Impact of Climate Change on Nunavik's Marine

Home , Great Whale River, Koksoak River, Kuujjuarapik, Lake Tasiujaq, Quaqtaq, Salluit Airport

KNOWLEDGE SYNTHESIS: IMPACT OF CLIMATE CHANGE ON NUNAVIK’S MARINE AND COASTAL ENVIRONMENT

Report presented to the Ministère des Transports du Québec

Final Report

July 2020

IMPACT OF CLIMATE CHANGE ON NUNAVIK’S MARINE AND COASTAL ENVIRONMENT: KNOWLEDGE SYNTHESIS Final Report May 2020

PROJET COORDINATION Geneviève Trudel, MTQ Stéphanie Bleau, Ouranos PRODUCTION AND CARTOGRAPHY Sonia Hachem, Ouranos Stéphanie Bleau, Ouranos Raphaël Desjardins, Ouranos SCIENTIFIC REVISION Antoine Boisson, Université Laval Isabelle Charron, Ouranos Hélène Côté, Ouranos Julie Cunningham, Ouranos Dany Dumont, UQAR Yves Gauthier, INRS-ETE LINGUISTIC REVISION Nom de réviseur/use linguistique 1, Affiliation TRANSLATOR Julien Sandiford (Knowledge Synthesis and English outreach tools) Nom du traducteur (Inuktitut Executive Summary and flyer) PAGE LAYOUT Prénom Nom, Affiliation Project title: Impact of Climate Change on Nunavik’s Marine and Coastal Environment: Knowledge Synthesis Project number: T2.1 (Action 28.2 PACC 2013-2020)

Suggested citation: Hachem S. and Bleau S., (2020). Impact of Climate Change on Nunavik’s Marine and Coastal Environment: Knowledge Synthesis. Report presented to the Ministère des transports du Québec. Ouranos. Montréal. 70 p. + appendices.

The results and opinions presented in this publication are entirely the responsibility of the authors and do not commit Ouranos or its members. Any further use of the document will be at the sole risk of the user, and the authors shall bear no liability nor be subject to legal action.

ACKNOWLEDGEMENTS

This Knowledge Synthesis is an MTQ initiative falling under the framework of action 28.2, priority 28 of the 2013-2020 Climate Change Action Plan (PACC, Plan d’action sur les changements climatiques). This non-technical document was financed by the Fonds vert 2020 and coordinated by Ouranos. The study findings presented in this decision-support and knowledge transfer tool come mainly from various research projects carried out as part of a research initiative led by the Ministère des Transports du Québec under the name “Évaluation de l’impact des changements climatiques sur les infrastructures maritimes du Nunavik et solutions d’adaptation (Evaluation of Climate Change Impacts on Nunavik Marine Infrastructure and Adaptation Solutions),” started in 2009 in the context of the Plan d’action sur les changements climatiques 2006-2012. It will continue as part of the PACC 2013- 2020. The authors would like to acknowledge the great contribution made by the members of the monitoring committee who participated in the project and in the various revision phases that helped improve the tool: Anick Guimond (MTQ), Geneviève Trudel (MTQ), Laurie Beaupré (Makivik), Véronique Gilbert (ARK), Jean-Denis Bouchard (MSP), Julie Veillette (MELCC) and Frédérique Gosselin-Lessard (MTQ). The authors would also like to thank Isabelle Charron and Julie Cunningham from the Knowledge Mobilization team, and Hélène Côté from the Climate Science and Climate Services team, who served as internal Ouranos revisers, as well as external scientific revisers Antoine Boisson from CEN, Dany Dumont from ISMER and Yves Gauthier from INRS-ETE, whose comments and corrections helped make the text clearer, simpler and more concise. This initiative brought together many participants from various public regional and local government organizations, as well as academic and private organizations. The MTQ and Ouranos would like to thank the following collaborators and representatives for their participation in this research and development initiative from its inception in 2009: • Kativik Regional Government (KRG) • Makivik Corporation • Ministère de la Sécurité publique (MSP) • Ministère de l'Environnement et de la Lutte contre les changements climatiques (MELCC) • INRS-ETE • UQAR-ISMER • CEN, ULaval • CIMA+ • Environnement Illimité Inc. • LaSalle NHC

SUMMARY

The Ministère des Transports du Québec (MTQ) has coordinated and/or financed several studies on the impacts of climate change on Nunavik’s marine and coastal environment between 2009 and 2020 in the context of various iterations of the PACC. In order to ensure access to and the effective transfer of the understanding developed in the context of these research projects, as well as the use of the acquired knowledge in the application and monitoring of adaptation measures by various stakeholders at the regional and local government and private sector levels, the results have been collected in a decision-support tool that presents the key findings. Context of the project The impacts of climate change are already more significant in Nunavik, and these effects will continue to intensify. Increases in air temperature (4 to 7.5°C) and precipitation (20 to 30%) will be greater than in the rest of Quebec by the end of the 21st century, mainly in the fall and winter. Since coastal infrastructure will be affected by these changes, the MTQ (in collaboration with the Kativik Regional Government, or KRG) wanted to document and improve the understanding of several hydrological and climate hazards with a potential impact on the medium- and long-term integrity and longevity of Nunavik’s coastal facilities. The objective of the approach is to optimize the planning, maintenance and rehabilitation of the infrastructure and establish measures to strengthen its climate change resilience. This Knowledge Synthesis deals with Nunavik specifically, providing a portrait of the hazards at the regional level and the level of its 14 communities. It provides integrated insights on the vulnerabilities in order to evaluate risks and direct the sustainable adaptation of marine infrastructure. It is accompanied by a technical report explaining the scientific and technical concepts that made the research and findings possible, as well as a Powerpoint presentation and flyer presenting the results in plain language. Main study findings As result of the later ice formation period in the fall and earlier melting in the spring, the ice-free period could be longer by six weeks to two months. The period of unstable sea and landfast ice could be longer. In December, ice concentrations could decrease to 40 and 60% on the coasts from Ivujivik to Kangiqsualujjuaq and ice formation could cease entirely between Ivujivik and Inukjuak by the 2040- 2070 period. Extreme high and low water levels that previously had 100-year return periods could occur every 50 years by approximately the middle of the 21st century. Most coastal sites in Nunavik could experience positive storm surges of up to 1 m and negative storm surges at the same levels as in the recent past. However, post-glacial rebound in Nunavik could compensate for the higher global sea level, and could even lead to lower or unchanged relative sea levels at several sites (especially in Hudson Bay) by the end of the 21st century, thus reducing the impact of positive storm surges. On the other hand, extreme negative storm surges and the reduced rise in the global sea level could lead to lower extreme low levels with greater coastal impacts. Knowledge integration for the implementation of adaptation measures Coastal risks exist where potential hazards coincide with the presence of populations or infrastructure. In Nunavik, the population of the 14 Inuit communities and Deception Bay are located on the coastline and are thus potentially at risk. The understanding of climate risks and the factors that influence the vulnerability and risks of Nunavik’s marine and coastal environments is growing, providing an already solid foundation for solution implementation. Decision makers and local populations are more

conscious of climate issues than ever (Shah, 2018), and conditions are ideal for the implementation of adaptation measures. In fact, there have already been a number of initiatives aimed at successful adaptation in northern environments. Construction standards specific to the North have already been established, and action and outreach programs to facilitate an understanding of climate change and risk-reduction measures are also in place. Of course, some knowledge-related challenges remain, but these issues should not prevent the implementation and monitoring of adaptation measures. These measures will continue to be improved as social acceptance grows and our understanding becomes clearer and more detailed.

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...... i SUMMARY ...... ii LIST OF FIGURES ...... vi List OF TABLES ...... viii LIST OF ABBREVIATIONS AND acronyms ...... x GENERAL INTRODUCTION ...... 1 1. CONTEXT OF THE KNOWLEDGE SYNTHESIS ...... 2 1.1 POPULATION AND INFRASTRUCTURE PROFILE ...... 2 1.2 Nunavik CLIMATE PROFILE ...... 4 1.3 EVALUATING THE CHANGES IN CLIMATE AND HYDROLOGICAL HAZARDS AND THEIR ASSOCIATED CHARACTERISTICS ...... 6 1.4 EVALUATION OF THE VULNERABILITY OF THE COASTAL ENVIRONMENT AND OF COMMUNITIES ...... 8 1.5 THE KNOWLEDGE SYNTHESIS AS A STARTING POINT FOR THE ADAPTATION PROCESS ...... 8 1.6 DISCLAIMER ABOUT THE UNCERTAINTY OF THE FINDINGS IN THE VARIOUS REPORTS PRODUCED BY THE MTQ AND ITS PARTNERS ...... 9 2. KEY FINDINGS ...... 11 2.1 Nunavik ...... 12 2.1.1 General Profile ...... 12 2.1.2 Relative Sea Level ...... 13 2.1.3 Extreme Water Levels, Frequency and Seasonality ...... 14 2.1.4 Storm Frequency ...... 19 2.1.5 Ice Conditions ...... 20 2.1.6 Waves ...... 22 2.2 HUDSON BAY ...... 23 2.2.1 General Profile ...... 23 2.2.2 The Communities of Hudson Bay ...... 25 2.3 HUDSON STRAIT ...... 36 2.3.1 General Profile ...... 36 2.3.2 The Communities of Hudson Strait ...... 37 2.4 UNGAVA BAY ...... 42 2.4.1 General Profile ...... 42

2.4.2 The Communities of Ungava Bay ...... 43 3. KNOWLEDGE INTEGRATION FOR THE IMPLEMENTATION OF ADAPTATION MEASURES ...... 46 3.1 HAZARDS, VULNERABILITIES AND RISKS IN NORTHERN QUEBEC ...... 46 3.2 DATA AND KNOWLEDGE SUPPORTING ADAPTATION ...... 47 3.2.1 Analyzing Vulnerabilities ...... 48 3.2.2 Spatial Analysis of At-Risk Areas ...... 50 3.3 GOING FURTHER: DECISION-MAKING TOOLS ...... 50 3.3.1 Design and Adaptation in a Context of Uncertainty: A Few Points of Reference ...... 50 3.3.2 Useful Information for Climate Change and Land-use Professionals and Municipal Managers ...... 51 3.4 PROMISING RESEARCH AVENUES FOR NUNAVIK’S COASTAL AND MARINE ENVIRONMENTS...... 52 3.4.1 Improving the Scientific Understanding of the Environment ...... 52 3.4.2 Improving Knowledge for a More Efficient and Robust Adaptation Process ...... 54 GENERAL CONCLUSION ...... 56 Glossary ...... 58 References ...... 61 Appendix – Water level and coastal sensitivity maps ...... 66

LIST OF FIGURES

Figure 1: Data collection locations for the hazards studied in Nunavik, with collection periods and names of maintenance networks. The data were collected by meteorological stations (temperatures, wind and humidity) tide gauges (water level on the coast), wave gauges (water level at sea), cameras (photos of the sea and the coast) and Landsat and MODIS images (satellite images)...... Erreur ! Signet non défini. Figure 2: Extreme storm surge levels, with a 100-year return period, from the recent past (1980- 2009) and (a) negative storm surge and (b) positive storm surge variations for the 2049-2069 and 2070-2099 periods...... Erreur ! Signet non défini. Figure 1: Data collection locations for the hazards studied in Nunavik, with collection periods and names of maintenance networks. The data were collected by meteorological stations (temperatures, wind and humidity) tide gauges (water level on the coast), wave gauges (water level at sea), cameras (photos of the sea and coast) and Landsat and MODIS images (satellite images)...... 3 Figure 2. The adaptation process according to Ouranos, 2020 1. Identify; 2. Prepare; 3. Implement; 4. Adjust...... 9 Figure 3. Tidal range in Nunavik (CanCoast data product, source data from Canadian Hydrographic Service). (Boisson, 2019)...... 12 Figure 4. Extreme storm surge levels, with a 100-year return period, from the recent past (1980- 2009) and (a) negative storm surge and (b) positive storm surge variations for the 2049-2069 and 2070-2099 periods...... 16 Figure 5. Average concentration of sea ice in Nunavik in December 1980-2010 (a) and 2040-2070 (b) (Senneville, 2018a)...... 21 Figure 6. Kuujjuarapik coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods ...... 26 Figure 7. Coastal Sensitivity Index, positive storm surges for the 1989-2009 period and positive storm surge variations for future periods (2040-2069 and 2070-2099)...... 49 Figure 1. Umiujaq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods ...... 66 Figure 2. Inukjuak coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods ...... 67 Figure 3. Puvirnituq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049- 2069 and 2070-2099 periods ...... 68

Figure 4. Akulivik coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods ...... 69 Figure 5. Ivujivik coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods ...... 70 Figure 6. Baie Déception coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods ...... 71 Figure 7. Kangiqsujuaq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049- 2069 and 2070-2099 periods ...... 72 Figure 8. Quaqtaq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods ...... 73 Figure 9. Kuujjuaq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods ...... 74

vii

LIST OF TABLES

Summary of key climate data for Nunavik ...... Erreur ! Signet non défini. Extreme storm surge levels (negative and positive): Nunavik average, frequency and seasonality ...... Erreur ! Signet non défini. Nunavik ice cover and duration risks in future periods, in winter, spring and fall .. Erreur ! Signet non défini. Summary of key climate data for Nunavik (IRIS 4, 2nd edition – personal communication) ...... 4 List of simulated data for the different studied sites in Nunavik ...... 7 Relative sea level (cm ± 95% confidence interval) according to climate scenarios RCP4.5, RCP8.5 and RCP8.5 with the collapse of the West Antarctic Ice Sheet (WAnt), at certain sites in Nunavik for the 2081-2100 period, as compared to 1986-2005. These scenarios take post-glacial rebound according to James et al. (2014) into account...... 14 Extreme negative and positive storm surge levels with a 100-year return period simulated for the recent past and for the near and distant future (Massé and Gallant, 2016) ... 16 Extreme storm surge levels (negative and positive): Nunavik average, frequency and seasonality ...... 17 Return period of extreme low total water levels in the current and future climates in Umiujaq, Ivujivik and Quaqtaq (from Figures 5.15, 5.14 and 5.13 in Massé and Gallant, 2016) 18 Return periods of extreme high total water levels in current and future climates in Umiujaq, Ivujivik and Quaqtaq (from Figures 5.9, 5.8 and 5.7 in Massé and Gallant, 2016) ...... 19 Nunavik ice cover and duration risks in future periods, in winter, spring and fall ...... 21 NRCan modeling of post-glacial rebound rates for Hudson Bay communities ...... 24 Summary of ice observations obtained by camera for Kuujjuarapik from 2015 to 2018 (Bernier et al., 2017; 2019; Poulin et al., 2018) ...... 27 Duration of the various stages of the ice cover as captured by camera in Kuujjuarapik from 2015 to 2018 (Bernier et al., 2017; 2019; Poulin et al., 2018) ...... 28 Minimum negative storm surge and total levels in the future climate in Umiujaq (from Figure 5.15 in Massé and Gallant, 2016) ...... 29 Return periods for certain minimum extreme total water levels in current and future climates in Umiujaq (from Figure 5.15 in Massé and Gallant, 2016) ...... 29 Maximum positive storm surge and total levels in the future climate in Umiujaq (from Figure 5.15 in Massé and Gallant, 2016) ...... 30 Return periods for certain maximum extreme total water levels in current and future climates in Umiujaq (from Figure 5.15 in Massé and Gallant, 2016) ...... 30

viii

Duration of the various stages of the ice cover as captured by camera in Umiujaq from 2016 to 2018 (Bernier et al., 2017; 2019; Poulin et al., 2018) ...... 31 Minimum negative storm surge and total levels in the future climate in Ivujivik (from Figure 5.14 in Massé and Gallant, 2016) ...... 34 Return periods for certain minimum extreme total water levels in current and future climates in Ivujivik (from Figure 5.14 in Massé and Gallant, 2016) ...... 34 Maximum positive storm surge and total levels in the future climate in Ivujivik (from Figure 5.8 in Massé and Gallant, 2016) ...... 35 Return periods for certain maximum extreme total water levels in current and future climates in Ivujivik (from Figure 5.8 in Massé and Gallant, 2016) ...... 35 Ice processes in Ivujivik (Bernier et al., 2017; 2019; Poulin et al, 2018) ...... 35 Rates of post-glacial rebound in Hudson Strait communities according to NRCan model (James et al., 2014) ...... 37 . Ice processes in Deception Bay (Bernier et al., 2017; 2019; Poulin et al, 2018) ...... 38 Minimum negative storm surge and total levels in the future climate in Quaqtaq (from Figure 5.13 in Massé and Gallant, 2016) ...... 39 Return periods for certain minimum extreme total water levels in current and future climates in Quaqtaq (from Figure 5.13 in Massé and Gallant, 2016) ...... 40 Maximum positive storm surge and total levels in the future climate in Quaqtaq (from Figure 5.7 in Massé and Gallant, 2016) ...... 40 Return periods for certain maximum extreme total water levels in current and future climates in Quaqtaq (from Figure 5.7 in Massé and Gallant, 2016) ...... 41 Ice processes in Quaqtaq (Bernier et al., 2017; 2019; Poulin et al, 2018) ...... 41 Ice processes in Aupaluk (Bernier et al., 2017; 2019; Poulin et al., 2018) ...... 44 Sensitivity indices used to construct the Coastal Sensitivity Index (CSI) (from Manson et al., 2019) ...... 48

LIST OF ABBREVIATIONS AND ACRONYMS

CBA Cost–benefit analysis CEN Centre d’études nordiques (Centre for Nordic Studies) CICE Sea ice model (CICE5: version 5 of CICE) CSI Coastal Sensitivity Index (CSI) – Manson et al., 2019 DJF December, January and February, which represent the winter months GHG Greenhouse gas INRS-ETE Institut national de la recherche scientifique - Eau, terre et environnement (National Institute for Research – Water, Earth and Environment) IPCC Intergovernmental Panel on Climate Change ISMER Institut des sciences de la mer de Rimouski (Rimouski Institute for Marine Sciences) JJA June, July and August, which represent the summer months KRG Kativik Regional Government MAMH Ministère des Affaires Municipales et de l’Habitation MELCC Ministère de l’Environnement et de la Lutte contre les Changements Climatiques (Ministry of Environment and the Fight Against Climate Change) MSP Ministère de la Sécurité publique (Ministry of Public Security) MTQ Ministère des Transports du Québec (Quebec Ministry of Transport) NRCan Natural Resources Canada PACC Plan d’action 2013-2020 sur les changements climatiques (2013-2020 Climate Change Action Plan) RCM Regional Climate Model RCP Representative Concentration Pathways ROM Regional Ocean Model UQAR Université du Québec à Rimouski (University of Quebec in Rimouski)

GENERAL INTRODUCTION

The Intergovernmental Panel on Climate Change (IPCC) has projected 1°C of global warming from preindustrial levels (IPCC, 2014). This increase could reach up to 2°C by 2031-2050 and 4.3°C by 2081-2100 (IPCC, 2019) according to the high-emissions RCP8.5 scenario (Inset 1). The warming is most significant in the Arctic, where temperatures have increased at more than double the average global rate over the last two decades (IPCC, 2019). In the Canadian Arctic, the warming is even more significant, with average annual air temperatures increasing at three times the average annual increase (Bush et al., 2019). Projections indicate that the warming will escalate in all seasons in all the considered greenhouse gas (GHG) emissions scenarios (Inset 1). However, extreme summer temperatures (hottest temperatures) and extreme winter temperatures (coldest temperatures) could increase even more significantly (Bush et al., 2019). The warming of the climate is also being strongly felt in sub-Arctic regions like Northern Quebec. In the context of increased warming in northern regions, and in collaboration with the Kativik Regional Government (KRG), the Government of Quebec wanted to measure the magnitude of the changes with potential future impacts on the marine environment. This Knowledge Synthesis brings together the results of the studies coordinated by the Ministère des Transports du Québec (MTQ) under the name “Évaluation de l’impact des changements climatiques sur les infrastructures maritimes du Nunavik et solutions d’adaptation (Evaluation of Climate Change Impacts on Nunavik Marine Infrastructure and Adaptation Solutions),” financed by the Fonds vert in the context of the 2006-2012 Climate Change Action Plan (PACC, Plan d’action sur les changements climatiques). It will continue as part of the PACC 2013-2020. The document is aimed at various users, including climate change and land-use professionals (e.g. MAMH, MSP, MELCC, KRG) and municipal managers. It will provide users working in a northern environment with a decision-support tool to which various stakeholders can refer. It also makes available information specific to the coasts of Hudson Bay, Hudson Strait and Ungava Bay that can be used to design, implement and monitor adaptation measures in coming years (for example, in the context of land occupancy planning and infrastructure and facilities maintenance along the northern coastline). In short, the Knowledge Synthesis can support the implementation of climate change adaptation solutions and decision-making and planning processes. This document is accompanied by a glossary of scientific terms (which are italicized in the text when they first appear), an inset that explains the complex scientific concepts, a bibliography that presents the references used in the body of the synthesis and an appendix with the maps of the communities. The Technical Synthesis (not part of the Knowledge Synthesis) documents the measurement devices and observation methods used in the recent past (including the calibration and parameters of the models used in the work) and provides specifications on the models and greenhouse gas emission scenarios used in the studies to evaluate future changes. The Knowledge Synthesis is divided into three chapters: The context of the projects implemented by the MTQ in Nunavik since 2009 (Chapter 1). The results of the MTQ studies for Nunavik, for each coastal region, and finally for each community (Chapter 2). A demonstration of the utility of the knowledge presented in the preceding chapters when it is integrated into certain key stages of the adaptation process (Chapter 3).

