffic loading, rutting was the primary cause of failure in the pavements with increasing temperatures and decreasing subgrade modulus values. There was less than 10% alligator cracking predicted and the amount of longitudinal cracking decreases with increased temperature and decreased subgrade modulus. Therefore, pavement failure was defined as 0.75” of total rutting for the purposes of this project. The temperature-pavement life relationship was developed by running the MEPDG analysis of this pavement structure using climate files from cities along the eastern US coast. The number of months to a total rut depth of 0.75” was determined from the MEPDG output. This time to failure (defined as 0.75” total rut depth) was converted to years and then plotted against the maximum pavement temperature determined for that location using LTPPBind. A linear regression was performed to determine the relationship between years to failure and pavement temperature. To determine the subgrade modulus-pavement life relationship, MEPDG analyses were conducted using the NH climate and a range of reduced modulus values. The original modulus value for the subgrade was 26,500 psi. Modulus values of 95%, 90%, 80%, and 70% of the original value were used for MEPDG analysis. The number of years to failure (defined as 0.75” total rut depth) was determined from the MEPDG output and plotted against the subgrade modulus percentage. A linear regression was used to determine the relationship between years to failure and subgrade modulus. Results - Climate change and impacts The model was run for a time span of 100 years, and the results for the example pavement are shown in Figures 3 through 7. As expected, because of predicted changes in the climate, the maximum pavement temperature, the number of 100% saturation months and number of inundations do increase over time, and result in reduced effective subgrade modulus and minimum HMA modulus, and average pavement life. For the modeled pavement, the maximum temperature increases by 1.8C, the number of inundations increases by 10, and the number of months at which the soil is expected to be at or close to saturation increases from 2 to 4, over a time span of 100 years. The corresponding change in effective resilient modulus of the subgrade soil is from 23,500 to 20,500 psi, and that for the modulus of the HMA is from 625,000 to 11 400,000 psi. Figure 6 shows the most important finding - the change in average pavement life over the years, decreases from 16 to 4 years, over a time span of 100 years. The net effect of all of these changes is reflected in Figure 7, which shows the percent increase in cost with and without considering the effect of climate change. It can be seen that due to maintenance the overall cost decreases initially, but starts increasing after 20 years for both cases. The case with no consideration of climate change impact is a linear increase, with a positive increase only after 45 years, whereas for the case with the climate change impacts it is a non-linear increase with a positive increase after 35 years. At the end of 100 years, the scenario without climate impact increase is below 60% whereas the one with the climate impact is above 160% - a significant difference. Note that in these calculations the effect of rising costs has not been considered. Note that in this paper an asphalt mix overlay has been considered as the maintenance activity. 60.0 59.8 59.6 59.4 59.2 59.0 58.8 58.6 58.4 58.2 58.0 57.8 12 11 10 9 8 7 6 5 4 3 2 1 0 0 20 40 60 Time, Years 80 Number Temperature, C While these results are dependent on the specific pavement and traffic used and models that have been developed in this study, and will change as more accurate models are applied, it is clear that this reduction in life is caused by increasing damage that would result from detrimental effects on key material properties (such as modulus) caused by higher temperature, rainfall, sea water level and more Category 3 hurricanes. The key observation is that the changes in these factors (which are considered to be critical indicators of climate change) do cause a significant change in performance and deterioration of pavements. To protect our future generations from the burden of these expected detrimental changes, work is needed to improve pavement design and construction, to make pavements more resistant to the effects of climate change. 