IMPACTS OF FLOODING ON CONCRETE PAVEMENT Remi Oyediji, Research Associate, Centre for Pavement and Transportation Technology, University of Waterloo, Waterloo, Ontario, Canada

Susan L. Tighe, Norman W. McLeod Professor in Sustainable Pavement Engineering, Centre for Pavement and Transportation Technology, University of Waterloo, Waterloo, Ontario, Canada

Corresponding Author: ooyediji@uwaterloo.ca

KEYWORDS: Jointed Plain Concrete Pavement (JPCP), Flooding, Inundation, Pavement Performance, Climate Change

Conflict of Interest: None

1

1. ABSTRACT

With shifting paradigms in usual climatic events and increased occurrence of flood hazards, vulnerability assessment and adaptation of road infrastructure is essential. Road pavements are critical in sustaining socio-economic activities and their vulnerability to flood hazards could have serious cost consequences. Therefore, a conscientious decision to consider pavement materials, designs and alternatives that are resilient to recurring flood events is desired. Based on previous investigations into how pavements types, classes and configuration respond to extreme events, concrete pavements are reported as better flood-resilient systems in countries that have experienced intense flooding and inundation. Although Canada has experienced some of the worst flood incidences in history and owns a number of concrete pavement infrastructure, no study has been conducted to better understand its performance under extreme conditions. To provide insight on concrete pavement flood response, the use of the state of the art AASHTOWare Pavement ME Design (PMED) program is employed to model various flood scenarios on concrete pavement types and configurations common to two Canadian provinces, Ontario and Manitoba. The performance of the various pavement classes in terms of flood resilience, service life and cost feasibility is analyzed and results provide insight on the resilience and adaptive capacity of rigid pavements to flood hazards in Canada.

2

2. INTRODUCTION

Climate change is increasing the occurrence of climate hazards across Canada. Based on a report published by Public Safety Canada, flooding is reported to have occurred 241 times more than other climate hazards between1900 to 2005 (Sandink et al. 2010) and this continues to be on the rise. The likelihood of extreme precipitation, storm surge, and flash floods occurrence in Canadian provinces during the spring season have remained high over the past two decades and future projections under climate change reports increases in frequency in major parts of the country (Gaus et al 2018).

In the wake of these extreme events, two critical infrastructure systems, water and transportation, are reported vulnerable to flood impact.  Out of these two, transportation is of major interest as it is pivotal to the sustainability of socioeconomic activities such as agriculture, natural resources, fisheries, tourism, insurance and health. (Warren and Lemmen 2014). Transport Canada estimated the national contribution of transport services to Canada’s gross domestic product (GDP) to the tune of 4.2%, totaling over 100 billion dollars in the year 2011. The main component of these services in Canada are air, marine, rail and roads systems. Out of this four however, road transportation was noted as the most important asset for passenger and freight transportation, local (intra-city) and intercity transportation, intra-provincial transportation activities, and trade between Canada and the United States (in terms of value transported) (Transport Canada, 2011). Road pavements, an operational and functional component of Canada’s road infrastructure, are not inexorable to climate hazards being that they were mostly designed and engineered to consider climate conditions only at the time of construction. Consequently, frequencies and extremities of flood hazards impacts on their performance, relatively reducing service life. As it is of national interest that pavements be preserved and kept within serviceable limits therefore, climate change impact on pavement do need to be addressed (Tighe 2015).

Based on previous investigations into how pavements types, classes and configuration respond to flood events, concrete pavements is reported to provide better performance in various countries that have experienced hurricanes, typhoons, intense flooding and extreme inundation (Gaspard et al. 2006, Chen and Zhang 2014, Khan et al. 2017, Lukefahr 2018, Powell 2018).  From these studies, rigid pavement could be suggested as a better alternative for roads in flood plain areas. However, no study on concrete pavement flood impact has been conducted to understand how they will perform under Canadian extreme flood events. To provide insight on flood impact, the state of the art AASHTOWare Pavement ME Design (PMED) program is employed to model plausible flood scenarios on concrete pavement types and configurations common to two Canadian provinces, Ontario and Manitoba.