1. CONTEXT OF THE KNOWLEDGE SYNTHESIS

This first chapter contextualizes the projects implemented by the MTQ in Nunavik since 2009. The first sections present a profile of the population and infrastructure (Section 1.1) and a Nunavik climate profile (Section 1.2). Next, the hazards that affect coastal stability and the important characteristics for understanding the impact of climate change in Nunavik are provided (Section 1.3). Each of these characteristics was studied through measurement devices set up in several communities. A regional synthesis map catalogs the observation types, as well as the locations and acquisition periods for the observations. The chapter also shows why it is important to analyze both the hazards and the vulnerability of an environment to understand how the climate-related risks will change in the future (Section 1.4). The Knowledge Synthesis can be used as a starting point for adaptation (Section 1.5). A number of uncertainties remain at different levels of the various projects, but they will be resolved as our understanding improves (Section 1.6).

1.1 POPULATION AND INFRASTRUCTURE PROFILE

Located between the 55th and 62nd parallels, Nunavik is the northernmost region in Quebec.1 Its approximately 8000-km coastline (Ouranos, 2015b) borders Hudson Bay, Hudson Strait and Ungava Bay. According to the last Canadian census in 2016,2 the region is home to 12,000 Inuit residents and a little over one thousand non-Inuit residents. The populations have grown significantly (four times faster than the Quebec population between 2011 and 20163), and increased land occupancy and sea transport is expected in the context of the various Nunavik socioeconomic development agreements between Inuit communities, Makivik Corporation, the Kativik Regional Government (KRG) and the Government of Quebec (Comtois, 2020). This will potentially require expanded marine and coastal infrastructure, or the construction of new infrastructure better adapted to the demographic changes. The population is distributed in 14 coastal communities (Figure 1) on Hudson Bay, Hudson Strait and Ungava Bay. These remote communities are connected by air and maritime transportation, and there are no roads connecting communities to each other or to the rest of the province. Air transport (provided by Air Inuit and First Air) is used to move people and light cargo between the 14 communities, as well as between Nunavik and Montréal and Québec City. Air travel between communities takes place twice a day, stopping at each community from Salluit towards Kuujjuarapik or Kujjuaq, and ending in Montréal or Québec City. Sea access is essential for traditional hunting, fishing, subsistence gathering and leisure activities and for economic development.4 Indeed, for the conveyance of heavy and non-perishable goods (e.g. construction materials, vehicles, everyday consumer

1 Since Nunavik is located between 55°N and 62°N (Ivujivik is the northernmost community in Quebec), it is not considered part of Northern Canada in Environment and Climate Change Canada’s report (Bush et al., 2019). 2 Statistics Canada Census, 2016. Data retrieved on February 27, 2020 at https://www12.statcan.gc.ca/census- recensement/2016/dp-pd/prof/index.cfm?Lang=F. 3 Between 2006 and 2011, the population of the 14 Nunavik communities increased by 11.7%, and between 2011 and 2016, it increased by 12.7% (Statistics Canada 2016 Census). Meanwhile, the population of Quebec increased by 4.9% and 2.8% respectively (Institut de la statistique du Québec, 2019). 4 Makivik Corporation, 2019. Marine Infrastructure Program, overview on http://www.makivik.org/marine- infrastructure/ 2

goods, fuels, etc.) during the summer and fall, sea transport is preferred by communities and the mining industry for the supply of mining sites and ore exports (Comtois et al., 2020).

Between 1999 and 2011, marine infrastructure (breakwaters, access ramps, deep-water wharfs) was constructed across Nunavik’s northern communities, allowing for safe and reliable sea access for small vessels, the optimization of operations tied to maritime services, and the socioeconomic development of communities, all while facilitating traditional Inuit activities (Ministère des Transports du Québec, 2011). Between 2009 and 2020, measurement devices were set up off the coasts of different communities in order to provide data time series to the different models. The regional synthesis map shows the locations of the Nunavik communities, the current and past data collection sites, the types of collected data, the names of the networks that allow station maintenance as well as data compilation and dissemination, and the data

collection periods (Figure 1). This map reveals the numerous efforts deployed over the last decade to improve the understanding of the region’s hydro-climatic dynamic. It also shows the communities equipped with different kinds of measurement devices that made possible the collection of a large number of datasets over a relatively long period of time (Kuujjuarapik and Quaqtaq) and the communities not equipped with measurement facilities (Inukjuak, Salluit, Kangirsuk, Tasiujaq, Kangiqsualujjuaq and the Ungava Bay Marine Region). The integrity of the marine and coastal infrastructure could be affected by extreme weather events. The strength and frequency of these climate hazards could lead to damage to this infrastructure if they increased in the future. While the infrastructure was designed to resist “normal” weather events, it was not designed to resist future climate conditions (Comtois, 2020). Indeed, until now, natural risk levels were calculated according to the hypothesis that previous climate conditions would continue into the future. However, in the context of climate change, we cannot base the actions necessary today solely on past events if we are to anticipate future conditions (Ouranos, 2015). Since the effects of climate change are already being felt in Nunavik, the Government of Quebec wanted to evaluate future changes on the coast of Nunavik using climate projections. These future changes can be used to determine cost variations with respect to coastal infrastructure maintenance, rehabilitation and relocation, which will be important to measure. In the same manner, climate change will have repercussions on Inuit investment costs and sociocultural organization that are also important to anticipate.

1.2 NUNAVIK CLIMATE PROFILE

The Hudson Bay, Hudson Strait, Ungava Bay and Nunavik region experienced an increase in winter temperatures of 1.5°C per decade during the 1987-2016 period. During this period, summer temperatures increased by 0.5°C per decade. Forecasts indicate that this warming will continue. The region could experience annual air temperature increases of 4 to 5.1°C in the period from 2046 to 2064, and 4.1 to 7.5°C in the period from 2076 to 2100. The warming in winter months could be even more significant, with increases between 5.5 and 5.8°C in the 2046 to 2064 period, according to RCP4.5 and RCP8.5 (see Inset 1). According the same scenarios and for the same future periods, increases in air temperatures could reach 2.0 to 2.5°C in the summer (Integrated Regional Impact Studies (IRIS) 4, 2nd edition – personal communication). The warming will lead to an increase in total annual precipitation (total quantity of rain and snow) (Ouranos, 2015a). There is more uncertainty with respect to annual average precipitation, but there is consensus about the fact that precipitation has increased by 3% per decade since the 1950s in the whole region. (IRIS 4, 2nd edition – personal communication). According to the same scenarios (RCP4.5 and RCP8.5), in the period from 2046 to 2064, average annual precipitation could increase by 20 to 35% in Nunavik. Daily precipitation increases of between 0 and 0.5 mm per day are expected throughout the year. Extreme precipitation could increase in the entire region by 5 and 10 mm per day for the 2046-2064 and 2076-2100 periods respectively, according to RCP8.5. Changes in extreme precipitation are expected to be consistently greater than changes in annual precipitation. While total precipitation has increased, the warming appears to have contributed to a 13% decline in total solid precipitation (snow, hail, freezing rain) between 1980 and 2014. This decline in solid precipitation is even more pronounced in October and November (IRIS 4, 2nd edition – personal communication).

Summary of key climate data for Nunavik (IRIS 4, 2nd edition – personal communication) Recent past RCP4.5 RCP8.5 Period Season 1987-2016 2046-2064 2076-2100 2046-2064 2076-2100

↗ 0.5°C Av. summer air temp. (JJA) ↗ 2.0°C ↗ 2.5°C per decade ↗ 1.5°C Av. winter air temp. (DJFM) ↗ 5.5°C ↗ 5.8°C per decade ↗ 0.5 to 0.9°C Av. annual air temp. ↗ 4.0°C ↗ 5.1°C ↗ 4.1°C ↗ 7.5°C per decade Max. precipitation (mm/day) ↗ 5 ↗ 10 Annual av. daily precipitation Between 1 et 1.6 ↗ 0.33 ↗ 0.37 ↗ 0.39 ↗ 0.69 (mm/day) Av. annual precipitation (%) ↗ 20 to 35

Duration of snowfall (days) ↘ 23 ↘ 31 ↘ 31 ↘ 63

Thus, the changes in temperature and precipitation are already being felt in the Nunavik region and will escalate by the end of the 21st century, especially in the fall and winter, particularly in the northern part of Nunavik (IRIS 4, 2nd edition – personal communication).

Inset. IPCC Climate scenarios In 2013, the IPCC confirmed in its fifth report that it is extremely probable that global warming is linked to GHG emissions released by human activity (combustion of fossil fuels [coal, gas, natural gas], forestry and farming practices and urbanization). In 2014, the IPCC established four climate scenarios (Representative Concentration Pathways - RCP): RCP2.6, RCP4.5, RCP6.0 and RCP8.5. The scenarios are based on global economic and demographic growth projections and describe plausible future GHG emissions. RCP8.5 – The most pessimistic scenario, where it is assumed that GHG emissions are very high as a result of no significant action being taken by governments to reduce emissions (IPCC, 2019). RCP4.5 and RCP6.0 – The scenarios associated with moderate GHG emissions. They assume vigorous action taken by all governments to reduce emissions. RCP2.6 – The only scenario that leads to a temperature increase of less than 2°C by 2100 (IPCC, 2019). It assumes a very significant reduction in fossil fuel use and numerous mitigation measures aimed at reducing net GHG emissions to zero around the middle of the 21st century. Currently, this scenario is not considered realistic. In 2007, in its fourth report, the IPCC used other scenarios called SRES (Special Report on Emissions Scenarios). They use the same foundation with respect to demographic, societal, economic and technological changes. The SRES scenarios included four scenario families: A1, A2, B1 and B2. The SRES A2 emissions are comparable to the radiative forcing in RCP8.5. These different GHG emissions scenarios are used by climate models to produce climate projections. These projections in turn make it possible to estimate how the emissions scenarios will affect climate normals, as well as the variability and trends of climate variables for a given region and period (Charron, 2016).

1.3 EVALUATING THE CHANGES IN CLIMATE AND HYDROLOGICAL HAZARDS AND THEIR ASSOCIATED CHARACTERISTICS

The government of Quebec coordinated several studies5 on the impact of climate change on Nunavik’s coastal and marine environment between 2009 and 2020 in the context of the PACC 2006-2012 and the PACC 2013-2020. These studies, which are summarized in this document, were conducted or are ongoing at the INRS-ETE and ISMER research centres, at Ouranos and at the consulting firms Environnement Illimité Inc., CIMA+ and Lasalle NHC. Regardless of a given coast’s latitude or geophysical characteristics, the two main physical impacts that could occur are submersion and erosion (Boyer-Villemaire, 2016). The damage to the natural environment (i.e. disappearance of beaches as a result of submersion, erosion of coastal cliffs) and coastal infrastructure and facilities (i.e. submersion of buildings, washout of protective structure bases) could make access to the region and travel along the coast more difficult. To evaluate the changes that affect the Nunavik coastline, the studies mentioned above, which form the basis of the Knowledge Synthesis, each evaluated the changes with respect to the characteristics of particular natural hydro-climatic hazards6 (water level, waves, extreme winds and ice) that could intensify in the future and increase coastal disruption. The main disruptive coastal hazards that were studied are: • Storm regimes, even if storm intensity is lower • Extreme water levels • Strong wave frequency • Sea ice formation/disappearance dates and ice thickness The identification of hazard characteristics (average, intensity, occurrence probability (or frequency), spatial location or range and the potential duration of their impacts) is necessary to better understand the nature of hazards and their effect on the exposed environment. It is then a matter of collecting observation series with respect to these disruptors. These observations make it possible to assess the extent of changes that have already taken place. In Nunavik, few or no long observation series on water levels, shore and sea ice, and waves, have been collected, and few have been collected with respect to the climate conditions that influence them. Thus, starting in 2009, measurement equipment was installed at different sites in Nunavik in order to measure water levels at sea and on the coast, detect the presence of coastal and sea ice, and measure wave height and the meteorological conditions that influence these hazards (Figure 1). These observations were then processed and used to validate (or calibrate) the hydrodynamic models (for sea ice, waves and water levels) (Technical Synthesis) that produced simulations for the recent past (1989- 2009), near future (2040-2069) and distant future (2070-2099) (Technical Synthesis). Table 27 compiles the

5 These studies are used in the Knowledge Synthesis and the Technical Synthesis. References can be found in the bibliography (Clerc et al., 2012; Massée and Villeneuve, 2013; Massé and Gallant, 2016; Savard et al. 2014 et 2016; Bernier et al. 2016, 2017, 2018, 2019; Ropars, 2014; Senneville and St-Onge, 2013; Senneville, 2018 et Neumeir et al. 2019). 6 The interactions between the studied climate and hydrological (geophysical) hazards are explained in the Technical Synthesis. 7 For more information on the various data acquired, the equipment used, the measuring devices and their limitations, the gaps in data, the necessary processing operations, and the models produced, the reader may consult the Technical Synthesis. (Technical Synthesis). 6

data types and hydro-climatic time series simulated from the different observation series for the 14 communities in Nunavik and Deception Bay. Table 2 also indicates the reanalyses (long series of spatio- temporal, continuous and uniform data, consistent between different variables; when the observed data is too discontinuous, they are used in models as data for the recent past) and climate simulations used as inputs in hydro-climatic models.

List of simulated data for the different studied sites in Nunavik

Reanalyses Average water Maximum and Maximum water 1979-2012 levels in 2100 minimum water levels, Sea ice Name of Climate With post- levels, 100-year 2-year and 75-year Concentration Waves measurement simulations glacial return period return period Thickness 1980-2010 station 2011-2040 rebound 1980-2009 1980-2009 Volume 2071-2100 2041-2070 2040-2069 2040-2069 2041-2070 2071-2100 2070-2099 2070-2099 Hudson Bay Kuujjuarapik X X X X X X Umiujaq X X X Inukjuak X X X X Puvirnituq X X X Akulivik X X X Ivujivik X X X X Hudson Strait Salluit X X Deception Bay X X X Kangiqsujjuaq X X X Quaqtaq X X X X X Ungava Bay Kangirsuq X

Aupaluk X X Tasiujaq X Kuujjuaq X X X X Kangiqsualujjuaq X

All the collected data (Figure 1) were processed and used to validate the hydro-climatic models behind the simulations produced for the recent past (1989-2009 period), near future (2040-2069 period) and distant future (2070-2099) (see Technical Synthesis). The ice concentration simulation results are available in the ice condition atlas produced by Senneville in 2018, where ice conditions are catalogued on a monthly basis for future periods. The wave and storm simulations are still being studied and cannot be represented on a map either. The portion of the results obtained for storm surges is presented in Figure 4 (Section 2.1.3).

1.4 EVALUATION OF THE VULNERABILITY OF THE COASTAL ENVIRONMENT AND OF COMMUNITIES

In order to be comprehensive, a vulnerability study should include infrastructure and population characteristics in addition to the hazard characteristics studied in this document. Indeed, these hydro- climatic hazards only account for part of the probability or risk of significant modifications to the environment. The intrinsic vulnerability factors of the environment also contribute to this risk. The intrinsic vulnerability, which results from physical factors as well as social, economic and environmental factors, predisposes elements exposed to potential hazards to harm or damage (Morin, 2008b). In northern regions, the economic development potential and demographic growth, combined with the impacts of climate change, make the coastal vulnerability issues more complex (Ouranos, 2015b). For the coastal environment of Nunavik villages, the following vulnerable elements must be evaluated: • Environmental: the geomorphology of the coast, i.e. its shape (bay, fjord or mouth) and the presence of islands off the coast; the soil type, i.e. sedimentary soil (sandy or clay) or rocky coast; and the degradation of the permafrost (Allard et al., 2010; Allard et Lemay, 2013; L’Hérault et al., 2013). • Physical and economic: the number and age of the infrastructures, as well the materials used and their resistance to natural phenomena, the distance from the coast and the cost of repair or relocation, should all be considered in the context of hazard intensity and frequency. • Social: the historical attachment to the environment, social organization and demographics, and the level of sensitivity to impacts and possible solutions provide information on the adaptative capacity of the populations. In order to understand how the climate-related risks will change in the future, it is important to analyze both the hazards and the vulnerability of an environment. Infrastructure and population characteristics could also be added to this analysis.

1.5 THE KNOWLEDGE SYNTHESIS AS A STARTING POINT FOR THE ADAPTATION PROCESS

Adaptation is a process by which communities and ecosystems adjust to the effects of climate change in order to limit the negative consequences and take advantage of potential benefits (Larrivée, 2010). As Figure 2 shows, the implementation of adaptation measures must be supported by a robust understanding of climate change and the risks (consequences) that it could create in the region’s systems (natural and human). This synthesis can be used as a starting point for the marine and coastal environment adaptation process. It begins by summarizing the scientific work that evaluates the probable changes in the hydro-climatic hazards affecting Nunavik’s coastal and marine environments in the context of climate change, while also examining the anticipated impacts on the coast at different scales of analysis, from the regional (Nunavik) to the local (communities) levels (Chapter 2). Next, it demonstrates the utility of the understanding presented in Chapter 2 when integrated to certain key stages of the adaptation process (Chapter 3).

Figure 2. The adaptation process according to Ouranos, 2020 1. Identify; 2. Prepare; 3. Implement; 4. Adjust.