100 Maximum Pavement Temperature Number of Inundations Per Year Figure 3. Plot of number of inundations and maximum pavement temperature versus time 12 24,000 23,500 23,000 22,500 22,000 21,500 21,000 20,500 20,000 4.0 3.0 2.0 1.0 0.0 0 20 40 60 Time, Years 80 100 Resilient Modulus, psi Number of Months 5.0 Months with 100% Saturation Effective Resilient Modulus of Subgrade Figure 4. Plots of 100% saturation months and effective resilient modulus of subgrade versus time Minimum Modulus of HMA, psi 675,000 600,000 525,000 450,000 375,000 300,000 225,000 150,000 75,000 0 0 20 40 60 Time, Years 80 100 Figure 5. Plot of minimum HMA modulus versus time 13 Pavement Life, Years 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 0 20 40 60 Time, Years 80 100 Figure 6. Plot of average pavement life versus time 200 Increase in Cost, % 160 120 80 40 0 -40 -80 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95100 Time, Years With Climate Change Impact Without Climate Change Impact Figure 7. Increase in cost of maintenance, with and without the consideration of climate change impact on pavement performance 14 However, with rising costs and dwindling budgets, resources that are needed for such improvements are becoming scarce. The proper way to tackle this situation is to optimize pavement performance against the expected changes – and to do so, one needs accurate and reliable data regarding climate changes, particularly those changes that will have significant effects on pavement performance. While significant improvements in pavement design (both mix and structural) and construction (rehabilitation and maintenance) have taken place over the recent years, there is a big gap in understanding climate change and its effects on pavement performance in the future. For sustainable development, this information must be obtained through more collaborative research between climatologists and pavement engineers. Also, available climate, pavement and economic (as well as environmental) data must be integrated in system dynamics models and utilized for long term predictions. Such modeling and simulation will be crucial for investigating the individual and combined effects of multiple factors (which are not from the same discipline) and identifying the more critical ones. Furthermore, once the critical factors are determined, critical levels of these factors can be identified, that can trigger non-linear response of the key outputs, such as pavement life and costs of maintenance. Unless a holistic evaluation is made with system dynamics, it will not be possible to optimize our available resources to design and construct sustainable pavements. CONCLUSIONS AND RECOMMENDATIONS On the basis of this study, the following conclusions can be made. 1. System dynamics provide a practical way of evaluating the long term effect of climate change on pavement performance 2. A framework of using system dynamics to integrate climate change, pavement performance and economics of pavement maintenance has been proposed 3. Preliminary modeling and simulation show that the long term effects of changes in air temperature, rainfall, sea level rise and number of Category 3 hurricanes on pavement performance are significant 4. For the expected changes in climatic conditions, a significant reduction in stiffness properties of both subgrade and HMA can be expected 5. As a result of reduction in stiffness of the different layers, faster deterioration and a drastic reduction of average life of pavements are expected in the future 6. As evident from the evaluation of increase in cost of maintenance as a result of climate change, costs are expected to increase very significant (>160% in 100 years) and nonlinearly, in the future. 7. The significant and early increase in cost is not apparent from predictions that do not take climate changes and their effects on pavement performance into consideration. Recommendations for Future Work The work that is reported in this study clearly points out the necessity of the following. 