3. RESEARCH SIGNIFICANCE 

Albeit Canada has experienced some of the worst flood incidents in history and owns a number of concrete pavement infrastructure, no study on flood impact has been conducted to understand response and resiliency to extreme climatic events in a Canadian context.

3

4. PAVEMENT DESIGN DATA

Pavement design practice for Ontario and Manitoba provinces were sourced from research-based documents and reports published by road agencies and academia (ARA 2015a, ARA 2015b, Ahammed et al. 2013, Oberez et al. 2015). Traffic conditions, typical pavement structure, layer and sub-layer material properties for different road classes obtained from documents served as representative PMED inputs. For Ontario, typical municipal arterial and collector Jointed Plain Concrete Pavement (JPCP) road designs as shown in Figure 1 is modelled in PMED.

For Manitoba, twenty-seven (27) JPCP road classes are developed from a matrix of slab thicknesses, subgrade types and traffic levels to provide a holistic pavement flood impact evaluation. Also, concrete pavements are recorded higher in Manitoba than other Canadian provinces, hence the development of various road classes using design parameters provided in Table 1. Pavement configurations are classified to low, moderate and high traffic conditions with each traffic class having nine (9) different designs. Low traffic designs are denoted from C1 to C9, moderate traffic A1 to A9, and high traffic A10 to A18. Figure 2 represents typical pavement structure for C1.

(a) Collector Design (non-dowelled) (b) Arterial Design (dowelled)Figure 1. Pavement Structure for Typical Collector and Arterial Ontario Municipal Pavement

4

Table 1. Manitoba Matrix Design Parameters for Traffic, Slab thickness and Subgrade

Traffic PCC Slab Thickness (ST)

Subgrade

Type Two-Way

AADTT

Design Lane

Type Thickness (mm)

Type Subgrade (Mpa)

Moisture

Content

AASHTO Soil Class

Low 500 250 Thin 225mm Weak 35 23.80% A-7-6

Moderate

1000 500 Average

250mm Medium

73.1 13% A-6

High 2000 1000 Thick 275mm Strong 66.6 8.50% A-4

Figure 2. Typical Manitoba JPCP Pavement Structure (instance of Class C1)

5. PAVEMENT FLOOD IMPACT MODELLING

The PMED design tool program is employed to simulate performance of provincial typical pavement classes under no-flood and flood scenarios. A no-flood scenario or base case scenario

5

uses historical precipitation data for performance prediction while flood scenario harnesses climate-induced extreme precipitation values in performance simulation.

A. Historical Climate. For Ontario, a virtual station is created from interpolating two North American Regional Reanalysis (NARR) climate station data of proximity to a Toronto site location. Interpolated historical climate data include temperature, relative humidity, precipitation, wind and sunshine. PMED Enhanced Integrated Climate Model (EICM) uses this data for pavement climate impact modelling under no-flood conditions. For Manitoba, generated NARR climate data for the City of Winnipeg is integrated through PMED climate imputs. 

Plans towards using Modern-Era Retrospective Analysis for Research and Applications (MERRA) instead of NARR is currently underway to thus rigid pavement performance simulation as soon as PMED rigid models are recalibrated. (ARA 2018)