1.6 DISCLAIMER ABOUT THE UNCERTAINTY OF THE FINDINGS IN THE VARIOUS REPORTS PRODUCED BY THE MTQ AND ITS PARTNERS

This Knowledge Synthesis is based on a number of scientific research reports, and the methodologies and detailed uncertainties associated with these reports are summarized in the Technical Synthesis. For more complete information, the reader may consult the research reports by referring to the bibliography. The presented synthesis data should be verified by the user before being used for the design of works, navigation or other activities requiring exact data. Similarly, the extreme level statistics are intended for research and coastal planning applications. The use of the data for the design of works is the user’s responsibility. The main uncertainties inherent to any prospective climate study are indicated below. 1. Short-term or intermittent observational data. The data are used to calibrate the models that provide synthetic datasets. They can also be used to develop parameters that allow climate models to represent certain phenomena that cannot be correctly treated by their basic equations (Charron, 2016). If there are few data points (as in the case of wave regime) or the data spans too short a period (as in the case of extreme water levels), the data will underrepresent the variety of possible conditions. Furthermore, short-term or intermittent climate data cannot provide a reliable portrait of the observed climate (normals, interannual and decadal variability, extremes), which would require several decades of data. Thus, the developed parameters may prove to be of limited effectiveness. The results of the hydrodynamic and climate models cannot be evaluated with a few cases or episodes (with potentially little variation), and more importantly, their climate statistics cannot be correctly validated. To overcome the gaps in observed data, reanalyses are developed by meteorological research centres using their observation databases (Technical Summary). 2. Time and budget constraints. As a result of these constraints, it is not possible to carry out a very large number of calculations. However, additional calculations can always be carried out at a later date if there is a demand for more reliable projections for total extreme levels, wave height or ice conditions in the future climate (Massé and Gallant, 2016; Senneville, 2018; Neumeier, 2019). 3. Models are simplifications of reality. The climate is a very complex and nonlinear (chaotic) system, which means that the interrelations between the different subsystems are not all well understood or represented in the models. The models vary significantly with respect to their degree of simplification and the manner in

which they represent smaller-scale physical phenomena (Charron, 2016). However, their realism is constantly improving as a result of the inclusion of new processes, the perfection of the numerous existing processes and the use of more recent and up-to-date reanalyses aimed at improving model parameters (Technical Synthesis). 4. The magnitude of climate change in the future depends on the GHG emissions scenario (IPCC, 2014). These scenarios, which are all considered plausible (Charron, 2016; IPCC, 2019), depend on changes in economic processes and political involvement in the establishment of tools aimed at reducing GHG emissions (Inset 1). With each new IPCC report, these scenarios are updated according to political and economic changes. Thus, the results that use these scenarios are valid for the scenarios available at the time of the simulations. The magnitude of climate change falls in the range of the climate projections resulting from the two most extreme scenarios (RCP2.6 and RCP8.5 inset) in the selected time periods. 5. Results projected over 30-year periods. These projections should not be viewed as predictions that could come true in these precise time periods, but rather as typical conditions that are likely to appear in the approximate period indicated by these windows of time. 6. Comparisons based on projections made using different GHG scenarios for different time periods. When comparing results produced by different models, it is important to verify that the projections are based on the same scenarios (RCP or SRES) and were made for the same reference period and time span (30 years). Some research teams use 20-year time periods. 7. Spatial uncertainty. UQAM-Ouranos’ regional climate models (RCMs) provide climate data for the Nunavik region at 45-km and 25-km resolutions. The ice model uses 10-km cells, and the resolution is increased to 1 km near the coast. These spatial resolutions are sometimes too crude to represent certain very local effects that occur at the community level. Although the satellite images studied for ice conditions have a resolution of 30 m, they must be aggregated using the ice model cells (10 km or 1 km) for validation, and this comes at the price of certain details about the ice that could be of interest to communities.

2. KEY FINDINGS

The studies conducted in the Arctic on the hydro-climatic hazards that interest us here are far more numerous than those covering Hudson Bay and Nunavik. While it is not possible to transfer all the findings obtained for the Arctic, some of the studies offer interesting avenues for understanding the evolution of certain hazards in Nunavik, as the two environments have some similarities, especially the presence of sea ice. For example, it has been shown that in the Arctic, the disappearance of sea ice will be continuous for half a century (IPCC, 2019). The annual probability that the Arctic Ocean will be free of ice in September around 2050 is 1% if the increase in global air temperatures is limited to 1.5°C. This probability increases to between 10 and 35% in the case of a global air temperature rise of 2°C (IPCC, 2019). What is more, extreme water level events that had a return period of once per century in the recent past should happen at least once a year around the middle of the 21st century, according to all the RCPs (IPCC, 2019). With respect to storms, the most recent analyses focused on the Arctic Ocean and did not include Hudson Bay (see Technical Synthesis). In the Canadian Arctic Archipelago and Baffin Bay, winter storms (DJF) could be more frequent, but mixed conclusions have been drawn about their size and intensity. The Foxe Basin area could also see an increase in summer storms (JJA) by the end of the 21st century, unlike other Arctic regions (Akperov et al. 2019). These results are based on analyses combining models that are more recent, more numerous and that offer a higher resolution than those used in most of the previous work on the Arctic reported by Ford et al. (2016). As long as there is little or no data for Nunavik, we can carefully draw inspiration from the Arctic data. These two environments will experience warming far more significant than the global average and have in common the presence of sea ice and permafrost. However, they differ considerably with respect to the atmospheric circulation that determines storm formation. Thus, the results obtained for the Arctic are not sufficient to understand or know what is occurring in Nunavik more specifically. The studies coordinated by the MTQ thus contribute a new and distinct understanding of the characteristics of potential climate and hydrological hazards in Nunavik. Chapter 2 takes stock of the knowledge generated by recent studies on the impact of climate change on the hazards that global warming could generate on the Nunavik coastline. For each hazard, one key message summarizing the hazard characteristics is provided, followed by a more detailed explanation. The findings are presented at the regional scale for all of Nunavik (Section 2.1) in order to establish a general profile of the main findings, which apply to the entire region. Next, the findings are presented by ocean region (Hudson Bay, Hudson Strait and Ungava Bay) (Sections 2.2, 2.3, 2.4) in order to highlight each region’s main particularities. The key messages are then presented in a detailed manner for each of the communities in order to allow readers to easily consult the section dealing with a particular community of interest (Sub- sections 2.2.2., 2.3.2 and 2.4.2). A local synthesis map was created for the nine communities for which simulations were produced, as well as for Deception Bay. An example of this type of map along with a brief explanation of how to read it is provided in the section on Kuujjuarapik. The nine other local synthesis maps are included in the Appendix of this document. The presentation of communities starts with the southernmost community on the southwest coast of Hudson Bay, Kuujjuarapik. It follows the coastline, heading north and east and ending with Kangiqsualujjuaq, which is the furthest south and east of Ungava Bay. It should be noted that hazards are not documented in the same manner for all communities (Figure 1), and that other complementary studies carried out at the local level are mentioned in Chapter 3 (Allard et al., 2020a, b and c).

2.1 NUNAVIK

2.1.1 GENERAL PROFILE

The coastal environment (natural environment and infrastructure) can be damaged by submersion and/or erosion caused by hydro-climatic hazards (water levels, waves, winds and ice) according to its vulnerability (which depends on the physiography of the coastline and the presence of natural protection) and exposure level (intensity and duration). The different impacts that could affect the various Nunavik coasts are briefly described in the following paragraphs. The first describes impacts caused by tidal range, and the following describe impacts according to hydro-climatic risk. The relations between the different studied hazards are explained in the Technical Synthesis. • The tidal range is small on the coast of Hudson Bay, moderate on Hudson Strait and large on Ungava Bay (Figure 3). A small tidal range indicates sensitivity to flooding by submersion at low tide when a positive storm surge occurs. On the other hand, a site with a large tidal range indicates sensitivity to submersion only in cases when large positive storm surges coincide with high tide.

Figure 3. Tidal range in Nunavik (CanCoast data product, source data from Canadian Hydrographic Service). (Boisson, 2019).

• During a storm, the water level increases and waves are stronger, increasing the risks of submersion and erosion. The power (wind speeds in storms exceed 90 km/h) and the speed of the storm’s movement can be increased by the fetch (available space allowing winds to accelerate) and the absence of sea ice. Northern storms move more slowly than storms in the south of Canada, which further increases the duration of destructive wave exposure for northern coasts (Ford et al., 2016). Furthermore, even weaker storms can lead to cumulative damage through repetition if they appear

more frequently (Ropars, 2014). In the context of new storm studies and a more up-to-date understanding, the findings on storm presence in Hudson Bay produced by Savard et al. (2014) are considered too fragmentary. They must be updated using a larger number of more recent models that have demonstrated a better ability to simulate processes in an Arctic context that are important for Hudson Bay. See the Technical Synthesis for more details. • Stronger and more frequent waves (or breaking waves) and their approach angle (the direction of the wind and orientation relative to the coast) determine the coastal exposure. The modelling of waves in Nunavik is still ongoing, and several new findings are expected for 2021. • Very high total water levels can lead to more frequent coastal submersion. Meanwhile, very low total water levels can lead to receding water on a more frequent basis and over larger distances. • In winter, the complete landfast ice (or stable sea ice) cover diminishes the fetch and protects the coast against erosion. However, the beach and backshore, including the dunes, do suffer from erosion, and the maritime infrastructure can be damaged when the unstable sea ice linked to the partial cover in the spring and fall is pushed onto the shore (ice push) by strong storm waves and winds (Atkinson et al., 2016; Boisson, 2019; Bush et al., 2019). The ice piles up on the sublittoral bars, forming banks below and above the high tide zone, and digs into or shifts sublittoral sediments towards the land eroding the coast (bays and straits). As a result of rising temperatures, the period of partial and unstable ice cover is longer. Flooding through submersion is particularly relevant for land and infrastructure situated in low areas. Waves or ice could strike the top of the infrastructure and cause damage if it is not robust enough to resist the hydraulic pressure. Waves can submerge coasts by getting through the infrastructure and can also damage unprotected coastal facilities (Boisson, 2019). In the most extreme cases, the submersion can penetrate further into the coast. With the water’s retreat, coastal erosion, the undermining or washout of infrastructure bases and the loosening of rocks or other elements can occur. Sand accumulation at the mouths of rivers requiring dredging can compromise access to infrastructure and prevent the entry and exit of boats to and from ports (Boisson, 2019). Navigation activities can become perilous, and since the relative sea level is much lower than normal in these cases, the risk of grounding is increased. In general, access to the region can be compromised, both for traditional and harvesting activities and for industrial activities (Ouranos, 2015b). The impacts listed above can be more or less significant depending on the intensity, frequency and repetition of the hazards that cause them. As a result, it is important to consider the hazards and their characteristics in future periods, both in the near future (2040-2070) and distant future (2070-2100), in order to be able to adjust adaptation measures to the probable impacts. In the near future (2040-2070), the average and extreme water levels, wind characteristics and ice presence could increase the probability of high waves (Ford et al., 2016). In the distant future (2070-2100), projections show that in a large part of Nunavik, the relative sea level will be low, as the effect of post-glacial rebound will be greater than the rise in sea levels. Thus, extreme high water levels could be reduced (Ford et al., 2016). As a result of this, the impacts tied to the fall in the relative water level will gradually become more significant at the end of the century (Boisson, 2019).

2.1.2 RELATIVE SEA LEVEL

The rising relative sea level affects the entire North Atlantic. However, it will have a small effect on the coasts of Nunavik by the middle of the 21st century, then no effect at the end of the 21st century. Indeed, as a result of the compensating impact of post-glacial rebound, which is very significant in

the region (0.6 to 1.4 cm/year), Nunavik could experience a relative sea level drop of 40 to 90 cm depending on the community and the considered emissions scenario.

Higher sea levels lead to an increase in coastal submersion and erosion, depending on the physical nature of the coastline. For example, in communities with a very low tolerance to this type of risk (low altitude, silty and clayey soil surface, significant fetch), like Aupaluk and Inukjuak, higher sea levels could cause significant damage. As a result, the sea level change projections are very important for predicting the risks for populations, infrastructure planning and maintenance, and environmental management (Bush et al., 2019). According to RCP8.5, the IPCC (IPCC, 2019) expects the global relative sea level (or the absolute sea level averaged over the entire planet) to rise by 71 cm between 2081 and 2100, and by more than 84 cm by the end of the 21st century. In the Arctic, however, post-glacial isostatic adjustments will reduce the predicted rate of change for sea levels (Ford et al., 2016). On the coasts of Nunavik, post-glacial rebound is currently occurring at a rate of 0.6 to 1.4 cm per year depending on the considered coast. The combination of this effect with that of rising temperatures could lead to a fall in the relative water level of 40 to 90 cm by the end of the 21st century, depending on the community and scenario (Table 3). Thus, by the 2070-2100 period, the effects of post-glacial rebound will outweigh the impact of higher global sea levels for a large part of Nunavik. • The maximum height and frequency of extreme high water levels will decrease. • The anticipated damages will result mainly from extreme low water levels, which could be even lower and more frequent than in the recent past. In addition, 65 cm could be added to the global average sea level if a portion of the West Antarctic Ice Sheet melted by the end of the 21st century (James et al., 2014). However, even with this water, the effect of post- glacial rebound would remain high in Kuujjuarapik and Inukjuak, where water levels would be unchanged, while in Kuujjuaq and Salluit, water levels would increase by 40 cm (Table 3). By mid-century, in the 2040-2070 period, the effects of post-glacial rebound will not be great enough to compensate for the impact of storm surges. Also, the effects of positive surges, such as coastal submersion, could manifest.

Relative sea level (cm ± 95% confidence interval) according to climate scenarios RCP4.5, RCP8.5 and RCP8.5 with the collapse of the West Antarctic Ice Sheet (WAnt), at certain sites in Nunavik for the 2081-2100 period, as compared to 1986-2005. These scenarios take post-glacial rebound according to James et al. (2014) into account. Water level reached by 2100 Water level reached by Water level reached by 2100 Site according to RCP4.5 2100 according to RCP8.5 according to RCP8.5 + WAnt Kuujjuaq -45.9 ± 27.1 cm -36.2 ± 33.2 cm 40 cm

Salluit -42.0 ± 23.7 cm -32.5 ± 29.5 cm 40 cm

Inukjuak -71.2 ± 21.7 cm -58.8 ± 25.4 cm 0 cm

Kuujjuarapik -87.1 ± 23.4 cm -72.6 ± 23.4 cm 0 cm

2.1.3 EXTREME WATER LEVELS, FREQUENCY AND SEASONALITY

Nunavik could be affected by more frequent extreme water levels by the 2040-2069 period. While extreme high water levels (positive storm surges) could become progressively lower by the end of

the 21st century, extreme low water levels (negative storm surges) could also fall because of the additional effect of the drop in relative sea level. None of the water levels below take post-glacial rebound into account. Extreme water levels could increase or decrease by up to ±1 m. Between 1980 and 2010, negative storm surge water levels fell (by 54 cm to 1 m). Compared to this reference period, negative surge levels could be even lower in the future, resulting in water levels lower by 1.5 cm by 2040-2069 and 6 cm by 2070-2099. Between 1980 and 2010, positive storm surge water levels rose (by 68 cm to 1.5 m). Compared to this reference period, positive surge levels could be even higher in the future, resulting in water levels higher by 10 cm. The frequency of extreme water levels could increase by 2040-2069. The extreme levels that currently occur once every 100 years could occur once every 35 or 50 years. The storm surge season could lengthen in the near future and be even longer in the distant future. By the end of the 21st century in Nunavik, the lower relative sea level will lead to reduced extreme high total water levels (Savard, 2016). The word “total” indicates that the level includes the relative sea level, the tide, storm surges and coastal processes like the wave setup and run-up. Since the studies summarized in this document did not consider coastal processes, it would be useful to consider them in future studies (Chapter 3). The adjective “extreme” is used to describe water levels that reach rare maximum or minimum levels (i.e. levels with a 100-year return period). More frequent very high water levels could lead to more frequent erosion and submersion. On the other hand, the receding waters and very low water levels could cause the formation of beaches. The extreme total water levels that could occur in coastal systems throughout the 21st century should generally be considered for infrastructure adaptation, as these levels occur when strong storm surges coincide with spring or neap tides and the wind is blowing towards the coast (Boyer-Villemaire et al., 2016). The study on storm surges (Massé and Gallant, 2016) focused on three characteristics: the levels likely to be reached by positive and negative storm surges, their probability of occurrence (or frequency) and their seasonality (the season with the highest probability of occurrence). 1- Storm surge extreme water levels. The water levels caused by the tide are understood and consistent, and thus predictable. However, the additional storm surge levels are not well understood and could have serious consequences. This is why Massé and Gallant (2016) subtract the tidal signal from the total simulated water levels to obtain the storm surge water levels (negative and positive storm surges). Only extreme water levels with 100-year return periods from 1980 to 2009 were considered in this chapter.8 All of the findings on extreme water levels presented in Figure 4 and Tables 4 and 5 are from simulations that did not include post-glacial rebound. Figure 4 presents two regional synthesis maps that illustrate the results of the storm surge extreme water level simulations: (a) negative storm surges and (b) positive storm surges. Water levels are said to be extreme when they reach very rare maximum heights with 100-year return periods. On these maps, the water levels are indicated in centimetres (cm). The negative and positive storm surge levels from 1980 to 2009 are indicated by a circular symbol, with levels closest to zero cm indicated in blue and the most extreme levels indicated in red. For the future periods, the variations in the extreme levels (i.e. the extreme level for the

8 Smaller return periods (75, 50 and 30 years) were also calculated and can be found in the report by Massé and Gallant (2016).

considered future period minus the extreme level from the recent past) are indicated by an arrow. A positive value indicates an increase, represented by a rising arrow, while a negative value indicates a decrease, represented by a descending arrow. If the variation is small (0 to 5 cm), the arrow is blue. If the variation is moderate (5 to 10 cm), the arrow is green. If the variation is large (10 to 15 cm), the arrow is orange. Finally, if the variation is very large (15 to 20 cm), the arrow is red.

(a) (b)

Figure 4. Extreme storm surge levels, with a 100-year return period, from the recent past (1980-2009) and (a) negative storm surge and (b) positive storm surge variations for the 2049-2069 and 2070-2099 periods.

Table 4 shows the values of the extreme positive and negative surges (with a 100-year return period) that were simulated for the recent past and the two future periods for six communities on Hudson Bay, three communities on Hudson Strait and Kujjuaq on Ungava Bay.

Extreme negative and positive storm surge levels with a 100-year return period simulated for the recent past and for the near and distant future (Massé and Gallant, 2016) Negative storm surge levels (in cm), 100-year Positive storm surge levels (in cm), 100-year return period return period (2040- (2070- (2040- (2070- Study site 1980- 2040- 2069) 2070- 2099) 1980- 2040- 2069) – 2070- 2099) – 2009 2069 – 2099 – 2009 2069 (1980- 2099 (1980- 2009) 2009)

(1980- (1980- 2009) 2009) Kuujjuarapik -87.3 -80.3 7 -83.8 3.5 147 166 19 162.7 15,7

Umiujaq -94.2 -90 4.6 -95 -0.9 156 172 15.6 173 15,6

Bay Inukjuak -78.7 -81.4 -2.7 -90.5 -11.8 112 129.6 17.6 132.9 20.9

Puvirnituq -91.7 -96.9 -5.2 -108.4 -16.7 129 139.8 10.8 143.2 14.2

Hudson Akulivik -79.6 -85.9 -6.3 -93.5 -13.9 118 127.2 9.2 130.4 12.4 Ivujivik -66.5 -74 -7.1 -80 -13.3 101 111 8.1 107 6.1

Deception Bay -65.1 -68 -2.9 -71.5 -6.4 75.1 81.7 6.6 78.5 3.4

Kangiqsujjuaq -57.8 -58.6 -0.8 -59.9 -2.1 68.3 78.4 10.1 71.7 3.4 Strait Hudson Hudson Quaqtaq -54.3 -55 -0.4 -55 -0.4 69.7 81 9.9 73 4.1 Ungava Kuujjuaq -71.8 -70.6 1.2 -69.4 2.4 101 91.9 -9.1 103.4 2.4 Bay Nunavik average -77,0 -76.1 -1.3 -80.7 -6.0 107.7 117.9 9.8 117.6 9.8 *From Figures 5.18a and 5.18b of the report by Massé and Gallant, 2016.