1. To obtain more accurate and reliable data regarding relevant climate change factors, that are specific to regions, particularly those that are expected to be more vulnerable, such as coastal regions 2. To use system dynamics approach to integrate the multidisciplinary topics of climate change, pavement design and performance, and economics into comprehensive studies 3. To develop better models and refine the accuracy of predictions 15 4. To conduct region or site specific modeling and simulations for accurate prediction of impacts of climate change on pavement performance, disruption of transportation networks and associated economic factors These recommended actions will allow us to have a clear idea about future needs, prepare research plans for developing appropriate materials and methods, identify vulnerable areas, plan properly for appropriate budget allocations and optimize the spending of our precious tax dollars. ACKNOWLEDGEMENTS The authors gratefully acknowledge the help of Dr. Paul Kirshen in writing this paper. The ICNet project funded by the U.S. National Science Foundation research grant (CBET-1231326) facilitated this collaboration. REFERENCES 1. National Academies, Adapting to the Impacts of Future Climate Change 2010 2. Katsman CA, Church JA, Kopp RE, Kroon D, Oppenheimer M, Plag HP, Rahmstorf S, Ridley J, von Storch H, Vaughan DG, van der Wal RSW (2008a) High-end projection for local sea level rise along the Dutch coast in 2100 and 2200. In: Exploring high-end climate change scenarios for flood protection of the Netherlands. KNMI/Alterra, The Netherlands, pp 15–81. Available from <http://www.knmi.nl/bibliotheek/knmipubWR/WR2009-05.pdf> accessed July 27, 2013. 3. Pfeffer, W.T., J.T. Harper, and S. O'Neel, "Kinematic Constraints on Glacier Contributions to 21st-Century Sea-Level Rise", Science 321 no. 5894, pp. 1340-1343, 5 September 2008. DOI: 10.1126/science.1159099. 4. Smith, J.B., Schneider, S.H., Oppenheimer, M., Yohe, G.W., Hare, W., Mastrandrea, M. D., Patwardhan, A., Burton, I., Corfee-Morlot, J., Magadza, C. H. D., Fussel, H. M., A. Barrie Pittockk, Atiq Rahmanl, Avelino Suarezm, and Jean-Pascal van Ypersele. Assessing dangerous climate change through an update of the Intergovernmental Panel on Climate Change (IPCC) ‘‘reasons for concern’’. Proc. Natl. Acad. Sci. U.S.A. 106, 4133–4137, 2009. 5. Vecchi G. A and Soden, B. J. Effect of remote sea surface temperature change on tropical cyclone potential intensity Nature 450, 1066-1070 (13 December 2007) | doi:10.1038/nature06423; Received 23 July 2007; Accepted 26 October 2007. 6. Vermeera, Martin, and Rahmstorfb, Stefan. Global sea level linked to global temperature. www.pnas.org/cgi/doi/10.1073/pnas.0907765106< http://www.pikpotsdam.de/~stefan/Publications/Journals/vermeer_rahmstorf_2009.pdf > accessed February 18, 2013. 7. Pachauri, R. K., IPCC, Intergovernmental Panel on Climate Change - statement on November 28, 2012 < http://www.ipcc.ch/news_and_events/docs/COP18/Pachauri_COP18_address.pdf accessed February 18, 2012. 8. IPCC, Climate Change 2007 - The Physical Science Basis: Contribution of Working Group I to the Fourth Assessment Report of the IPCC, CAMBRIDGE UNIVERSITY PRESS, New York, NY 10013-2473, USA, 2007. 16 9. Transportation Research Board Special Report 290, “Potential Impacts of Climate Change on U.S. Transportation”, National Research Council of the National Academies, Washington, D.C., 2007. 10. Mills, B., Tighe, Susan., Andrey, J., Smith, James T., Parm, Suzanne., Huen, Ken. The Road WelTravelled: Implications of Climate Change for pavement Infrastructure in Souhern Canda. Final technical report. University of Waterloo, 2007. 11. Mills, B., Tighe, Susan., Andrey, J., Smith, James T., Huen, Ken. Climate Change Implications for Flexible Pavement Design and Performance in Southern Canda. Journal of Transportation Engineering, ADCE 135 (10) 2007. 12. Truax, D., Heitzman, M., Takle, G., Herzmann, D. and Timm, D. Developing MEPDG Climate Data Input Files for Mississippi. Final Report FHWA/ME-DOT-RD-11-232, 2011. 