B. Future Climate. To estimate extreme precipitation values, Intensity Duration Frequency Climate Change Tool (IDF_CC Tool 3.0) is used to generate extreme precipitation values for the provincial locations with projection period from 2018 to 2100. The IDF_CC tool is an open source information which estimates precipitation accumulation depths for a variety of return periods (2, 5, 10, 25, 50 and 100 years) and durations (5, 10, 15 and 30 minutes and 1, 2, 6, 12 and 24 hours) for the Canadian environment. The tool engages 24 Global Circulation Models (GCMs) and 9 downscaled GCMs using rigorous downscaling method such as spatial and temporal downscaling, statistical analysis and optimization to update pre-estimated IDF from historical precipitation data to IDF under Representative Concentration Pathway (RCP 2.6, 4.5 and 8.5W/m2) climate change scenarios (Simonovic et al. 2016). RCP 4.5, compared to other RCP scenarios, reportedly has the least uncertainty in projected increase or decrease in flood frequencies across Canada (Gaur et al 2018). Also, since greenhouse gas concentrations for RCP 4.5 may peak in the year 2040 (Meinshausen et al. 2011) and considering a 25-year JPCP pavement design life, extremities of RCP 4.5 climate scenario would be represented in the pavement analysis. Consequent upon this, extreme magnitudes of recurrent intervals (50-year and 100-year) gridded at a resolution of 300 arc-seconds (0.0833 degrees, or roughly 10 km) is generated for Toronto, Ontario and Winnipeg, Manitoba locations under RCP 4.5. Extreme precipitation magnitudes are shown in Table 2

Table 2. Magnitudes of 50-year and 100-year recurrence interval under RCP 4.5

Location (Lat, Long) Duration

RCP 4.550-year recurrence interval

RCP 4.5100-year recurrence interval

Toronto, Ontario(43.81174, -79.41639)

Winnipeg, Manitoba(43.86200 -79.37000)

24hr

24hr

151.94mm 168.84mm

127.35mm

C. Flood Impact Simulation. van de Lindt et al. 2009 grouped flood loading into five (5) major classes namely flood depth, flood duration, flood velocity, flood debris and flood contaminants. These five (5) are potentially damaging to road pavements and should be considered in impact analysis. However, a modelling method for integrating all flood loads is yet

6

to exist. As of current, only flood duration and flood depth can be integrated into PMED climate file to define inundation conditions. Frequency of extreme events could likewise be repeated to model event recurrence. 

In this study, one, two and three extreme precipitation event cycles under RCP 4.5 is represented in PMED for concrete pavement flood performance simulation as Gaur et al 2018 projected higher flood frequencies in southwestern Ontario Northwest Territories, Yukon Territory, Nunavut compared to other Canadian provinces. His study reports the likelihood of a 100-year historical flood getting reduced to a frequency of 10–60 years by the end of the 21st century in these provinces. However, return flood in northern prairies (Manitoba for instance) and north-central Ontario could experience lower flood frequencies, with a return period of 100-year historical floods becoming 160–200 years return period in the future. From this information, extreme precipitation magnitudes for 50-year and 100- year recurrence intervals with repeated cycles of flood event under RCP 4.5 are generated for Ontario flood impact assessment, and RCP 4.5 100-year flood event for the Manitoba case study using an ensemble of climate prediction models. 

D. Flood Damage Estimation. Plausible extreme precipitation events and event cycles under climate change are introduced into PMED performance prediction and relative impact on pavement performance assessed. Relative change in performance under no-flood and flood scenario under RCP 4.5 is then calculated and computed as percentage flood damage using International Roughness Index (IRI) parameter. IRI is a function of joint faulting and slab cracking alongside climate (frost) and subgrade factors (AASHTO 2008), giving a holistic look into pavement overall performance. Equation 1 and 2 are used to estimate flood damage and pavement loss of life. 

δ IRI (% )=IRI f−IRI nf

IRI nf(1)

LSm(days)=365∗[( IRI f∗nIRI nf

)−n] (2)Where δ IRI (%) is the flood damage or damage ratio,IRI f is the terminal IRI (m/km) under RCP 4.5 Extreme Precipitation (EP) or flood conditions, IRI nf is the terminal IRI (m/km) at base-case or no-flood scenario,LSis the loss of pavement service life (days) and n is the pavement design life in years

6. RESULTS AND DISCUSSION – ONTARIO MUNICIPAL PAVEMENT

A. Ontario Flood Impact Assessment. At one event cycle, the same damage was observed across 50 and 100-year return periods in the municipal collector study. This demonstrates the

7

possibility of a higher return period having the same damaging effect as a lower return period at lower cycles of extreme precipitation. The opposite was the case for municipal arterial as damage ratio increase from 1.06 % to 1.86%. However, damage magnitude in the arterial study was lower to the collector study.  