In the near and distant future, negative storm surge water levels could fall (Figure 4a and Table 4) and positive storm surge levels could rise (Figure 4b and Table 4) in all of Nunavik. In Hudson Bay and Ungava Bay, negative storm surge water levels were lower than in Hudson Strait, while positive storm surge water levels were higher than in Hudson Strait (Figures 4a and 4b and Table 4). Along Hudson Bay, negative storm surges with 100-year return periods could lower the minimum water level by 0.7 m to 1 m by 2040-2069 and by 0.8 m to 1.1 m by 2070-2099 (Table 4). Extreme positive storm surges are everywhere projected to be greater than 1 m (between 1.1 m and 1.7 m), with a maximum of 1.7 m in Umiujaq (Table 4). However, on the coasts of Hudson Bay, where the tidal range is small (4 m to 50 cm or less) (Figure 3), extreme positive surge water levels have a significant impact on total water levels. Extreme positive surges will always cause flooding along all the coasts of Hudson Bay (Savard et al. 2016), whether or not they coincide with the spring tide. On the other hand, where the tidal range is large (from 4 to 10 m and more) (Figure 3) on the coasts of Ungava Bay and Hudson Strait, total water levels result mainly from the tide rather than from storm surges. Currently, in these regions, the infrastructure is adapted to withstand the high waters associated with exceptional spring tides (more than 10 m) without damage. Given that positive storm surges with a 100-year return period (the highest surges) may not exceed 1 m on these coasts in the two future periods and that it is possible that only Kuujjuaq could experience 1-m surges in 2070-2099 (Table 4), 1-m positive storm surges will only have an impact on the coastal environment and marine infrastructure if they coincide with a spring tide (10 m). Negative storm surges with a 100-year return period could lower the water level by up to 72 cm in the two future periods, which is equivalent to current negative storm surge levels (Table 4). Table 5 summarizes the main simulated results for storm surges (negative and positive surges) in all of Nunavik.

Extreme storm surge levels (negative and positive): Nunavik average, frequency and seasonality

Storm surges

Period Negative Positive

1980-2010 between -1 m and -50 cm between +68 cm and +1.5 m Average of 100-year return period extreme 2040-2069 ↘ 1.5 cm ↗ 10 cm levels in all of Nunavik 2070-2099 ↘ 6 cm ↗ 10 cm Return period of 2040-2069 50 years 50 years extreme levels with current 100-year 2070-2099 100 years 100 years return period ↗ July to August 2040-2069 ↗ December to January ↗ August to February Seasonality (seasonal ↘ February to March frequency) ↗October to February ↗↗ July to August 2070-2099 Appearing in March and summer ↘↘ December to January months

2- Return period of extreme water levels. To know the probability of extreme water level occurrence, the lowest or highest tide levels (high spring tide or low neap tide) must be considered in conjunction with extreme positive and negative storm surges, i.e. those that occur once every 100 years (Savard, 2016). In this manner, it is possible to rank the highest and lowest extreme water levels according to their probability of occurrence. The rarest are those that reach the most extreme high or low water levels. When a negative storm surge takes place at low tide, the water level is very low, and more rarely, when an extreme negative storm surge coincides with a low neap tide, the resulting total water level is extremely low (Table 6). Extreme simulations show that the probability of very low levels will be higher by 2040-2069 for the three studied communities, going from once every 100 years in the three communities to once every 50 years in Umiujaq, once every 30 years in Ivujivik and once every 45 years in Quaqtaq (Table 6).

Return period of extreme low total water levels in the current and future climates in Umiujaq, Ivujivik and Quaqtaq (from Figures 5.15, 5.14 and 5.13 in Massé and Gallant, 2016) Community Extreme low total Return period in current Return period from Return period from 2070- water level (cm) climate (years) 2040-2069 (years) 2099 (years) Umiujaq -152 100 50 100 Ivujivik -214 100 30 55 Quaqtaq -552 100 45 100

The very high water level that occurs when a positive storm surge coincides with high tide has a higher probability of occurrence than the extreme high water level that happens when an extreme storm surge (once every century) coincides with a high spring tide. Finally, the likelihood of an extreme storm surge (once every century) coinciding with the century’s highest high tide (once every century) is extremely low. The probabilities of extreme surge water levels show that they will be more likely to occur by 2040-2069 for the three studied communities, going from once every 100 years for the three communities to once every 40 years in Umiujaq, once every 35 years in Ivujivik and once every 50 years in Quaqtaq (Table 7). These same extreme water levels will be less frequent by 2070-2099 for Ivujivik and Quaqtaq but not for Umiujaq, where the probability of occurrence will increase to once every 35 years (Table 7).

Return periods of extreme high total water levels in current and future climates in Umiujaq, Ivujivik and Quaqtaq (from Figures 5.9, 5.8 and 5.7 in Massé and Gallant, 2016) Community Extreme high total Return period in current Return period from Return period from 2070- water level (cm) climate (years) 2040-2069 (years) 2099 (years) Umiujaq 209 100 40 35 Ivujivik 235 100 35 Over 100 Quaqtaq 530 100 50 100

However, the water level studies conducted by Savard, 2016, Massé and Villeneuve, 2013, and Massé and Gallant, 2016, did not take into account the fall in the relative level due to post-glacial rebound for any of the climate simulations. The rate of post-glacial rebound is very high in Umiujaq (1.15 cm/year), which could reduce the extreme water levels as well as the probabilities of occurrence presented above by compensating for the rise in global sea level. 3-Time span. The simulations by Massé and Gallant (2016) show that the seasonality of storm surges varies in the three maritime regions. During the 1980-2009 period, the extreme surge season occurred during the months of September, October and November in the entire study area. In the central part of Hudson Bay, there could be more extreme positive surges during this season by 2070-2099. Extreme negative storm surges occurred during the same season in the central part of Hudson Bay. However, negative storm surges tended to be more present in winter in northern Hudson Bay and in Hudson Strait, with March being the main month for this phenomenon (Massée and Gallant, 2016). By 2040-2069, the lowest negative storm surge levels (less than -0.60 m) could be more frequent from July to August and from December to January, while they could be less frequent from February to March. By 2070-2099, the lowest negative storm surges could be more frequent in spring and much less frequent in winter (as is currently the case) (Table 5). By 2040-2069, the largest positive storm surges (0.75 m to 0.95 m) could be more frequent from August to February in the entire region. By 2070-2099, they could reach 0.85 m to 0.95 m and be more frequent from October to February in the entire region (Table 5). In the distant future (2070-2099), positive storm surges could also occur in March, in the spring and in the summer, despite being practically totally absent during these periods in the current climate (Table 5).

2.1.4 STORM FREQUENCY

Future storm trends in Hudson Bay are not clear from our current level of understanding.

Since the changes in storm patterns in Hudson Bay are still poorly understood, we recommend studying them with the new climate models, which are more effective for this region (Technical Synthesis). The current understanding of storms in northern marine environments focuses mainly on the Arctic Ocean and was developed through the CORDEX ARCTIC ensemble simulations. While atmospheric circulation does not occur in the same manner in Hudson Bay and the Arctic Ocean, the occurrence of storms during landfast ice formation, when the ice is unstable and mobile, has drastic consequences on coastal environments in both cases. For example: • The ice displacement caused by strong winds blowing towards the coast could further reduce ice stability on the coast.

• Storms could increase water levels and the height and strength of waves, increasing the risks of submersion and erosion. • Whether or not there are changes in storm intensity and frequency, the projected sea ice changes could lead to cumulative damage through more frequent repetitive stress.

In the warmer future climate, the earlier melting of ice and the higher summer temperatures will also increase the heat reservoir that builds up in the waters of Hudson Bay during the summer. From June to August, the atmosphere is very stable because the very cold surface water cools the air in the lower atmosphere, leading to lower rates of storm formation (Savard et al., 2016). In the future, however, the heat reservoir could last into the fall, increasing atmospheric instability and postponing ice freeze-up. These two factors could lead to an increase in the intensity and/or number of storms (Savard et al., 2016), although other mechanisms are most likely also involved. This hypothesis will thus need to be validated with new simulations and the other relevant mechanisms. Recommendations are provided in the Technical Synthesis.

2.1.5 ICE CONDITIONS

The rapid increase in winter temperatures above Hudson Bay and Nunavik will lead to later ice freeze- up and earlier ice breakup. - The duration of the ice season could be reduced by more than six weeks by 2041-2070 and more than two months by 2100. - The increased mobility of the ice could lead to a higher probability of ice-related erosion. - The coasts of Nunavik could see a decrease in average ice concentration of around 30% in the month of May by 2040-2070. This concentration could be 10 to 40% lower in the 2040-2070 climate than it was for 1980-2010. According to the original simulations conducted by Senneville (2018), by 2040-2070, the ice is expected to be 15 cm thinner in November and 80 cm thinner in June. The thinner ice cover could break if strong winds and higher water levels occur simultaneously. In spring and autumn, the coasts could therefore be more exposed to erosion through ice pushes. • In winter, the complete stable sea ice cover diminishes the fetch and protects the coast against erosion (Table 8). • The spring and fall are problematic periods because of the presence of mobile ice. The partial and unstable (or mobile) sea ice cover increases the risk of erosion and can cause damage when strong winds push this mobile ice towards the coastline (Table 8). In spring, the mild temperatures activate thermal and mechanical ice cover degradation processes. By accelerating the maturation process, this degradation leads to ice floes that are more mobile. The atmospheric temperature can also delay the stable ice cover formation process in the fall (Clerc et al., 2012) – (see Section 2.3.2.4. on Quaqtaq). • However, during the fall ice freeze-up, the periods of mobile ice for the studied sectors are far shorter and the mobile ice far thinner than during the spring thaw (Clerc et al., 2012). Thus, the likelihood of ice-related damage in the fall is generally less high than in the spring. But in the context of interacting hydro-climatic parameters or detached pack ice that returns towards the coast as a result of a number of combined factors (waves, strong winds and wind direction, tides, etc.), there could still be a high likelihood of damage in the fall, as on the coasts of Hudson Strait, for example. • The coasts of Nunavik could experience a 30% decrease in average ice concentration in the month of May in the 2040-2070 period. This concentration could be lower by 10 to 40% in the 2040-2070 climate, as compared to the 1980-2010 period. The ice’s average thickness could decrease by up to 15 cm in November and 80 cm in June by the 2040-2070 period, as compared to the recent past.

Nunavik ice cover and duration risks in future periods, in winter, spring and fall

Ice stability Stable ice Winter Shorter complete cover, Coastal erosion protection Spring Mobile ice

Fall Longer partial cover, Increased erosion risk

In Northern Quebec and the Canadian Arctic, the decreased ice cover and shorter ice season could allow for a longer navigation season, while travel on the pack ice in the context of the traditional activities of local populations will become more dangerous (Lemmen et al., 2016). The duration of the ice season in the bodies of water surrounding Nunavik could be reduced by more than six weeks by 2041-2070, according to the SRES-A2 scenario (Senneville and St-Onge Drouin, 2013), and by more than two months by 2100, according to the same scenario (Joly et al. 2011). Climate simulations at a horizontal resolution of 10 km are currently the best source of information for ice concentration and thickness projections for the 2040-2070 period (Senneville, 2018b). Improvements to the ice model in order to obtain ice condition simulations on a more local scale are currently being studied (Technical Synthesis). The results of the simulations were compiled in the monthly climate atlas of Nunavik ice concentration and thickness (Senneville, 2018a). They also indicate a possible reduction in the concentration and thickness of sea and landfast ice (Senneville, 2018a). For example, Figure 5 shows that in December, the coasts from Ivujivik to Kangiqsualujjuaq had average ice concentrations of 90 to 100% in the recent past, while in the 2040-2070 period, the concentration could decrease to 40 and 60% on average. From 1980 to 2010, the coasts between Ivujivik and Inukjuak had average ice concentrations of 60 to 80%. By 2040-2070, ice concentrations could be practically non-existent (around 10%) in December.

Figure 5. Average concentration of sea ice in Nunavik in December 1980-2010 (a) and 2040-2070 (b) (Senneville, 2018a). 21

2.1.6 WAVES

No wave data existed for Nunavik before the CAIMAN project installed wave height measurement stations in 2017 in Kuujjuarapik and Quaqtaq for the Neumeier team (UQAR-ISMER). However, the processes that govern waves are understood and hypotheses can be made.

When powerful storms cause stronger winds, the number of positive surges and the amount and energy of waves reaching the coast may increase. Thus, in Nunavik, the shortening of the sea ice season by about 40 days would have the effect of increasing the total wave energy produced by storms reaching the coast. Also, by 2040-2070, average and extreme water levels, winds and ice presence could increase the likelihood of high waves occurring.

At the same time, however, the vulnerability of the coast to these waves could be reduced by the end of the 21st century as post-glacial rebound gradually lowers the relative sea level. In coastal systems, wave patterns are influenced by the direction and strength of the wind, especially during storms. They are also influenced by wind persistence (duration of wind), fetch, the presence of an ice cover, and physical obstacles (islands) in certain configurations (Boyer-Villemaire et al., 2016). As swells approach the coast, water depth decreases and the waves break. The shape of the waves is influenced by the typology of the seabed, the bathymetry, the water depth and currents. If the bottom is flat and gently sloping, the waves are flat and soft. On the other hand, if the bottom is steeply sloping, the waves are deep and powerful. The physical modeling of waves makes it possible to describe their shape according to the wind and the typology of the seabed. In Nunavik, in the winter, wave formation is inhibited by the presence of a sea ice and landfast ice cover. Thus, the total wave energy produced by storms reaching the coast will be increased by the shortening of the sea ice season by about 40 days. If storms become more frequent, there would be an increase in strong winds and a corresponding increase in wave height on the coast. Future storm intensity and frequency trends cannot be reliably established from our current level of understanding. However, with post-glacial rebound causing a gradual decrease in the relative sea level at the end of the 21st century, coastal vulnerability to strong waves may decrease. Two research projects are currently underway in order to gain a better understanding of wave patterns. The first is project CC09.2, “Suivi des conditions des glaces de rive et de vagues à l’aide de caméras et d’imagerie satellitaire à proximité d’infrastructures maritimes au Nunavik dans un contexte de changements climatiques: Kuujjuarapik, Umiujaq, Ivujivik, Baie Déception, Quaqtaq et Aupaluk (Monitoring landfast ice conditions and waves using cameras and satellite imagery near marine infrastructure in Nunavik in a climate change context: Kuujjuarapik, Umiujaq , Ivujivik, Deception Bay, Quaqtaq and Aupaluk) (2017-2020)” (Bernier et al., 2019). It will generate observations for a digital wave model developed at ISMER as part of the second project, called project CC16.1, “Suivi, analyse et modélisation des conditions de vagues en milieu côtier au Nunavik en fonction des conditions de glace dans un contexte de changements climatiques (Monitoring, analysis and modeling of wave conditions in Nunavik’s coastal environment as a function of ice conditions in a climate change context)” (Neumeier, 2019). It will allow for the acquisition of wave, ice current and bathymetric data near marine infrastructure in Quaqtaq and Kuujjuarapik. The datasets collected by these two projects between 2017 and 2020 will allow the ISMER digital wave model to be tested under different climate scenarios in order to assess the impact of climate change on wave regimes. Details on these two projects are provided in the Technical Summary.

2.2 HUDSON BAY

2.2.1 GENERAL PROFILE

GEOMORPHOLOGY OF THE COAST The coast of Hudson Bay is mainly composed of sedimentary soil. It is currently advancing toward the sea. Between Kuujjuarapik and Inukjuak, the coast of Hudson Bay is made up of cuestas interspersed with rich sedimentary environments. Between Inukjuak and Ivujivik, the coast is composed of low outcrops, small islands, pocket beaches and bays (e.g. Kovik) (Boisson, 2019). Boisson (2019) has identified numerous spits in formation southeast of Hudson Bay that will eventually attach themselves to small rocky islands that are part of Nunavut.

HAZARDS IN THE CURRENT CLIMATE The Hudson Bay coast is very sensitive to flood risks, but not to erosion. Storms, high water levels, high waves and a longer ice-free period will amplify these risks. Ice presence decreased by 10.5% between 1968 and 2010. Because of its low to moderate tidal range (0.5 m to 4 m at most) (Figure 3) and its exposure to variations in extreme water levels and waves, the coast of Hudson Bay is an environment that is sensitive to flooding by submersion (L'Hérault, 2017a and 2017b). However, in spite of its sedimentary composition, the coast is not sensitive to erosion (Boisson, 2019). The coast of Hudson Bay is characterized by the presence of breaking waves (Ford et al., 2016). The risk of higher breaking waves could be amplified if the “storm season” coincides with the bay’s ice-free season (Savard et al., 2014), as the ice has a mitigating effect on winds and high waves. Storm surges in Hudson Bay are the result of drops in atmospheric pressure caused by intense low-pressure systems passing through the bay (inverted barometer effect) and storm surges from the Labrador Sea. Hudson Bay has not been well represented in various past global climate models and an understanding of the development and movement of storms on this body of water is has not yet been developed. With respect to waves, the models are currently being processed (Technical Synthesis). Summer ice cover in Hudson Bay fell by 10.5% per decade between 1968 and 2010 (Derksen et al., 2012).

HAZARDS IN THE FUTURE CLIMATE Hudson Bay could experience the most significant increase in total winter precipitation of all the Nunavik coasts.

Rising winter temperatures could reduce the length of the ice season by nearly two months, as well as ice concentration and thickness. Waves could be higher in the fall due to the absence of ice, which increases the fetch and reduces the coast’s protection.

Given the relative sea level changes, positive surges could be more frequent but less high. Negative storm surges could be more frequent and/or lower around 2100. Thus, the risks associated with positive storm surges should decrease in the long term, as the falling relative sea level gradually reduces the potential for damage from the rise in positive surge extreme water levels.

On Hudson Bay, winter negative storm surges could occur more often in early winter and in March.