13. Meagher, W., Daniel, J. S., Jacobs, J. M., Linder, E. Method for Evaluating Implications of Climate Change for Design and Performance of Flexible Pavements Transportation Research Record: Journal of the Transportation Research Board Volume 2305 / 2012, Washington, DC. 14. Forrester, J. W. (1971). World dynamics. Cambridge, Mass, Wright-‐Allen Press. 15. Meadows, D. H., Meadows, D,L., Randers, J. and William W. Behrens III (1972). The Limits to Growth., New York, Universe Books. 16. Meadows, D. H., Meadows, D,L., and Jorgen Randers (1992). Beyond the limits. Post Mills, VT, Chelsea Green Publishing Company. 17. Meadows, D. H., Meadows, D,L., and Jorgen Randers. (2004). Limits to growth the 30 year update. White River Junction, VT, Chelsea Green Publishing Company. 18. Sterman, J. D. (2012). Sustaining Sustainability: Creating a Systems Science in a Fragmented Academy and Polarized World. Sustainability Science: The Emerging Paradigm and the Urban Environment. Springer 21-‐58. 19. Radzicki, M. J. (1995). A System Dynamics Approach to Sustainable Cities. The 1995 International System Dynamics Conference, Tokyo. 20. Bueno, N. P. (2011). A simple system dynamics model for the collapse of complex societies. The 29th International Conference of the System Dynamics Society, Washington DC, The System Dynamics Society 21. Bockermann, A., et al. (2005). "Modelling sustainability." Journal of Policy Modeling 27(2): 189-‐210. 22. Beck, A., Stave, K. (2011). Understanding Urban Quality of Life and Sustainability. The 29th International Conference of the System Dynamics Society, Washington. 23. Beck, A., Stave, K. (2012). Understanding Urban Quality of Life and Sustainability: Model Development and Validation. Proceedings of the 30th International Conference of the System Dynamics Society. , St. Gallen, Switzerland, The System Dynamics Society. 24. Grosskurth, J. (2007). "Ambition and reality in modeling: a case study on public planning for regional sustainability." Sustainability: Science, Practice, & Policy 3(1). 25. Maggio, G. and G. Cacciola (2012). "When will oil, natural gas, and coal peak?" Fuel 98: 111-‐123. 26. Schmidt-‐Bleek, F. (2008). "Factor 10: The Future of Stuff." Sustainability: Science, Practice, & Policy 4(1). 17 27. Hostetler, S.W., Alder, J.R. and Allan, A.M., 2011, Dynamically downscaled climate simulations over North America: Methods, evaluation and supporting documentation for users: U.S. Geological Survey Open-File Report 2011-1238, 64 p. 28. Hayhoe K, CP Wake, J Bradbury, T Huntington, L Luo, MD Swartz, J Sheffield, B Anderson, A DeGaetano, D Wolfe, E Wood (2007) Past and future changes in climate and hydrological indicators in the U.S. Northeast. Climate Dynamics 28, 381–407. doi: 10.1007/s00382-006-0187-8. 29. Kunkel, K.E., L.E. Stevens, S.E. Stevens, L. Sun, E. Janssen, D. Wuebbles, J. Rennells, A. 39 DeGaetano, and J.G. Dobson, 2012: Climate of the Northeast U.S. Draft (prepared as NOAA 40 Technical Memorandum). 30. Guswa, A.J., M.A. Celia, and I. Rodriguez-Iturbe. 2002. Models of soil moisture dynamics in ecohydrology: A comparative study. Water Resources Research, 38(9), 1166. 31. Kirshen, P., C. Watson, E. Douglas, A. Gontz, J. Lee, and Y. Tian. Coastal Flooding in the Northeastern USA under High and Low GHG Emission Scenarios, Mitigation and Adaptation Strategies for Global Change, Mitigation and Adaptation Strategies for Global Change, 13:437–451, 2008. 32. Done, J.M., G.J. Holland, C.L. Bruyère, L.R. Leung, and A. Suzuki-Parker. 2012. Modeling High-Impact Weather and Climate: Lessons from a Tropical Cyclone Perspective. Modeling high-impact weather and climate: Lessons from a tropical cyclone perspective. NCAR Technical Note NCAR/TN-490+STR, DOI:10.5065/D61834FM 33. NCHRP, Appendix DD-1:Resilient Modulus As Function Of Soil Moisture-Summary Of Predictive Models, 2004. 34. Zhang, Z., Wu, Z., Martinez, M., and Gaspard, K. Pavement Structures Damage Caused by Hurricane Katrina Flooding. J. Geotech. Geoenviron. Eng. 134, Special Issue: Performance of Geo-Systems during Hurricane Katrina, 633–643., 2008. 35. Mallick, R., Pelland, R and Hugo, F. Use of Accelerated Loading Equipment for Determination of Long Term Moisture Susceptibility of Hot Mix Asphalt. International Journal of Pavement Engineering, Volume 6, Issue 2, 2005. 18