At two event cycles, damage ratios in collector pavements remained the same as a one-cycle event, no further damage beyond 2.22% is predicted for both 50 and 100-year extreme event. Though arterial JPCP experienced lower damage magnitude compared to collector JPCP, there is an increase in damage across return periods, that is, from 1.59% to 2.39% for 50 and 100 years respectively. 

At three event cycles, an increase in damage ratio is noted across RCP return floods in the collector pavement. Increase in the number of cycles (two-cycle to three-cycle) resulted in a slight increase under the 50-year flood (from 2.22% to 2.5%) and larger increases under 100-year event (2.22% to 5.56%). Also, at three-cycle, arterial pavement damage increased from 1.59% to 2.92% and 2.39% to 2.92% for 50 and 100-year EP event respectively. Overall, the presence of dowels in the arterial pavement contributed to relatively low damage ratios even though it constituted more traffic loading.

B. Reduction in Pavement Life. Pavement life loss is higher in collector compared to arterial pavements, peaking at 507days to 266days respectively after three-cycle extreme precipitation as shown in Table 3.  Generally, increase in event cycles resulted in more loss of pavement life in the pavement classes. Summary of Ontario flood damage results is presented in Table 3 and Figure 3 and 4.

Table 3. Ontario Flood Damage under RCP 4.5

Flood Damage and Loss in Pavement Life

Pavement Class

Event 50-year return period

Loss in Pavement Life (50-year flood)

100-year return period

Loss in Pavement Life

(100-year flood)

Collector 1-cycle 2.22% 203 2.22% 203

2-cycle 2.22% 203 2.22% 203

3-cycle 2.50% 228 5.56% 507

Arterial 1-cycle 1.06% 97 1.86% 169

2-cycle 1.59% 145 2.39% 218

3-cycle 2.92% 266 2.92% 266

8

50-year,1-cycle 50-year,2-cycle 50-year,3-cycle 100-year,1-cycle

100-year,2-cycle

100-year,3-cycle

0.00%

1.00%

2.00%

3.00%

4.00%

5.00%

6.00%

2.22% 2.22%2.50%

2.22% 2.22%

5.56%

RCP 4.5 Extreme Precipitation Scenario

Dam

age

Rat

io (%

)

Figure 3. Flood Damage (%) against Return Flood of Ontario Municipal Collector Pavement under RCP 4.5 

50-year,1-cycle 50-year,2-cycle 50-year,3-cycle 100-year,1-cycle 100-year,2-cycle 100-year,3-cycle0.00%

1.00%

2.00%

3.00%

4.00%

5.00%

6.00%

1.06%

1.59%

2.92%

1.86%

2.39%

2.92%

RCP 4.5 Extreme Precipitation Scenario

Dam

age

Rat

io (%

)

Figure 4 Flood Damage (%) against Return Floods of Ontario Municipal Arterial Pavement under RCP 4.5 

9

7. RESULTS AND DISCUSSION – MANITOBA PAVEMENT

A. Low Traffic Designs. Figure 5 and 6 present flood damage and reduction in service life for road Class C1 to C9 having a matrix of slab thickness and subgrade levels under low traffic designs.  Flood damage had minimal influence on pavement Classes C2 and C3 and maximum impact on Class C7. Regarding service life, Class C9 possessed better service life than all other pavement class but sustained approximate average damage. Class C9 may be regarded as an overdesign as it not economically feasible to provide 275mm thick slab and strong subgrade for a 500 AADTT traffic. Low traffic should realistically be provided with thin corresponding slab thicknesses. Therefore, Classes C1-C6, having thin to medium slab thicknesses, are assessed based on damage intensity, service life, and cost feasibility. Based on these three, Classes C2 and C3 are better optimized designs. 