The eastern coast of Hudson Bay could experience the most significant increase in total winter precipitation in Nunavik, with 25 to 45% more precipitation in the mid-21st century (Logan et al., 2011). Eastern Hudson Bay and Ungava Bay are among the areas where even larger decreases in ice thickness are projected (Ouranos, 2015a). Freeze-up is expected to occur 25 to 30 days later, breakup is projected to occur 22 to 24 days earlier, and winter ice thickness is set to decrease by 30 to 50% in the 2041-2070 period, as compared to the 1961-1990 period (Joly et al., 2011; Senneville and St-Onge Drouin, 2013). This would mean a two-month extension of the open water and ice-free period. If the development of landfast ice coincided with the heavy storm period, the formation and stability of landfast ice would be further reduced by the displacement of ice pushed by strong winds. Because warmer temperatures could last into the winter, the unstable atmospheric conditions in the fall could also extend into January, February and even March towards the end of the 21st century (2041 to 2099). Indeed, in a future warmer climate, the ice-free season would last longer and end later. Summer heat accumulation in the waters of Hudson Bay will increase and continue into early winter, which will delay freeze-up and potentially increase atmospheric instability, prolonging the positive and negative storm surge season and increasing surge frequency and magnitude (from 0.5 to 1 m). In addition, the season during which extreme levels can coincide with strong waves, once limited to the fall, may increasingly extend to the first months of winter. Fall weather conditions are already considered difficult (frequent fog, freezing rain, strong winds, etc.) but could worsen and last several weeks longer in a few decades. However, this hypothesis remains to be verified (Savard, 2016). The strong isostatic adjustment in this area, which mitigates the rise in relative sea level resulting from other factors, should not be overlooked. In fact, the coast will have risen by about 1.4 m from Kuujjuarapik to around Puvirnituq by 2100 (Ouranos 2015a). Taking this increase into account, projections show that the frequency of very high water levels could decrease, while the frequency of very low levels is expected to increase more rapidly than in the past (James el al., 2014). Negative storm surges with a 100-year return period could reduce the minimum water level by 70 cm to 1 m by 2040-2069 and by 80 cm to 1.1 m by 2070- 2099. A decrease in total water levels could be even greater towards the south, from Akulivik to Umiujaq, due to the combined effect of a lower average relative sea level and higher negative storm surge intensity (Ropars, 2014). In central and northern Hudson Bay, there could be a decrease in negative storm surge heights in early winter, and an increase between February and March. In the distant future, these winter negative storm surges could shift to the spring (Savard, 2016). On the coasts of Hudson Bay, extreme positive storm surges will always cause flooding whether or not they coincide with spring tides. That being said, the projected positive storm surges will reach water levels close to 1 m, which is very close to the tidal range for several communities. There could be an increase in the combined effect of waves/water levels and coastal wave strength. The natural hazard maps produced by Michel Allard’s CEN team at the request of the MSP (Carbonneau et al., 2015, Allard et al., 2020) for Hudson Bay communities use a level 2.5 m above the high tide level. This higher level is used in order to increase the degree of safety in the context of positive storm surge value uncertainties. In Hudson Bay, a positive storm surge coinciding with a 3- to 4-m run-up is a major problem, especially given that at its current speed of around 1 cm per year, post-glacial rebound (Table 9) will not reduce this risk for a long time (over 50 years) (Michel Allard, personal communication).

NRCan modeling of post-glacial rebound rates for Hudson Bay communities

Station Post-glacial rebound rate Station Post-glacial rebound rate

Ivujivik 0.79 cm/year Inukjuak 1.18 cm/year

Akulivik 0.86 cm/year Umiujaq 1.15 cm/year Puvirnituq 0.89 cm/year Kuujjuarapik 1.38 cm/year

As indicated in Chapter 1, hydro-climatic hazards only account for part of the risk of significant changes in a given environment. Indeed, this risk also depends on the environment’s intrinsic vulnerability factors, which fall into four main categories associated either with the natural environment (environmental and physical) or with the human environment (economic and social). Recent research carried out by teams from CEN and NRCan (Allard et al., 2020a, b and c and Manson et al., 2019) has led to the mapping of the coastal environment’s hydro-climatic hazards and environmental and physical vulnerability factors. The local maps produced by CEN for each community specify the spatial extent of the natural hazards identified in Nunavik (Allard et al., (2020a, b and c). They complement the results presented here. NRCan has developed a climate change Coastal Sensitivity Index (CSI) (discussed in detail in Chapter 3) that includes the external hazards (changes in sea level and wave height, including effects of changes in sea ice) and physical characteristics (i.e. shoreline composition, slope, presence of ground ice) of the coastal environment that can change over time (Manson et al., 2019). NRCan offers a map at a Canada-wide scale, and it is possible to zoom into the map and obtain the CSI at the Nunavik and community scales (see Appendix). As the map uses the same external hazards while adding physical characteristics, it is a relatively effective complement to the findings obtained for the MTQ. A synthesis map with the characteristics mentioned above was created for each of the communities for which water levels were simulated, i.e. for nine communities and Deception Bay. The map provided below is for the first community in the first coastal region (Section 2.2.2), i.e. Kuujjuarapik (Sub-section 2.2.2.1). The map is labelled Figure 6 “Kuujjuarapik and climate change.” The maps of the other communities are included in the Appendix. A description of these maps is provided in the following paragraph. On each of the synthesis maps presented in the main text (Figure 6) and in the Appendix (Figures 9 to 17), the CSI has been superimposed on the positive and negative storm surge extreme water levels simulated for the recent past and the extreme water level variations for future periods. On these maps, the water levels are indicated in centimetres (cm). The negative and positive storm surge levels from 1980 to 2009 are indicated by circular symbols that contain a wave and water level. In the negative storm surge symbol, the wave is situated below the “zero” water level, while in the positive surge symbol, it is above the zero level. The symbols change colour according to the variation from the zero level, with levels closest to zero cm indicated in blue and the most extreme levels indicated in red. For the future periods, the variations in the extreme levels (i.e. the extreme level for the considered future period minus the extreme level from the recent past) are indicated by an arrow. A positive value indicates an increase, represented by a rising arrow, while a negative value indicates a decrease, represented by a descending arrow. If the variation is small (0 to 5 cm), the arrow is blue. If the variation is moderate (5 to 10 cm), the arrow is green. If the variation is large (10 to 15 cm), the arrow is orange. Finally, if the variation is very large (15 to 20 cm), the arrow is red.

2.2.2 THE COMMUNITIES OF HUDSON BAY

2.2.2.1 Kuujjuarapik

As a result of post-glacial rebound, the relative sea level in the Kuujjuarapik area could fall by 87.1 cm under RCP4.5 and by 72.6 cm under RCP8.5 at the end of the 21st century.

It is likely that Kuujjuarapik will not experience any extreme negative storm surge level variation towards the end of the 21st century, as compared with the recent past. Extreme positive surges, which were calculated without taking post-glacial rebound into account, could be 19 cm higher in 2040-2069 and 15.7 cm higher in 2070-2099. Currently, storms are very present and regular throughout the ice-free period. Calm seas are almost non-existent off the coast of Kuujjuarapik. The three years (2015-2016, 2016-2017 and 2017-2018) of ice observation in Kuujjuarapik do not cover a long enough period to establish a trend with respect to the length of the ice season or ice thickness, freezing or melting.

Figure 6. Kuujjuarapik coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods

In the Kuujjuarapik region, the post-glacial rebound rate is the highest in Nunavik at 1.4 cm/year (Table 9). This is so high that “the community should not experience any rise in sea level, even if the melting West Antarctic Ice Sheet makes an additional contribution to the rising sea level (Ford et al., 2016).” The relative sea level in the Kuujjuarapik area is projected to drop by 87.1 cm under RCP4.5 and 72.6 cm under RCP8.5 (Table 3) by the end of the 21st century, which would necessarily lead to an equivalent decrease in 100-year positive surges as compared to the 1980-2009 period. Table 4 and Figure 6 are based on the calculations of Massé and Gallant (2016) and do not take post-glacial rebound into account. According to Massé and Gallant (2016), this positive surge level will reach 1.63 m towards the end of the 21st century (Table 4). When factoring in post-glacial rebound, this would correspond to a height of 0.29 m (1.63 - 1.34). For the airstrip to be submerged, a 4 m surge would be required (Boisson, 2019).

Negative and positive storm surges reaching nearly 1 m have a large impact on the water level because of the low tidal range (Figure 3) and must taken into consideration. Given the 1.43-m average tidal range and 2-m maximum tidal range (Fisheries and Oceans Canada, 2016), these levels are equivalent to the average storm surge levels. Extreme negative storm surges could increase water levels in both future periods. Projections suggest that there is a positive difference between future negative surges and those from the recent past, indicating future negative surges lower than those measured in the recent past (7 cm lower in 2040-2069 and 3.5 cm lower in 2070-2099, Table 4 and Figure 6). Regular (non-extreme) positive storm surges can easily reach 1 m (Hill et al., 2003 and Neumeier et al., 2017). Extreme positive surges with a 100-year return period could be 19 cm higher in 2040-2069 and 15.7 cm higher in 2070-2099 (Table 4 and Figure 6). These increases will not cause damage to the current roads (roads or airstrip) or houses (Boisson, 2019). During the observation year (2017-2018), Neumeier et al. (2019) found that there were 12 storm events with waves exceeding three metres. There was significant storm presence in August and early September. However, this data does not yet make it possible to confirm interannual storm event variability or seasonal patterns. The last storm measured before the complete freeze-up of Hudson Bay occurred on December 15, 2017, and the first storm event detected after the melt occurred on July 3, 2018. The Manitounuk Islands provide coastal protection by reducing wave strength, but the protection is less effective in the case of strong westerly winds, which generate waves that can come through Manitounuk Strait (Neumeier et al., 2019). Northern winds generate larger waves to the south and off the mouth of the Great Whale River. Waves measuring between 1.5 and 2 m reach and pass through the Manitounuk Islands but are strongly attenuated. In Manitounuk Strait, waves do not exceed 1.5 m in height. The passages between Neilsen and Merry Islands in front of Kuujjuarapik’s marine infrastructure appear to be the most important passages in Manitounuk Strait for offshore waves related to this type of storm. Westerly winds generate larger waves off the Manitounuk Islands and north of the community at the mouth of Manitounuk Strait. The waves entering Manitounuk Strait are higher (>1.5 m) and travel further. The results of the two previous simulations show that waves appear to be rapidly attenuated at the mouth of the Great Whale River and do not exceed 1 m in height in the river. Off Kuujjuarapik, winter wind speed trends are slightly increasing (Ouranos, 2015a). Ice observations in Kuujjuarapik for the three years 2015-2016, 2016-2017 and 2017-2018 (Tables 10 and 11) do not cover a long enough period to establish a trend. Indeed, during the most recent winter, the first ice seems to have appeared later, but full freeze-up occurred more quickly (in just two days) than during the previous two winters. In the spring, melting appears to have started later, but the last ice observed also appeared later than in 2015-2016.

Summary of ice observations obtained by camera for Kuujjuarapik from 2015 to 2018 (Bernier et al., 2017; 2019; Poulin et al., 2018)

North camera First First signs of (Marine First ice sighting Complete freeze-up complete Last ice sighting water Infrastructure) clearing of ice

2015-2016 27 November 2015 23 December 2015 29 May 2016 06 June 2016 16 July 2016

2016-2017 23 November 2016 14 December 2015 26 May 2017 07 June 2017 9 June 2017

15 July 2017-2018 10 December 2017 12 December 2017 13 June 2018 26 June 2018 2018

Duration of the various stages of the ice cover as captured by camera in Kuujjuarapik from 2015 to 2018 (Bernier et al., 2017; 2019; Poulin et al., 2018)

Observations 2015-2016 2016-2017 2017-2018

Freeze-up duration (days) 26 22 2

Number of days of complete cover 158 162 182

Number of days between first sighting of water 8 13 14 and ice outflow Number of days between outflow and the last 40 2 19 sighting of ice

Number of mobile ice episodes 2 1 1 15 (5 and Average duration of mobile ice episodes (days) 1 1 25)

According to the coastal classification carried out by L’Hérault et al. (2017), Kuujjuarapik is made up of sand banks or beaches and the sandy spit is experiencing progradation. Parts of Manitounuk Strait may be at risk for landslides. Indeed, a significant event that occurred in this strait in 2005 caused the formation of a new bay. This type of event could occur again, as the trigger depends on local geological conditions rather than on the climate (Boisson, 2019). It also appears that erosion and flooding are not a risk at this time. Thus, Kuujjuarapik's infrastructures should not be vulnerable to extreme surges. The recent marine infrastructure located in the Manitounuk Strait is protected from offshore waves and the consequences of potential extreme positive storm surges (Boisson, 2019). On the other hand, the wharf located in the Great Whale River is more exposed to the risks associated with the increased intensity and frequency of extreme negative storm surges. Indeed, access could be more difficult as a result of the continuous sanding up of the Great Whale River delta (Boisson, 2019).

2.2.2.2 Umiujaq

At the end of the 21st century, the relative water level could be 92 cm lower than it is today. In the 2040-2069 period, extreme negative storm surge levels could be lower than in the past and remain unchanged from 2070-2099. Positive storm surge levels could rise by nearly 16 cm in the 2040-2069 and 2070-2099 periods. Because of the small tidal range, positive and negative storm surge levels have an impact on the total water level. The return period for extreme storm surges could decrease. The extreme total water level, which previously reached 2.09 m once every 100 years, could be reached once every 40 years in the 2040- 2069 period and once every 35 years in the 2070-2099 period. In the future, the duration of the ice cover period could decrease by 29 days: the formation of landfast ice could be delayed by 20 days and melting could occur 9 days earlier. In addition, the mobile ice period could last several weeks.

With post-glacial rebound reaching 1.15 cm/year (Table 9), the relative water level at the end of the 21st century could be 92 cm lower than today if the higher global sea level is not taken into account. In Umiujaq, around the end of the 21st century, extreme negative storm surge levels with a 100-year return period could be higher by 2040-2069 and roughly unchanged (-0.95 m) by 2070-2099 (Table 4 and Figure 9). The lowest total water level could reach 1.54 m in 2040-2069 and 1.51 m in 2070-2099 (Table 12). The return period for this lowest level could be halved, i.e. once every 50 years in the 2040-2069 period (Table 13). On the other hand, in the 2070-2099 period, extreme negative storm surges with a 100-year return period could have the same return period as they do now. Currently, extreme negative storm surges take place in September, October and November, and medium-sized negative storm surges take place in the winter (Massé and Gallant, 2016).

Minimum negative storm surge and total levels in the future climate in Umiujaq (from Figure 5.15 in Massé and Gallant, 2016) Negative storm surge levels (residual level) in cm Total level (tide + residual level) in cm Return period 2040-2069 2070-2099 2040-2069 2070-2099 2 years -65 -66 -121 -120 5 years -73 -76 -131 -127 10 years -76 -80 -136 -134 20 years -80 -85 -142 -140 25 years -82 -85 -144 -140 50 years -86 -91 -149 -146 100 years -90 -95 -154 -151

Return periods for certain minimum extreme total water levels in current and future climates in Umiujaq (from Figure 5.15 in Massé and Gallant, 2016) Extreme low total water level Return period in current Return period from 2040- Return period from 2070- (cm) climate (years) 2069 (years) 2099 (years) -152 100 50 100 -149 75 45 75 -146 50 35 50 -143 30 25 30

In Umiujaq, strong and persistent westerly winds cause most of the major positive storm surges. The average tidal range is 1.5 m (small range) at Gillies Island (near Umiujaq) (Figure 3). Extreme positive surges in the 2040-2069 and 2070-2099 periods could be 15.6 cm higher than those in the recent past (Table 4). Positive surges with a two-year return period could reach 1 m, while the most extreme positive surges could reach up to 1.7 m in the two future periods (Table 14). The highest 100-year total water level could reach 226 cm by 2070-2099 (Table 14). Factoring in the 92-cm drop towards the end of the 21st century, this 100-year extreme total level would therefore reach 134 cm (226 - 92). This level could thus be lower than the current extreme total water level of 209 cm (Table 15). That being said, these extreme surge levels are still nearly double the height of the average tidal range.

Maximum positive storm surge and total levels in the future climate in Umiujaq (from Figure 5.15 in Massé and Gallant, 2016) Positive storm surge levels (in cm) Total level (tide + positive level) Return period 2040-2069 2070-2099 2040-2069 2070-2099 2 years 110 110 160 160 5 years 125 130 175 175 10 years 140 140 190 190 20 years 150 150 200 200 25 years 150 155 205 205 50 years 162 163 213 215 100 years 172 173 224 226

While it is possible that water levels will fall, it is important to consider that the return period could also decrease. For example, the total water level of 2.09 m currently has a return period of 100 years. The same level could have a return period of 40 years from 2040 to 2069 and 35 years from 2070 to 2099 (Table 15). Even with extreme water levels that are lower than today, the higher probability of occurrence could also cause significant damage through repetitive stress (Boisson, 2019).

Return periods for certain maximum extreme total water levels in current and future climates in Umiujaq (from Figure 5.15 in Massé and Gallant, 2016) Extreme low total water level Return period in current Return period from 2040- Return period from 2070- (cm) climate (years) 2069 (years) 2099 (years) 209 100 40 35 205 75 30 25 200 50 20 20 190 30 10 10

Extreme positive storm surges could occur in September, October and November (Massé and Gallant, 2016). Weaker storm surges could appear around June and February. Also, storm surges could increase overall between August and December, with a significant increase in September (Massé and Villeneuve, 2013). In the spring, the mobile ice period could extend over several weeks at Umiujaq (Clerc et al., 2012 and Bernier et al., 2017; 2018; 2019) due to the effect of strong currents that occur throughout winter (Clerc et al., 2012). The Nastapoka Islands (Gillies, Curran) protect the Umiujaq coastline from bad weather from the open sea. They also allow the formation of stable landfast ice in winter and limit the inflow of exogenous ice from Hudson Bay during the spring melt, when the coastline is most vulnerable to drift ice (Clerc et al., 2012). At the same time, however, the ice from Lake Tasiujaq (also known as Guillaume Delisle Lake) exits into Nastapoka Strait for several weeks, then drifts from the south to the north, passing close to Umiujaq. These ice floes are a significant infrastructure and navigation risk (Clerc et al., 2012). Over the 2009-2017 period, freeze-up occurred quickly (1 to 3 weeks) in Umiujaq (Clerc et al., 2012, Bernier et al, 2017; 2018; 2019) (Table 16). In the future, the formation of landfast ice could be delayed by 20 days

and melting could occur 9 days earlier, resulting in a 29-day decrease in the duration of ice cover (Sennevile and St-Onge Drouin, 2013). In Umiujaq, ice conditions do not pose a problem for the marine infrastructure, whether during the spring melt or freeze-up (Clerc et al., 2012). The two main reasons are probably the small tidal range and the protection provided by the Nastapoka Islands in front of the community. The only risks associated with freeze-up are fractures in the pack ice as a result of high water levels or strong winds (Clerc et al., 2012). Ice observations in Umiujaq for the three years 2015-2016, 2016-2017 and 2017-2018 do not cover a long enough period to establish a trend (Tables 16). Indeed, the shortest freeze-up (8 days) was observed in the first year, while the longest (19 days) was observed the next year.

Duration of the various stages of the ice cover as captured by camera in Umiujaq from 2016 to 2018 (Bernier et al., 2017; 2019; Poulin et al., 2018)

Observations 2015-2016 2016-2017 2017-2018

Freeze-up duration (days) 8 19 12

Number of days of complete cover 154 159 170

Number of days between first sighting of water 4 5 19 and ice outflow

The Umiujaq coastline is formed by a sandy beach bordered by a 2 m low bluff, which is being eroded and pushed back by the intensity and frequency of storm waves. Even in the event of a 1.7-m extreme positive storm surge (Table 4), the road network and houses would be unaffected, with the exception of a few wooden cabins (some of which have already been removed). On the other hand, storage sites near the marine infrastructure are at risk of flooding (Boisson, 2019). Thus, Umiujaq's marine infrastructure could become less and less vulnerable to high water levels due to the effect of post-glacial rebound, although accessibility to the basins could become a problem (Ropars, 2014). In the short term, however, it is very important to take into account the increased frequency of extreme water levels despite the decline in total and extreme water levels expected at the end of the 21st century. The hazard maps established by Carbonneau et al. (2015) indicate that Umiujaq is also sensitive to thawing permafrost and to mass movements that cause landslides.

2.2.2.3 Inukjuak

By the end of the 21st century and as a result of post-glacial rebound, the sea level could fall by 71.2 cm under RCP4.5 and 58.8 cm under RCP8.5. The extreme negative storm surge water levels could reach -0.81 m by 2040-2069 and -0.91 m by 2070-2099. The extreme positive storm surge water levels could reach nearly 1.3 m in the 2040-2069 and 2070- 2099 periods, which corresponds to increases in surge levels of more than 20 cm, as compared to 1989-2009.

Storm surges have a very strong effect on the total water level because of the small tidal range (~22 cm) (Figure 3). In the Inukjuak region, with the post-glacial rebound rate reaching 1.2 cm/year (Table 9), the relative water level could fall by 71.2 cm by the end of the 21st century under RCP4 .5 and by 58.8 cm under RCP8.5.

100-year extreme negative storm surges could decrease by 2.7 cm, reaching -0.81 m by 2040-2069, and by 11.8 cm, reaching -0.91 m by 2070-2099 (Table 4 and Figure 10). Taking into account the effect of post-glacial rebound, this could lead to extreme negative surge levels of -1.62 m (0.91 + 0.71) under RCP4.5 and -1.50 m (0.91 + 0.59) under RCP 8.5 by the end of the 21st century. Extreme positive storm surge levels could reach 1.30 m by 2040-2069 and 1.33 m by 2070-2099 (Table 4), which corresponds to an increase of more than 20 cm, as compared to 1989-2009 (Table 4). Taking into account the effect of post-glacial rebound, this could lead to extreme positive storm surge levels of 0.59 m (1.30– 0.71) under RCP4.5 and 0.74 m (1.33 - 0.59) under RCP8.5. If they occur more frequently, these water levels could represent a flood risk for several houses located near the coast, for the area around the marine infrastructure, and for some road sections where the altitude of the built environment is less than 30 m and the topography is flat with some rocky hills and valleys (Boisson, 2019). This community is built on sandy terraces along the eastern shore of the Innuksuak River (L’Hérault et al., 2017).

2.2.2.4 Puvirnituq

As a result of post-glacial rebound, the sea level could fall by 71.2 cm under RCP4.5 and 58.8 cm under RCP8.5 by the end of the 21st century. Extreme negative storm surge water levels could reach -0.97 m by 2040-2069 and -1.08 m by 2070- 2099. Extreme positive storm surge water levels could reach nearly 1.4 m in the 2040-2069 and 2070-2099 periods, which corresponds to a surge level increase of more than 13 cm, as compared to 1989-2009.

As a result of the micro-tidal coast, storm surges have a very strong effect on the total water level (Figure 3). The rate of post-glacial rebound, which is currently 0.89 cm/year (Table 9), could lead to a 71.2-cm decrease in the relative water level at the end of the 21st century, without factoring in the global sea level rise. The cumulative influences of the average sea level rise and post-glacial rebound should cause a small drop (less than 10 cm with a 30- or 50-year return period) in the total relative water levels in Puvirnituq (Ropars, 2014). Extreme negative storm surge water levels, which reached -0.92 m in the recent past and could reach -0.97 m by 2040-2069 and -1.08 m by 2070-2099 (Table 4 and Figure 11), would be amplified by the fall in relative sea level linked to post-glacial rebound. At the end of the 21st century, extreme negative surge water levels could then reach -1.79 m (1.08 + 0.71). Extreme positive storm surge water levels, which reached 1.30 m in the recent past, could reach 1.40 m between 2040 and 2069 and 1.43 m between 2070 and 2099 (Table 4), which corresponds to increases of almost 11 cm and 14 cm respectively, as compared to 1989-2009 (Table 4). Located on the north shore of the Puvirnituq River, less than 10 km from Hudson Bay, the community of Puvirnituq has a coastline characterized by small coves, beaches and rocky islets. Most of the buildings are located at an altitude of less than 10 m (Boisson, 2019). Mass movements, erosion and permafrost thawing are the most important hazards for this community (Carbonneau et al., 2015). The expected extreme positive storm surges do not generally pose a risk to existing houses (Boisson, 2019). However, flooding would affect some houses, road sections, and marine infrastructure surroundings located about 10 m from the area swept by positive storm surges (Boisson, 2019).

2.2.2.5 Akulivik

As a result of post-glacial rebound, the sea level could fall by 68.8 cm by the end of the 21st century. Extreme negative storm surge water levels could reach -0.90 m by 2040-2069 and 2070-2099. Extreme positive storm surge water levels could reach nearly 1.30 m in the 2040-2069 and 2070-2099 periods, which corresponds to a surge level increase of more than 9 cm, as compared to 1989-2009.

As a result of the micro-tidal coast (the tidal range is less than 2 m), storm surges have a very strong effect on the total water level (Figure 3). The rate of post-glacial rebound, which is currently 0.86 cm/year (Table 9), could lead to a 68.8-cm decrease in the relative water level at the end of the 21st century, without factoring in the global sea level rise. The cumulative influences of the average sea level rise and post-glacial rebound should result in relatively stable total relative water levels, with variations below 10 cm expected in Akulivik (Ropars, 2014). Extreme negative storm surge water levels, which reached -0.80 m in the recent past and could reach -0.86 m by 2040-2069 and -0.94 m by 2070-2099 (Table 4 and Figure 12), would be amplified by the decrease in relative sea level linked to post-glacial rebound. At the end of the 21st century, extreme negative storm surge water levels could then reach -1.63 m (0.94 + 0.69). Extreme positive storm surge water levels, which reached 1.18 m in the recent past, could reach 1.27 m between 2040 and 2069 and 1.30 m between 2070 and 2099 (Table 4), which corresponds to increases of almost 9 cm and 14 cm respectively, as compared to 1989-2009 (Table 4). Akulivik is the most vulnerable to the coastal submersion hazard because it is located at a very low altitude and has a micro-tidal coast (Figure 3) (Boisson, 2019). Extreme hydrological events are the most significant hazard (Carbonneau et al., 2015). A 1.18-m surge level (Table 4), which could occur more frequently than in the past, would cause extensive damage to infrastructure (Boisson, 2019). This would lead to the flooding of around ten houses and the roads leading to the port and airport (Boisson, 2019).

2.2.2.6 Ivujivik

By the end of the 21st century, the sea level could be roughly the same as in the recent past. Extreme negative storm surge water levels could reach -0.75 m by 2040-2069 and 2070-2099. Instead of a 100-year return period, the lowest extreme total water level from the recent past (-2.14 m) could have a return period of 30 years in 2040-2069 and 55 years in 2070-2099. Extreme positive storm surge water levels could reach nearly 1.30 m in the 2040-2069 and 2070-2099 periods, which corresponds to a surge level increase of more than 9 cm, as compared to 1989-2009. Instead of a 100-year return period, the highest extreme total water level from the recent past (2.35 m) could have a return period of 35 years in 2040-2069. Ice observations in Ivujivik for the three years 2015-2016, 2016-2017 and 2017-2018 do not cover a long enough period to establish a trend.

The rate of post-glacial rebound, which is currently 0.79 cm/year (Table 9), could lead to a 63.2-cm decrease in the relative water level at the end of the 21st century, without factoring in the global sea level rise. The tidal range is medium (between 2 and 4 m) (Figure 3), so storm surges have little effect on the total sea level. The cumulative effect of the average sea level rise and post-glacial rebound should result in total relative levels with a high level of stability. This is in line with the predicted drop of a few centimetres (Ropars, 2014).

Extreme negative storm surge levels, which reached -66.5 cm in the recent past, could be even lower, reaching -74 cm by 2040-2069 and -80 cm by 2070-2099 (Table 4 and Figure 13), or a decrease of 7 and 13 cm respectively (Table 4). The lowest 100-year extreme total levels could decrease up to -2.28 m by 2040- 2069 and -2.18 m by 2070-2099 (Table 17), without factoring in post-glacial rebound. The lowest extreme total water level in the recent past was -2.14 m (Table 18). This same water level would have a return period of 30 years by 2040-2069 and 55 years at the end of the 21st century (Table 18).

Minimum negative storm surge and total levels in the future climate in Ivujivik (from Figure 5.14 in Massé and Gallant, 2016) Negative storm surge levels (residual level) in cm Total level (tide + residual level) in cm Return period 2040-2069 2070-2099 2040-2069 2070-2099 2 years -44 -46 -180 -177 5 years -52 -56 -194 -189 10 years -57 -62 -202 -195 20 years -63 -66 -210 -202 25 years -64 -69 -213 -204 50 years -69 -74 -220 -211 100 years -74 -80 -228 -218

Return periods for certain minimum extreme total water levels in current and future climates in Ivujivik (from Figure 5.14 in Massé and Gallant, 2016) Extreme low total water level Return period in current Return period from 2040- Return period from 2070- (cm) climate (years) 2069 (years) 2099 (years) -214 100 30 55 -212 75 25 50 -209 50 17 35 -203 30 12 25

In the recent past, medium-sized positive storm surges (-0.5 m) mostly occurred in the winter. In the future climate, on the other hand, large and extreme negative storm surges (-0.6 to -1 m) could be increasingly numerous in the fall, in October, November, March and April (Massé and Gallant, 2016). That being said, negative storm surges exceeding these levels could be less frequent in late fall and winter (Savard, 2016). Extreme positive storm surge levels, which reached 1.01 m in the recent past, could reach 1.11 m by 2040- 2069 and 1.07 m by 2070-2099 (Table 4), an increase of 10 cm and 6 cm respectively. The highest 100-year extreme total water level could reach 2.48 m by 2040-2069 and 2.31 m by 2070-2099 (Table 19), without factoring in post-glacial rebound. In the recent past, the highest extreme total water level was 2.35 m. This same water level would have a return period of 35 years by 2040-2069 (Table 19). Extreme positive storm surges could occur in September, October and November (Massé and Gallant, 2016).

Maximum positive storm surge and total levels in the future climate in Ivujivik (from Figure 5.8 in Massé and Gallant, 2016) Positive storm surge levels (residual level) in cm Total level (tide + residual level) in cm Return period 2040-2069 2070-2099 2040-2069 2040-2069 2 years 64 65 195 191 5 years 76 76 210 202 10 years 85 85 220 209 20 years 93 92 230 215 25 years 95 94 231 216 50 years 103 100 239 224 100 years 111 107 248 231

Return periods for certain maximum extreme total water levels in current and future climates in Ivujivik (from Figure 5.8 in Massé and Gallant, 2016) Extreme high total water Return period in current Return period from 2040- Return period from 2070- level (cm) climate (years) 2069 (years) 2099 (years) 235 100 35 Over 100 233 75 30 Over 100 230 50 25 80 224 30 15 45

Freeze-up of Hudson Bay and Hudson Strait can occur within days, depending on the temperatures, winds and tides pushing ice into the bay (Bernier et al., 2017; 2018; 2019). Table 21 presents observations taken from Bernier et al. (2017-2019) and Poulin et al. (2018). Ice observations in Ivujivik for the three years 2015- 2016, 2016-2017 and 2017-2018 do not cover a long enough period to establish a trend.

Ice processes in Ivujivik (Bernier et al., 2017; 2019; Poulin et al, 2018)

Observations 2015-2016 2016-2017 2017-2018

Freeze-up duration (days) 18 6 9 Number of days of complete cover 196 162 181 Number of days between first sighting of water 6 14 28 and ice outflow

Located at the bottom of a small valley, the community of Ivujivik is built almost entirely on rocky and slightly undulating terrain (L’Hérault et al., 2017). An extreme storm surge of 1 m (Table 4) could cause some minor damage to the built environment. The area around the marine infrastructure and two uninhabited cabins would be flooded (Boisson, 2019). Mass movements, erosion and permafrost thawing are also very significant hazards (Carbonneau et al., 2015).

2.3 HUDSON STRAIT

2.3.1 GENERAL PROFILE

GEOMORPHOLOGY OF THE COAST The coast is rocky. Erosion occurs more slowly than on Hudson Bay. The southern coast of Hudson Strait is located between Cape Wolstenholme and Cape Hopes Advance. Photographic surveys taken by Boisson (2019) have shown that this rocky coast (Ford et al., 2016) consists mainly of high promontories and steep fjords (submerged glacial valleys) with scree slopes.

HAZARDS IN THE CURRENT CLIMATE The tidal range is large, so storm surges have very little effect on extreme variations in the total water level. The extent of ice cover has decreased in the past. Mobile ice is present for longer periods and could cause damage to infrastructure if it is accompanied by strong winds blowing towards the coast. In Hudson Strait, prevailing winds are westerly, tidal currents are strong and the tidal range is large between Salluit (3.3 m) and Quaqtaq (8.1 m) (Figure 5). Winds are often caused by depressions in the Labrador Sea, so the probability of strong waves occurring with very high water levels is likely lower than in Hudson Bay (Savard 2016). Storm surges have very little effect on extreme variations in the total water level, as the tidal range exceeds the positive and negative storm surge levels and the total water level is close to the level of the tide (Massé and Villeneuve, 2013). The extent of ice cover in Hudson Strait decreased by 16% over the 1968-2010 period (Derksen et al., 2012). In winter, the ice cover is not continuous in Hudson Strait. For example, off Quaqtaq, icebergs and floes (patches of ice resulting from the breakup of pack ice) are observed throughout the year due to strong winds blowing from west to east, along the length of the strait (Clerc et al., 2012). Arctic ice shedding also leads to ice flowing in this direction, posing a risk to navigation and any infrastructure located in poorly protected areas (Boisson, 2019).

HAZARDS IN THE FUTURE CLIMATE Much work remains to be done in order to understand the impacts of climate change on Hudson Strait. Very little is known about the oscillations and low-pressure systems that affect the region. Climate change will be of relatively little significance in relation to the tidal range and is unlikely to significantly affect total water levels. On the shores of Hudson Strait, the risks associated with positive storm surges will be limited and those associated with negative storm surges will be very limited, especially since the rate of post- glacial rebound is low. In fact, the coast will rise by 60 to 80 cm by 2100. All the sites on Hudson Strait could experience an average increase in 100-year positive storm surges of around 10 cm by 2040-2069 and of only 4 cm by 2070-2099, as compared to 1980-2009 (Table 3). These surges could reach nearly 80 cm by 2040-2069 and nearly 75 cm by 2070-2099, while they reached nearly 70 cm in the recent past. Given the very large tidal range, these extreme water levels would only be damaging if they occur at the same time as spring or neap tides. Extreme low levels could remain unchanged in Hudson Strait (Table 3). The risks associated with negative storm surges are therefore very low in Hudson Strait, especially since the rate of post-glacial rebound is lower than in Hudson Bay (Boisson, 2019). Indeed, the coast will have risen by 60 to 80 cm on the shores of Hudson Strait by 2100 (Table 22). 36

Rates of post-glacial rebound in Hudson Strait communities according to NRCan model (James et al., 2014)

Station Post-glacial rebound rate

Salluit 0.75 cm/year Kangiqsujuaq 0.67 cm/year Quaqtaq 0.63 cm/year

2.3.2 THE COMMUNITIES OF HUDSON STRAIT

2.3.2.1 Salluit

By the end of the 21st century, the sea level could fall by as much as 60 cm due to post-glacial rebound. The coastal risks (erosion and submersion) linked to storm surges are and could remain moderate to high. The rate of post-glacial rebound, which is currently 0.75 cm/year (Table 22), could lead to a drop in the relative water level of 60 cm at the end of the 21st century, without factoring in the global sea level rise. The coast is macro-tidal (tidal range between 4 and 6 m) (Figure 3), so storm surges have a small effect on the total water level (Massé and Villeneuve, 2013). Water level and ice condition measurement instruments have not been set up in Salluit (Figure 1). However, given its geomorphological configuration, increases in the intensity and frequency of extreme positive storm surges could cause damage to infrastructure in Salluit in the near future (Boisson, 2019). Located between Ivujivik and Deception Bay, Salluit could experience extreme positive storm surges of nearly 1 m (Table 4), which would result in the flooding of the first row of houses along the coast and the road leading to the marine infrastructure (Boisson, 2019). The areas around the mouth of Kuuguluk Creek would also be affected. The rockfill put in place to protect the houses closest to the coast demonstrate the community's vulnerability to this coastal hazard and their efforts to protect themselves from it (Boisson, 2019). Thus, as a result of its location in the fjord, Salluit is protected from fall erosion linked to ice pushes from the sea (Boisson, 2019). However, the sandy and low-lying soil is sensitive to erosion (Carbonneau et al., 2015). More frequent high water levels could increase the risk of summer erosion. However, the current significant degree of sensitivity could remain the same or decrease in the distant future (2070-2099 period) because of the cumulative effects of post-glacial rebound and the global sea level rise.

2.3.2.2 Deception Bay

Extreme negative storm surge water levels could be slightly lower (-3 cm) by 2040-2069 and moderately lower (-6 cm) by 2070-2099, as compared to the recent past. Extreme positive storm surge water levels could be moderately higher (7 cm) by 2040-2069 and slightly higher (3 cm) by 2070-2099, as compared to the recent past. The time series of ice condition measurements does not cover a long enough period to establish a trend. As a result of the macro-tidal coast (the tidal range is between 4 and 6 m) (Figure 3), storm surges have a small effect on the total water level (Massé and Villeneuve, 2013).

Extreme negative storm surge water levels, which reached -0.65 m in the recent past, could reach -0.68 m by 2040-2069 and -0.72 m by 2070-2099 (Table 4 and Figure 14), which corresponds to decreases of 3 cm and 6 cm respectively, as compared to 1989-2009 (Table 4). Extreme positive storm surge water levels, which reached 0.75 m in the recent past, could reach 0.82 m between 2040 and 2069 and 0.79 m between 2070 and 2099 (Table 4), which corresponds to increases of almost 7 cm and 3 cm respectively, as compared to 1989-2009 (Table 4). Deception Bay is built on fragile marine clay. The hazards most likely to cause damage are mass movements and erosion (Carbonneau et al., 2015). Indeed, Deception Bay has already experienced a landslide and the risks associated with this hazard could be higher for the marine infrastructure (Boisson, 2019). A camera-based landfast ice observation system was installed in 2015. The collected data was studied for 2015 to 2018 (Table 23). The time series of measurements does not yet cover a long enough period to establish a trend.

Ice processes in Deception Bay (Bernier et al., 2017; 2019; Poulin et al, 2018)

Number of days Number of days Number of days Ice cover stages in Deception Bay 2015-2016 2016-2017 2017-2018 Freeze-up duration 39 16 35 Permanent ice cover duration 208 185 197 Period between first sighting of water and ice outflow 2 5 13 during spring melt

2.3.2.3 Kangiqsujuaq

By the end of the 21st century, the sea level could fall by 53.6 cm due to post-glacial rebound. The rise in the average sea level would therefore only be somewhat offset by post-glacial rebound, and an increase in total water levels (around 30 cm for 30- or 50-year return periods) could take place by 2040-2074.