Figure 5.  Flood Damage (%) of Pavement Classes (C1 - C9) under Low Traffic Condition

10

Figure 6. Estimated Service Life of Pavement Classes (C1 - C9) under Low Traffic Condition

B. Moderate Traffic Designs. For pavement classes with moderate traffic and weak subgrade conditions, increases in slab thickness did not contribute to the predicted service life both before and after a flood event as observed in Classes A1, A6 and A7 as shown in Figures 7 and 8. Therefore, Class A1 is considered an economically viable option for this type of soil conditions and traffic group. 

For the average subgrade conditions, Classes A2, A5 and A8 are representative designs. Percentage of flood damage in Class A5 and A8 had a magnitude of 2.69%, less than A2 of 3.02% damage. Initially, the predicted service life of these classes before flood impact was the same. However, after 100-year event, A5 and A8 had better service life compared to A2. Therefore, considerations would be given to A5 and A8 based on its service life. In terms of economic feasibility, Class A5 has more leverage as its slab is of a medium thickness. This 250mm slab thickness is also typically preferred for moderate traffic volume in Manitoba.

Classes A3, A4 and A9 represent pavements with strong subgrade, moderate traffic, and varied slab thicknesses. Estimated damage for Classes A3, A4 and A9 are 2.73%, 2.74% and 3.74% respectively. Class A3 had a better service life compared to other road classes even after flood event. Classes A3, A4 and A9 had the same predicted service life before flood impact but after flood impact, Classes A3 and A4 outperformed Class A9 in terms of service life. The reduced cost of having a thin pavement perform better than a medium or thick pavement is a viable option as this holistically reduce the cost implications.

11

Figure 7. Flood Damage (%) of Pavement Classes (A1 - A9) at Moderate Traffic Conditions

Figure 8. Estimated Service Life of Pavement Classes (A1-A9) at Moderate Traffic Conditions

C. High Traffic. Classes A12, A13 and A16 represent pavement of low subgrade and high traffic condition having thin, medium and thick slab thicknesses respectively. Class A16

12

experienced the lowest damage magnitude, having a value of 0.31% compared to Classes A12 and A13 which had flood damage of 3.00% and 3.02% as shown in Figure 9. Class A16 has a thick slab thickness while Classes A12 and A13 have thin and medium slab thickness respectively. Figure 10 shows that the three pavement classes had the same predicted performance before the influence of flood. However, after flood impact, the pavement reduction in service life is more pronounced in the Class A12 and A13 compared to A16. Class A16 shows more sustainability and resilience to both high traffic conditions and flood-induced damage.  Generally, pavement with thick slabs such as we have in A16 is often required under high traffic loading conditions, especially when the subgrade is made of weak soils. Supposing a road agency decides to use Class A13 and A12 based on their equivalent performance and lower cost implication under historical climate conditions, the performance of these pavement classes could significantly reduce if relatively compared to Class A16 performance in the wake of a major flood event.

Class A11, Class A14 and Class A17 represent pavements with average subgrade and high traffic volume with respective slab thicknesses.  These classes all had the same damage ratio (2.67%) irrespective of slab thickness and the same predicted service life before and after extreme precipitation or flood hazard. Relatively implying cost and performance under extreme climate events could be optimized in Manitoba if underlying soil is an average subgrade as indexed in this study.

Classes A10, A15 and A18 represent pavements with strong subgrade, high traffic volume and varied slab thicknesses. The same damage magnitude is observed across the three classes (2.72%). This further reinstates the ability of the subgrade soil to contribute to pavement overall performance both under flood or no-flood conditions. A good subgrade could invariably optimize the performance, service life and cost of the pavement structure as it allows more flexibility in selecting JPCP slab thicknesses both under flood and no-flood scenarios. 