The rate of post-glacial rebound, which is currently 0.67 cm/year (Table 22), could lead to a 53.6-cm decrease in the relative water level at the end of the 21st century, without factoring in the global sea level rise. The coast is hyper-tidal (the tidal range is between 8 and 10 m) (Figure 3), so storm surges have a small effect on the total water level (Massé and Villeneuve, 2013). The effect of climate change on positive storm surges would have relatively low significance compared to the tidal range and will probably not have a large effect on total levels, which should not significantly vary (Massé and Villeneuve, 2013; and Ropars, 2014). On the other hand, since the average sea level rise is only somewhat compensated by post-glacial rebound in Kangiqsujuaq, an increase in total water levels (around 30 cm for 30- or 50-year return periods) could realistically occur in the future climate (2040-2074) (Ropars, 2014). Extreme negative storm surge levels, which reached -0.58 m in the recent past, could reach -0.59 m by 2040- 2069 and -0.60 m by 2070-2099 (Table 4 and Figure 15), which corresponds to insignificant decreases of 1 cm and 2 cm respectively, as compared to 1989-2009 (Table 4). Extreme positive storm surge levels, which reached 0.68 m in the recent past, could reach 0.78 m by 2040- 2069 and 0.72 m by 2070-2099 (Table 4), which corresponds to increases of almost 10 cm and 3 cm respectively, as compared to 1989-2009 (Table 4).

The community of Kangiqsujuaq is located in a valley surrounded by high rock faces covered with unconsolidated deposits (Boisson, 2019), on Wakeham Bay in a fjord about 10 km from Hudson Strait. This location in a relatively protected bay should limit exposure to risks associated with ice from the open sea, although a few houses located between the two small rivers that run through the community could be affected by a 1-m extreme surge. (Boisson, 2019). The road leading to the marine infrastructure would not experience flooding (Carbonneau et al., 2015). Thus, the hazards most likely to cause damage in Kangiqsujuaq are mass movements and permafrost thawing (Carbonneau et al., 2015).

2.3.2.4 Quaqtaq

The rise in the average sea level will cause the total water level at Quaqtaq to increase by 30 to 40 cm. Despite their extreme nature (100-year return period), storm surges have little effect on the total water level because of the very large tidal range and low storm surge projections (10 cm for positive storm surges and -4 mm for negative storm surges), as compared to 1980-2009. This situation could continue into the future. The contribution of storm surges to the total water level is not expected to change in the future. 100-year total extreme water levels are expected to be more frequent, occurring every 50 years by 2040-2069. Extreme positive storm surges could take place in September, October and November and last into December and January, and extreme negative storm surges could especially take place in March and April. The riskiest season is currently spring due to the presence of mobile ice. However, the likelihood of more frequent extreme storm surges in the fall could also increase mobile ice risks during this season.

In Quaqtaq, the rate of post-glacial rebound is very low, around 0.63 cm/year (Table 22). It could cause the land to rise by nearly 50 cm by the end of the 21st century. Post-glacial rebound will only somewhat compensate for the foreseeable increase in average sea level. In the future, the relative water level is expected to increase by 30 to 40 cm for events with return periods of 30 to 50 years (Ropars, 2014). The coast of Quaqtaq is hyper-tidal (the tidal range is between 8 and 10 m) (Figure 3). Thus, total water levels are mainly the result of the tide, and the impact of storm surges (even extreme 100-year surges) is very small. Between 2015 and 2017, the tidal range reached more than 9 m during the spring tide and less than 4 m during the neap tide (Neumeier et al., 2019). The negative storm surge level was -54.3 cm for the 1980-2009 period. It could reach -55 cm in the 2040- 2069 and 2070-2099 periods (4 mm less than the level from the recent past) (Table 4 and Figure 16). Extreme low total levels could reach -5.6 m by 2040-2069 and -5.5 m by 2070-2099 (Table 24), while they reached - 5.5 m in the recent past (Table 25). The -5.5-m extreme low total water level, which currently has a 100-year return period, could be expected once every 45 years in the short term (Table 25). In Quaqtaq, during the 1980-2009 period, the negative storm surge season occurred in the winter rather than in the fall (Massé and Gallant, 2016)

Minimum negative storm surge and total levels in the future climate in Quaqtaq (from Figure 5.13 in Massé and Gallant, 2016) Return period Negative storm surge levels (residual level) in cm Total level (tide + residual level) in cm

2040-2069 2070-2099 2040-2069 2040-2069 2 years -36 -36 -515 -520 5 years -41 -41 -526 -515 10 years -45 -45 -535 -525 20 years -47 -47 -542 -535 25 years -48 -48 -545 -535 50 years -52 -52 -552 -544 100 years -55 -55 -559 -553

Return periods for certain minimum extreme total water levels in current and future climates in Quaqtaq (from Figure 5.13 in Massé and Gallant, 2016) Extreme low total water level Return period in current Return period from 2040- Return period from 2070- (cm) climate (years) 2069 (years) 2099 (years) -552 100 45 100 -549 75 40 75 -545 50 25 50 -540 30 17 30

Positive storm surge levels reached nearly 70 cm in the 1980-2009 period. They could reach 81 cm by 2040- 2069 (10 cm higher than the level from the recent past) and 73 cm by 2070-2099 (4 cm higher than the level from the recent past) (Table 4). Extreme high total levels could reach 5.4 m by 2040-2069 and 5.3 m by 2070- 2099 (Table 26), while they are currently at 5.3 m (Table 27). The 5.3-m extreme high total water level, which currently has a 100-year return period, could occur once every 50 years in the 2040-2069 period and once every 100 years in the 2070-2099 period (Table 27). Without factoring in post-glacial rebound, the rise in total levels as a result of storm surges is not expected to exceed the tidal range or the positive and negative storm surge levels from the recent past. For the surges to cause damage on the coast or have a greater effect than the tide, they must exceed 4 m during the neap tide period. This is unlikely according to the obtained simulations and the observations of Neumeier et al., 2019. Positive storm surges did not exceed 0.45 m between 2015 and 2017. Extreme surges occur mainly during the months of December and January.

Maximum positive storm surge and total levels in the future climate in Quaqtaq (from Figure 5.7 in Massé and Gallant, 2016) Positive storm surge levels (residual level) in cm Total level (tide + residual level) in cm Return period 2040-2069 2070-2099 2040-2069 2040-2069 2 years 46 47 492 480 5 years 56 55 505 494 10 years 64 60 512 505 20 years 67 64 520 512 25 years 70 65 522 515

50 years 75 69 530 523 100 years 81 73 537 531

Return periods for certain maximum extreme total water levels in current and future climates in Quaqtaq (from Figure 5.7 in Massé and Gallant, 2016) Extreme high total water Return period in current Return period from 2040- Return period from 2070- level (cm) climate (years) 2069 (years) 2099 (years) 530 100 50 100 529 75 45 75 524 50 25 50 519 30 17 30

Storms conducive to high wave formation occur mainly in the fall, from August to November, and with great regularity (Neumeier et al., 2019). However, some storms still occur in mid-December when ice is forming. These observations corroborate the recent climate simulations (1980-2009) by Massé and Gallant (2016), which show that the positive storm surge season extends from autumn into most of winter, and that negative storm surges occur more frequently in the winter than in the fall. In the future climate, the positive and negative storm surge season could be longer, lasting from August to January. Positive storm surges could even occasionally continue until February (Massé and Gallant, 2016). Extreme positive storm surges could take place in September, October and November and continue into December and January, and extreme negative storm surges could occur mostly in March and April (Massé and Gallant, 2016). For the most part, the ten strongest storms (out of 11 storms during the observation period from August 2017 to August 2018) originated from the northwest sector. They passed between Hearn Island and the mainland and were accompanied by significant storm surge heights exceeding 4.5 m. Only one storm came from the northeast (Neumeier et al., 2019). In the winter, the coastal environment of the Quaqtaq sector is protected by the presence of fast ice. For example, an ice foot accumulates on the breakwater and provides some protection, even while the pack ice gradually disappears during the spring breakup process. Towards the end of the melting period, when the local ice barrier is no longer present, these moving ice rafts can get closer to buildings and infrastructure (Clerc et al., 2012). They can then collide and damage coastal facilities, becoming a source of risk Over the 2009-2017 period, freeze-up generally took from 1 to 5 weeks, while melting took place over 3 to 6 weeks (Clerc et al., 2012; Bernier et al, 2017; 2018; 2019) (Table 28). However, ice observations at Quaqtaq over the three years 2015-2016, 2016-2017 and 2017-2018 (Tables 28) do not cover a long enough period to establish a trend. That being said, the simulations of (Sennevile and St-Onge Drouin, 2013) have shown that melting could occur 17 days earlier and freeze-up could be delayed by 22 days.

Ice processes in Quaqtaq (Bernier et al., 2017; 2019; Poulin et al, 2018)

Observations (days) 2015-2016 2016-2017 2017-2018

Freeze-up duration 17 11 21

Permanent ice cover duration 196 166 181

Period between first sighting of water and ice outflow 22 35 38

Generally speaking, in the Quaqtaq region, the coast could be exposed to mobile ice and high water levels characteristic of spring and fall for much longer periods (Sennevile and St-Onge Drouin, 2013). The risks associated with mobile ice are potentially higher in the spring, but extreme positive storm surges could be more frequent in the fall. The installation of elevated infrastructure outside the area (Boisson, 2019) that would be swept by a potential 1-m extreme positive storm surge (higher than the projections in Table 4) would reduce the risk associated with positive storm surges. In the summer, the coasts are more sensitive to mass movements and erosion (Carbonneau et al., 2015). This is especially true given that calm periods with no strong winds are non-existent during the ice-free period (Neumeier et al., 2019).

2.4 UNGAVA BAY

2.4.1 GENERAL PROFILE

Of the three maritime regions of Nunavik, Ungava Bay has received the least attention to date (Figure 1 – Chapter 1). Significant efforts will be needed in several research areas in order to understand the hazards and ocean dynamics in eastern Nunavik (Chapter 3).

GEOMORPHOLOGY OF THE COAST The coastline of Ungava Bay is relatively protected from erosion as a result of the many islands that dot the large estuaries, reducing the fetch. In some places, the land juts into the sea. The coastline of Ungava Bay is characterized by fjards (large open water space between groups of islands), low rocky coasts, bays and large estuaries (i.e. Arnaud, Leaf, False, Great Whale, George). These estuaries feature vast mudflats and sea marshes that are full of boulders visible at low tide (Boisson, 2019). Only Kuujjuaq is located inland. Tasiujaq and Aupaluk are the lowest elevation communities. Despite a fetch that is larger in Ungava Bay than on the eastern side of Hudson Bay, the many islands and large foreshore create natural barriers that attenuate wave energy and limit erosion on the coast (Boisson, 2019). Boisson (2019) has even observed progradation of beaches and the movement of marshes towards the sea. Only certain rocky areas experience erosion as a result of freeze-thaw weathering and storm waves (Fournier and Allard, 1992), such as the George River estuary. This phenomenon has been occurring since the glaciers retreated and will continue in the future. Surveys also show that current beaches are not experiencing erosion.

HAZARDS IN THE CURRENT CLIMATE Measuring devices have been set up in Ungava Bay, in Aupaluk and Kuujjuaq. However, the lack of measuring devices in other Ungava Bay communities limits our understanding of the impacts of climate change in the maritime region. The extent of ice cover has decreased in the past and mobile ice is present for longer periods. The ice can cause damage to infrastructure if its presence is accompanied by strong winds blowing towards the coast.

The tidal range in Ungava Bay is very large, ranging from 4 m to over 6 m (Figure 3) and reaching a maximum of 16.7 m in the Leaf River estuary (Fisheries and Oceans Canada, 2019). Also, storm surges have very little effect on the extreme variations in total water level. The coastlines of the isolated communities of Tasiujaq and Aupaluk have the largest tidal ranges known in Quebec. The tide generates particularly strong ebb and flood currents, which greatly increase the risks for navigation and continually break the ice cover that forms at the end of October (Boisson, 2019). On the other hand, in certain areas with raised beaches and raised boulder barricade backshore beaches, Boisson (2019) has observed regressing coastlines. These barricades also have a protective role, reducing the intensity of waves reaching the coast (Boisson, 2019). Aupaluk was equipped with cameras to observe ice conditions in 2015. The observations do not cover a long enough period to establish a clear trend.

HAZARDS IN THE FUTURE CLIMATE Since the communities on Ungava Bay have little or no measuring devices for long-term observations (Figure 1 Chapter 1), it is difficult to establish past trends with respect to the various hazards identified by the MTQ, and even more challenging to extrapolate trends into the future. Ungava Bay was included in the spatial extent or computational domain of the hydrodynamic model used in the work of Massé and Villeneuve, 2013, Savard et al. 2014 and Massé and Gallant, 2016, although this region is generally data deficient. The simulations were calculated only for Kuujjuaq. Among the five communities around Ungava Bay and upstream from the Koksoak River, only Kuujjuaq was equipped with a tide gauge at the mouth of the estuary, which unfortunately was unable to collect data due to the breakdown of the device (see Technical Summary). In order to obtain a more comprehensive picture of conditions in Nunavik, it would be helpful to improve hydrological and climatic knowledge of Ungava Bay, which represents a promising area for research and development (Section 3.4).

2.4.2 THE COMMUNITIES OF UNGAVA BAY

2.4.2.1 Kangirsuk

The community of Kangirsuk is separated into two neighbourhoods located on either side of a bay on the north shore of the Arnaud River. Built on coarse deposits, it stands on a solid rock base. The geographical location of this community, on a rocky hill with steep slopes, gives it a certain degree safety with respect to submersion risks. Thus, in the event of a 1-m extreme positive storm surge, the community’s road network, houses and other infrastructure would not be exposed. In Kangirsuk, there is thus no risk associated with positive storm surges, and this situation will continue in a climate change context. The strong ebb and flood currents in the Arnaud River estuary can currently make navigation more hazardous (Boisson, 2019). The hazards mapped at Kangirsuk are climatic and hydrological in nature (Carbonneau et al., 2015).

2.4.2.2 Aupaluk

Located south of Hopes Advance Bay, Aupaluk is built on the sandy deposits of raised beaches not exceeding 40 m in altitude. With Inukjuak, it is one of the lowest elevation communities (Boisson, 2019). Hydrological hazards have the most potential for damage, in addition to the hazards associated with mass movements and permafrost thawing (Carbonneau et al., 2015). A 1-m extreme positive storm surge could submerge some of the infrastructure built at the base of the marine terrace (Carbonneau et al., 2015). According to the data collected (Table 29) between 2015 and 2017, the freeze-up process takes around four weeks. The melting of the ice cover takes three to four weeks. During the melting, blocks of ice can settle in

the bay or be carried with the tides. However, the period of ice observation in Aupaluk over two years (2015- 2016 and 2016-2017) (Table 29) does not cover a long enough period to establish a trend.

Ice processes in Aupaluk (Bernier et al., 2017; 2019; Poulin et al., 2018)

2015-2016 2016-2017 Ice cover stages in Aupaluk Number of days Number of days

Freeze-up duration 27 32 Permanent ice cover duration 196 177 Period between first sighting of water and ice outflow during spring 21 30 melt

2.4.2.3 Tasiujaq

Tasiujaq is located at an altitude of approximately 3 m on a former marine terrace, on a bay on the south shore of the Leaf River basin, at the mouth of the Bérard River. Due to the region’s large tidal range, the foreshore is several kilometres wide and is composed of a field of ice blocks (Boisson, 2019). Possible hazards in Tasiujaq include hydrological hazards like erosion and permafrost thaw (Carbonneau et al., 2015). According to Boisson (2019), a 1-m extreme positive storm surge would not cause damage. On the other hand, 2.5-m positive storm surge would cause the first observable damages on the road leading to the marine infrastructure (Boisson, 2019). Still according to Boisson, climate change will not necessarily lead to an increase in coastal hazards. However, navigation in the Leaf River sector will continue to be very dangerous (Boisson, 2019).

2.4.2.4 Kuujjuaq

Kuujjuaq is located on the west bank of the Koksoak River, in an estuarine transition zone 53 km from the river’s mouth (Boisson, 2019; Bleau, 2012), where there is a 4.6-m tidal range (Figure 3). On the stretch of the river where the community and infrastructure are located, tides are significant. By 2100, the relative sea level could decrease by 45.9 ± 27.1 cm under RCP4.5 and 36.2 ± 33.2 cm under RCP8.5 (Table 3). When factoring in the melting of the West Antarctic Ice Sheet, it is projected that the relative sea level could rise by 40 cm by the end of the 21st century. Extreme negative storm surge levels, which reached -0.72 m in the recent past, could reach -0.71 m by 2040- 2069 and -0.69 m by 2070-2099 (Table 4 and Figure 17), which corresponds to insignificant decreases of 1 cm and 2 cm respectively, as compared to 1989-2009 (Table 4). Extreme positive storm surge levels, which reached 1.01 m in the recent past, could reach 0.92 m by 2040- 2069 and 1.03 m by 2070-2099 (Table 4), which corresponds to a decrease of 9 cm by 2040-2069 and an insignificant increase of almost 2 cm by 2070-2099, as compared to 1989-2009 (Table 4). The breakup processes and mobile ice period last several weeks in the spring. Local knowledge (Bleau, 2012) and camera observations show that between 2009 and 2012, it took 5 to 25 days of mild weather for signs of deteriorating ice cover to emerge near marine infrastructure. In addition, pieces of fast or drifting ice remained present for 25 to 55 days (Clerc et al., 2012). Tidal currents play a key role not only in the ice cover freeze-up, but also in the movement of ice floes, which lasts nearly three months (75 to 90 days) and 25 to 55 days during the melting period (Clerc et al., 2012; Bleau, 2012). The risks increase as the thickness and density of the ice in the estuary grow. Photographs from the CAIMAN network were not analyzed for Kuujjuaq after 2015 (Figure 1) (Bernier et al., 2017 and 2019). By the middle of the 21st century, the spring

ice melt could occur four days earlier, and in the fall, the appearance of landfast ice could be delayed by 12 days (Sennevile and St-Onge Drouin, 2013). At present, roads, houses and other infrastructure would not be affected in the event of an extreme positive storm surge (Boisson, 2019). The most significant natural hazards are associated with the gradual thawing of permafrost and hydrological hazards (Carbonneau et al., 2015), which include ice jams on the Koksoak, according to local knowledge (Bleau, 2012).

2.4.2.5 Kangiqsualujjuaq

Kangiqsualujjuaq is located on the shores of Akilasakallak Bay on the east shore of the George River estuary. This bay is characterized by an intertidal marsh and a large boulder mudflat. The risks associated with positive storm surges are not significant. Only part of the road leading to the port would be flooded by a 1-m surge. In a climate change context, this community will not be significantly affected by major coastal hazards (Boisson, 2019).

3. KNOWLEDGE INTEGRATION FOR THE IMPLEMENTATION OF ADAPTATION MEASURES

“Knowledge Synthesis: Impact of Climate Change on Nunavik’s Maritime and Coastal Environment” is a tool aimed at environment, climate change and land-use professionals, among others, as well as municipal managers. Its purpose is to make accessible the results of recent scientific research, allowing these stakeholders to understand the impacts of climate change while facilitating both the integration of findings into various land-use planning tools and the planning of marine infrastructure maintenance. Adaptation starts with raising awareness among all stakeholders that could be impacted by climate change. To be effective and robust, adaptation measures should generally be based on different types of knowledge and on the experience of a number of professionals and decision-makers. The acceptability of potential solutions can also be facilitated by informing the population of the impacts of climate change (Morin, 2008a). The aim of Chapter 3 is to suggest ways of integrating the knowledge presented in the previous chapters into certain key stages of the adaptation process (Figure 2, Chapter 1). The chapter is divided into four sections: First, a brief conceptual framework with respect to hazards, vulnerabilities and risks is provided (Section 3.1). Suggestions are then given for using the knowledge compiled in the previous chapters in order to better protect and develop the region (Section 3.2). Next, various useful tools for adaptation in Nunavik are presented (Section 3.3). The final section examines avenues for future research in the context of the particularities of northern environments and adaptation issues in these regions (Section 3.4).