13

Figure 9. Flood Damage (%) of Pavement Classes (A10 - A18) at High Traffic Conditions

Figure 10. Estimated Service Life of Pavement Classes (A10-A18) at High Traffic Conditions

From the observation of damage results, minimum and maximum IRI damage ratio estimated across pavement classes under RCP4.5 flood scenarios are respectively 1.06% and 5.56% in the Ontario municipal study, and 0.31% and 3.03% in the Manitoba study. Comparing these damage ratios to flood damage magnitude recorded two years after the Hurricane Katrina event on highly deteriorated flooded concrete pavements, both Ontario and Manitoba JPCPs would fall in the minor flood damage category under RCP 4.5 climate scenario. From the Katrina study conducted by Chen and Zhang (2014), 23.93% flood damage is estimated for highly deteriorated flooded concrete pavements. From these calculated metrics, one could hypothetically classify a minor flood damage to between 0 to 8%, moderate flood damage be from 8 to 16% and major damage be more than 16%.

8. CONCLUSIONS

An investigation is conducted to understand the impact of flooding on rigid pavements under Canadian extreme climatic conditions. Representative JPCP designs for two provinces, Ontario and Manitoba, are sourced and modelled using PMED to estimate their performance at no-flood and flood conditions under RCP 4.5 intermediate climate change scenario. Inference from study is below shown:

1.   In the Ontario study, flood damage is estimated by observing relative changes in International Roughness Index (IRI) prediction values for typical municipal arterial JPCP

14

and collector JPCP classes. No major damage is induced due to flood impact as there was only a slight reduction in pavement performance across road classes. However, estimated flood damage on pavement performance is more pronounced in collector (non-dowelled) pavements than arterial (dowelled) pavements. 

2. In the Manitoba case study, a total of 27 pavement design classes is developed based on a matrix of representative traffic levels, subgrade conditions and slab thicknesses in the province. Projected climate-induced flood hazards were simulated on the design classes to evaluate flood impact on concrete pavement performance. Results indicate diminutive flood damage and loss of life in all of the concrete pavement classes. Observed low flood damage ratios further reiterates the resilience and adaptive capacity of Jointed Plain Concrete Pavement (JPCP) to withstand extreme precipitation or flood conditions. Based on the developed matrix, performance of nine (9) structural designs alternatives for each low, medium and high traffic class is evaluated for flood resilience, cost feasibility and durability. Designs are compared and better performing alternatives proposed

3. Results of concrete flood damage generally provided insight into the resilience of concrete pavement to flood impact in Canada. Concrete pavement is therefore suggested as a rehabilitation measure for flood damaged pavements in Canadian floodplain zones.

9. ACKNOWLEDGEMENTS

The research was supported by the Natural Sciences and Engineering Research Council of Canada, Cement Association of Canada, Norman W. McLeaod Chair Partners and Centre for Pavement and Transportation Technology (CPATT) Partners.

10. REFERENCES

1. AASHTO. (2008) Mechanistic-Empirical Pavement Design Guide: A Manual of Practice

Interim Edition. (Jan 16 2019)

2. Ahammed, M. A., Kass, S., & Hilderman, S. (2013) Implementing the AASHTOWare

Pavement ME Design guide: Manitoba issues and proposed approaches.

https://www.scopus.com/inward/record.uri?eid=2-s2.0-

84896642865&partnerID=40&md5=84054c27916c6f4de89b5290a3c3f49e (Mar. 12

2019)

15

3. Applied Research Associates ARA Inc. (2015a). Methodology for the development of

equivalent pavement structural design matrix for municipal roadways-Ontario.

http://www.rediscoverconcrete.ca/assets/files/Typical-Ontario-Pavement-Designs-Final-

Report_012115.pdf (Sept 12 2018).

4. Applied Research Associates ARA Inc. (2015b). Methodology for the development of

equivalent pavement structural design matrix for municipal roadways: including

maintenance and rehabilitation schedules and life cycle cost analysis. Toronto, Ontario:

ARA

5. Applied Research Associates ARA Inc. (2018). AASTHO Pavement ME Design Build

2.5.0 Release Notes. http://www.me-design.com/MEDesign/data/AASHTOWare

%20Pavement%20ME%20Design%20Build%202.5%20Release%20Notes.pdf (Sept 20

2018).

6. Chen, X. and Zhang, Z. (2014). Effects of hurricanes Katrina and Rita flooding on

Louisiana pavement performance. pavement materials, structures and performances. In:

Geo-Shanghai 2014 Conference, 26-28 May, 2014 Shanghai, China: ASCE, 212-221.

7. Gaspard, K., et al., (2006). Impact of Hurricane Katrina on Roadways in the New

Orleans Area.” Technical Assistance Rep. No. 07-2TA, Louisiana Dept. of Transportation

and Development, Hammond, LA. (Dec. 2 2018).

https://www.ltrc.lsu.edu/pdf/2007/07_2ta.pdf

8. Gaur, A., Gaur, A., and Simonovic, S. P. (2018). “Future Changes in Flood Hazards

across Canada under a Changing Climate”. Water (Switzerland), 10(10). [Mar. 2 2018]

https://doi.org/10.3390/w10101441

9. Khan, M. U., et al., (2017). A case study on pavement performance due to extreme

moisture intrusion at untreated layers. International Journal of Pavement Engineering.

https://doi.org/10.1080/10298436.2017.1408272 (Oct 15 2018).

10. Lukefahr E. L. (2018). Continuously Reinforced Concrete Pavement Resiliency – A Case

Study. In ACPA 55th Annual Meeting 2018 New Orleans Marriot, United States.

16

Available from: http://www.acpa.org/wp-content/uploads/2018/12/Lukefahr-ACPA-

2018-CRCP-Resilience.pdf

11. Meinshausen, M., et al., (2011). The RCP greenhouse gas concentrations and their

extensions from 1765 to 2300: Climatic Change, 109: 213.

https://doi.org/10.1007/s10584-011-0156-z (Oct 9 2018).

12. Oberez, M. E., Kass, S., Hilderman, S., & Ahammed, M. A., (2015). “Estimating Base

Layers and Subgrade Moduli for ME Pavement Design in Manitoba.” In National

Transportation Association of Canada TAC Conference, 27-30 September 2015

Charlottetown. PEI: Transportation Association of Canada.

13. Powell J., (2018). CRCP roadways proved resilient after hurricane Harvey.

https://goo.gl/hQe1Do (Nov 7 2018).

14. Sandink, D., et al. (2010). Making Flood Insurable for Canadian Homeowners: A

Discussion Paper. Toronto. Institute for Catastrophic Loss Reduction & Swiss

Reinsurance Company Ltd.

15. Simonovic, S.P., et al., (2016) A web-based tool for the development of intensity

duration frequency curves under changing climate. Journal of Environmental Modelling

& Software. https://doi.org/10.1016/j.envsoft.2016.03.016 (Oct 18 2018).

16. Tighe, S. L., (2015). Vulnerabilities and design considerations for pavement

infrastructure in light of climate change. In: National Transportation Association of

Canada TAC Conference, 27-30 September 2015 Charlottetown. PEI: Transportation

Association of Canada.

17. Transport Canada (2011). Transportation in Canada 2011: Comprehensive Report;

Minister of Public Works and Government Services, Canada.

https://www.tc.gc.ca/eng/policy/report-aca-anre2011-index-3010.htm (Mar 10 2019]

18. van de Lindt, J., Taggart Mason, (2009). Fragility analysis methodology for performance-

based analysis of wood-frame buildings for flood. Natural Hazards Review, 10 (3), 113–

123.

17

19. Warren, F.J. and Lemmen, D.S., (editors) (2014). “Canada in a Changing Climate:

Sector Perspectives on Impacts and Adaptation.” Government of Canada, Ottawa, ON,

286p.

18