3.1 HAZARDS, VULNERABILITIES AND RISKS IN NORTHERN QUEBEC

The term risk refers to the point of contact between a hazard and vulnerability. In other words, there is a risk of harmful consequences when a hazard occurs where there are vulnerable elements (Morin, 2008b). The term “hazard” applies to any natural (flood, storm surge, avalanche, landslide, earthquake, etc.) or anthropogenic (train derailment, nuclear power plant failure, bombing, etc.) phenomenon with a certain probability of occurring in a given place. As described in Chapter 1, the characteristics that make it possible to establish the scale of a hazard in a given situation and environment are the intensity, probability of occurrence (or frequency) and possible spatial location (extent) of its effects (Morin, 2008b). In this context, the identification of hazards makes it possible to better understand the nature of a given phenomenon, to increase knowledge about the impacts of hazards on the environment and to anticipate any needs that may arise in the event of a hazard. The term “vulnerability” refers to “a situation resulting from physical, social, economic or environmental factors that increases the likelihood of damage to exposed elements should a hazard occur (Morin, 2008b).” Chapter 2 described the characteristics of the climate and hydrological hazards that could occur on the Nunavik coastline and the natural physical factors (generally geomorphological in nature) that influence vulnerability and the impacts of hazards. Coastal risks exist where potential hazards coincide with the presence of populations or infrastructure. In Nunavik, all populations and infrastructure are located on the coastline, around the mouths of rivers or estuaries, or around fjords or bays. The 14 Inuit villages and Deception Bay are thus always potentially at risk (Boisson, 2019 and L’Hérault et al., 2017a). Nunavimmiut have already modified some of their lifestyle habits as a result of the changes they have noticed in their environment. For example, they have already modified their roads and seasonal hunting and fishing habits to accommodate the fact that the ice cover now forms later in the season and breaks up more quickly than in recent decades (Clerc et al., 2011). There is so much variation from year to year that in some areas,

it is increasingly difficult to predict whether ice conditions are safe. This increases the risk of accidents (Clerc et al., 2011) and could restrict the use of snowmobiles or other transportation methods employed on the ice. As described in Chapter 2, changes in ice conditions can reduce coastal protection. In the winter, in Nunavik and particularly in Hudson Bay, the presence of immobile and stable sea ice (ice anchored to the sea floor or landfast ice) from January to June limits the impacts of waves and wind on the banks (Savard et al., 2016). If the winds are very strong, the ice may be unable to withstand the stress and become mobile (Atkinson et al., 2016). In the winter, the main travel-related dangers for Inuit are cracks and ice piles that can appear in certain areas sensitive to the impact of tides and sea currents (Clerc et al., 2011). However, it is in spring and autumn that the coasts are more at risk, since a combination of strong winds and high water levels (linked to strong tides) can cause the breaking of the ice cover, which is thinner as a result of milder temperatures. In past climate conditions, people had a good understanding of ice conditions. They are no longer able to rely on their acquired knowledge as a result of the changes they are observing. Thus, the presence of an ice cover may or may not protect the coast, depending on the air temperature and the shape and strength of storm waves that hit the coast. An understanding of ice behaviour and changing ice conditions will help identify both the period when ice is fragile and mobile and the coast is not protected by landfast ice,9 and the period when ice is solid and landfast (Bernier et al., 2017). Meanwhile, a consideration of positive storm surges will help identify new vulnerabilities in flood-prone areas, even if surges become less intense. It will also be possible to better analyze the location of future deep-water ports after examining the fall in relative sea level and the greater fall in negative storm surges, which on the Hudson Bay coast will be most extreme at the end of the 21st century (L'Hérault et al., 2017a). Since the hazards studied for the Nunavik coast have a natural origin, their characteristics cannot be voluntarily modified by humans. As Morin (2008a) points out, when we have little control over a hazard, we must address the vulnerability of the exposed infrastructure in order to reduce the risks. Thus, the only way to mitigate the risks in Nunavik is to reduce the vulnerability of the natural environment (physical factors), infrastructure and communities. This will involve the improvement of stakeholder skills, as well as the power or capacity of stakeholders to take concrete action aimed at reducing vulnerabilities.

3.2 DATA AND KNOWLEDGE SUPPORTING ADAPTATION

An understanding of hazards, vulnerabilities and risks can inform decisions linked to land occupancy and land-use planning. Indeed, this information can be used to analyze the potential impacts of climate change according to region, making it possible to identify and prioritize adaptation measures in the region. In order to show how this could be done in Nunavik, this section presents two examples of analyses that could be used while integrating new information in order to better protect and develop the region. The first is a sensitivity analysis, which combines information related to hazards and vulnerabilities, and the second is a spatial analysis of the risks associated with different climate hazards.

9 The landfast ice season has been defined as the number of days during which landfast ice concentration exceeds 30%.

3.2.1 ANALYZING VULNERABILITIES

An understanding of hazards and natural vulnerability factors makes it possible to assess where the greatest impacts will be felt. In coastal environments, the impacts related to submersion and erosion are likely to cause the most damage. Analysis grids, which assign significance scores to the various hazards in a given region, are one of the methods most frequently used for assessing vulnerabilities. The objective of this method is to analyze the sensitivities in the current climate and compare them to what is expected in the future climate in order to rank the vulnerabilities in order of importance and prioritize adaptation or planning measures. An NRCan study employs this type of grid to analyze the sensitivity of all of Canada’s coasts to different hazards (Manson et al. 2019). The study also uses a Coastal Sensitivity Index (CSI), which is a relative measure of the response of the coastal zone’s particular geomorphology (i.e. coastal composition, slope, presence of ground ice) to two hazards (i.e. changes in sea level and changes in wave height), while factoring in the effects of changes in sea ice (Manson et al., 2019). The study compares the 2006-2020 period to the end of the century. Six variables are categorized according to a sensitivity scale, with scores ranging from 1 for the lowest sensitivity to 5 for the greatest sensitivity (Table 30).

Sensitivity indices used to construct the Coastal Sensitivity Index (CSI) (from Manson et al., 2019)

Scores Variables 1 2 3 4 5 Sea-level change 2006 to 2020 and ≤ -0.33 - 0.32 to -0.20 - 0.19 to 0.20 0.21 to 0.70 > 0.70 2006 to 2100 (in m) Decadal mean wave height (2000 ≤ 0.25 0.26 to 0.75 0.76 to 1.50 1.51 to 2.25 > 2.25 and 2090) in m Permafrost does not Nil to Low and Low Low to Moderate Moderate to High high Ground Ice exist or ground ice is and Moderate Nil - Intrusive rocks - Sedimentary and - Blocks, and rubble - Thick and - Silt, and clay, locally - Metamorphic volcanic rocks with sand and silt continuous till containing stones rocks - Sedimentary rocks - Rubble and silt - Sand and locally - Silt, clay and fine sand - Volcanic rocks - Unknown bedrock - Sand and gravel gravel - Peat, muck and minor - Sand and gravel - Sand and minor inorganic sediments and locally silt - Fluid silty clay and Material diamicton - Stratified silt, clayey silt - Sand, gravel and sand, clay and - Ice and minor morainal pockets of finer gravel debris sediment - Sand, silt and gravel

Slope (in °) > 24 12.1 to 24 5.1 to 12.0 11.1 to 5.0 ≤ 10 Tidal range (in m) 2.1 to 4.0 4.1 to 6.0 1.1 to 2.0 > 6.0 ≤ 10

The index was calculated using the terrain’s hazards and geophysical particularities (the six variables in Table 30) in the form of spatialized data at different scales. These hydrodynamic and geophysical particularities were layered in a geographic information system called CanCoast 2.0. Different topographic maps were used to assess different features of the coast, including a 1:50,000 scale elevation model for slope. Others maps were added, a 1:5,000,000 scale map of geological surfaces, and a 1:7,500,000 scale permafrost map for instance. They are currently too old and should be updated. Projections for wave height, sea ice and post- glacial rebound were made with models using RCP8.5, whose outputs were resampled on a 50-km coastal 48

scale.. The sensitivity index can also be rendered as a qualitative coastal sensitivity scale (i.e. low, moderate or high sensitivity to climate change) that is easier to analyze. The advantage of CanCoast is that it is “open source,” which makes it possible to add more up-to-date Nunavik data and obtain a personalized CSI for each coastal region at a relatively precise 50-km scale. It is in this context that we mapped the Nunavik coastal region (Figure 7) and various communities, as well as Deception Bay (Figures 6 and 9 to 17 in the Appendix), by changing the scale in the current CanCoast2.0. We superimposed the CSI on the positive storm surge (Figure 7) and negative storm surge (Figures 6 and Figures 9 to 17) water levels for the 1980-2009 period and the variations for the two future periods.

Figure 7. Coastal Sensitivity Index, positive storm surges for the 1989-2009 period and positive storm surge variations for future periods (2040-2069 and 2070-2099).

Figure 7 could be completed by including the relevant hydro-climatic hazards (by replacing or adding to the variables in Table 30) as knowledges are improved. Figure 6 and Figures 9 to 17 could also be completed by including hazards that are more specific to each community and its corresponding environment. For example, in Hudson Bay and Hudson Strait, it would be useful to assess the combined effect of strong positive storm surges, strong waves and wave setup, while in Quaqtaq, it would be more relevant to analyze elements related to the mobile ice cover, high water levels and winds blowing towards the coast. In Aupaluk, in Ungava Bay, hydrological hazards should be carefully assessed because this community is built on the sandy deposits of raised beaches less than 40 m high. Consequently, an extreme positive storm surge of 1 m could submerge several pieces of infrastructure at the base of the marine terrace (Carbonneau et al., 2015).

3.2.2 SPATIAL ANALYSIS OF AT-RISK AREAS

The analysis of different areas or communities throughout the region makes it possible to identify and prioritize the necessary land occupancy actions for reducing vulnerabilities and risks. Mapping tools are certainly a great aid for land-use planning that takes climate risks into account. These maps make it possible to overlap different layers of information in order to generate a useful and adapted visual overview of all the hydro-climatic information described in the previous chapters. In addition to the results of studies coordinated by the MTQ, Chapter 2 presents elements that would be worth mapping for all the communities in order to obtain a more complete risk analysis for Nunavik as a whole. Among other things, it could be useful to map the areas at risk for erosion by carrying out morphodynamic monitoring of the coastline (Corriveau et al., 2016). In Nunavik, erosion (the retreat of the coast line) appears to takes place on some shores, as in Umiujaq, while progradation (the advance of land into the sea) is common on other shores, as in Kuujjuarapik (Boisson, 2019). These phenomena have not been thoroughly studied for all communities. The areas of flooding by submersion were mapped by L’Hérault et al. (2013) and Carbonneau et al. (2015) using the maximum tidal level as projected by the models of Massée and Villeneuve (2013), where the lower limit of the flood-prone coastal zone corresponds to the maximum height currently reached by positive storm surges and the upper limit corresponds to the height that may be reached by the end of the century (Boisson, 2019). Coastal hazard maps, which CEN updates for the MSP by community for all of the 14 Nunavik communities, present the flood zones according to the models of Massé and Gallant (2016), including the Akulivik, Kuujjuaq and Salluit flood zones (Allard et al., 2020a, 2020b, 2020c).

3.3 GOING FURTHER: DECISION-MAKING TOOLS

This section presents tools, programs and strategies that can inform and inspire Nunavik professionals in their search for solutions to coastal issues. Although they are often not explicitly defined as adaptation measures, several Arctic communities are already taking concrete actions to reduce the risks posed by climate hazards and their impacts on infrastructure and maritime and coastal environments. In addition, tools deployed outside Nunavik can serve as input to work in Quebec aimed at increasing the resilience of the province’s coastline.

3.3.1 DESIGN AND ADAPTATION IN A CONTEXT OF UNCERTAINTY: A FEW POINTS OF REFERENCE

It was mentioned previously that historical climate averages cannot be relied upon to make local or regional decisions. For example, the increased frequency and intensity of bad weather projected in Nunavik as a result of climate change means that building safety standards will have to be reviewed. Northern stakeholders will necessarily have to rely on climate projections (Ropars, 2014).

Professionals working in Nunavik can refer to the work carried out under the Northern Infrastructure Standardization Initiative (NISI)10 of the Standards Council of Canada (SCC). NISI oversees the design, planning and management of northern infrastructure. Specific standards for northern environments have been developed and training videos are available on the NISI information portal. For example, there are four standards for construction in permafrost regions11 and for adaptations to extreme weather conditions.12 In addition, the portal presents two standards designed by the Bureau de normalization du Québec,13 which have been approved for adaptation to climate change in the Canadian North: BNQ 2501-500: Geotechnical Site Investigation for Building Foundations in Permafrost Zones14 and BNQ 9701-500: Risk-based Approach to Community Planning in Northern Regions, scheduled to be published in 2021.15 These documents are important information sources to consider. The Engineering Protocol (Public Infrastructure Engineering Vulnerability Committee, or PIEVC16) is another useful tool that can serve northern needs. It makes it possible to systematically review historical climate data and project the nature, severity and likelihood of future climate events, as well as the adaptability of individual infrastructure elements based on their design, operation and maintenance. It includes an estimate of the severity of climate impacts on infrastructure components (i.e. deterioration, damage or destruction of infrastructure components) to help determine which elements are most at risk. This information can be used to make informed technical judgments about the different components requiring adaptation and the proper method for making these changes. The protocol is available free of charge under a license agreement with Engineers Canada (n.d.).

3.3.2 USEFUL INFORMATION FOR CLIMATE CHANGE AND LAND-USE PROFESSIONALS AND MUNICIPAL MANAGERS

In collaboration with the authorities and academia, northern stakeholders are already developing decision- support tools in response to the challenges posed by climate change in order to maintain their current activities and improve their capacity for action. Here are several examples: The northern engineering research program Arkuluk now offers technical solutions to the transport infrastructure performance problems associated with unstable permafrost.17 In addition, a quantitative risk analysis framework for linear infrastructure on permafrost, ranging from geotechnical properties to social impacts, has been developed and tested (Brooks, 2019). While it still needs to be refined, the tool was used

10https://www.scc.ca/en/stakeholder-participation/roadmaps-and-standardization-solutions/northern-Infrastructure- standardization-initiative 11https://www.scc.ca/en/nisi/building-in-permafrost 12https://www.scc.ca/en/nisi/extreme-weather 13https://www.bnq.qc.ca/ 14https://www.bnq.qc.ca/en/standardization/civil-engineering-and-urban-infrastructure/geotechnical-site- investigation-for-building-foundations-in-permafrost-zones.html 15https://www.scc.ca/en/nisi/designing-with-climate-change-and-risk 16https://pievc.ca/ 17https://arquluk.gci.ulaval.ca/en/mission-and-objectives/

to determine the hazards for the Salluit airport access road and to carry out a hazard, risk and profitability assessment for the Iqaluit International Airport. PermaSim is a computer application that simulates ground frost and permafrost depth and intensity for different types of soil and vegetation cover in Nunavik, using real climate data specific to each community. This tool is currently used in training sessions prepared by Laval University (Professor Michel Allard) with municipalities and organizations in Nunavik and Nunavut that are looking to improve their understanding of the phenomenon and their general skills. Recently, the MSP and CEN have made efforts to identify current and anticipated risks in the regions of the 14 Nunavik communities in a climate change context (Allard et al., 2020) (see Section 3.2.2). SmartICE (https://smartice.org/) is an award-winning technological innovation for the North developed at Newfoundland. It integrates traditional knowledge of sea ice with advanced data acquisition and remote monitoring technology. Community is an active partner in all project operations, research and decision- making In addition, certain summary works on the Arctic and northern environments are also contributing to the planning and decision-making processes. To this end, NRCan is preparing regional portraits of climate change impacts and adaptation in Canada. Quebec is one of the six regions that will be addressed in the report.18 The ArcticNet Network of Centres of Excellence is carrying out a second iteration of the Integrated Regional Impact Study on climate change for the Nunavik region (IRIS 4), published by Allard et al. (2012). Ouranos’ latest climate change Knowledge Synthesis (2015a) also offers relevant information for the region. In collaboration with its partners, this consortium has also produced a regional climate portrait for the Nunavik region in order to better support vulnerability and impact studies, as well as economic studies and research on sustainable solutions (Charron et al., 2015).

3.4 PROMISING RESEARCH AVENUES FOR NUNAVIK’S COASTAL AND MARINE ENVIRONMENTS

In the context of the adaptation process, it is important to understand hazard occurrence and intensity probabilities in order to carry out a vulnerability analysis and propose actions to reduce impacts. Chapter 1 presented the uncertainties facing researchers with respect to model imprecision (which will certainly be improved in the years to come), the uncertainties linked to the lack of observation data, and the need to continue observations at sites with special characteristics and extend data collection to a greater number of sites like these (Chapter 2). Thus, despite the enormous progress made so far, there are still knowledge gaps to be filled. That being said, the research avenues associated with these gaps are promising.

3.4.1 IMPROVING THE SCIENTIFIC UNDERSTANDING OF THE ENVIRONMENT

This section highlights the elements of interest raised in the various studies piloted by the MTQ and its partners and proposes the appropriation of an NRCan system that would make it possible to advance research and development in Nunavik’s coastal and maritime environments. Strengthen the nascent network of measurement stations

18https://www.nrcan.gc.ca/climate-change/impacts-adaptations/what-adaptation/canada-changing-climate-regional- perspectives/21092?_ga=2.127914442.615067387.1598816558-1732685042.1598816558

The scant data obtained at some of the stations makes it impossible to properly assess the climate models covering the Nunavik region (Ouranos, 2015b) or to calibrate the wave model, which could otherwise assess the probability of concurrent extreme water levels and strong waves (Savard, 2016). It would therefore be advisable to equip all communities with meteorological, tide gauge and wave measuring devices, and to ensure the continued long-term functioning of these facilities. See Figure 1. Apply weather forecasts to the water level model A model similar to the 2D hydrodynamic model used by Massé and Villeneuve (2013 and 2016) to reproduce variations in water levels could be driven with weather forecasts rather than reanalyses, providing coastal communities and managers with real-time operational water level forecasting infrastructure. Simulate the storm movements in Hudson Bay Storm modeling is still in its infancy given that many parameters are still unknown. Storm studies carried out in the Arctic (above 60°N) and in the middle latitudes (on the Atlantic and the Pacific at Canadian latitudes) have provided information on climate model capacity improvement and on the optimal ways to account for uncertainties, offering interesting clues about certain processes that are also important for Hudson Bay. However, the notable differences in atmospheric circulation, the presence or absence of sea ice and the seasonality of cyclonic activity mean that the results of these studies on the Arctic, the Pacific and the Atlantic cannot be easily transferred to Hudson Bay. It is essential to continue studies in this Canadian region, which has an effect on the Nunavik climate, using the new CORDEX-NA and CMIP6 model groups that will soon be available (they include more models with higher resolutions that make it possible to “see” Hudson Bay) – See Technical Summary. Increase the spatial resolution of ice simulations from 10 km to 2 km or less The spatial resolution increase in the ocean variable simulations conducted with the regional ocean model ( ROM) and in the ice simulations (CICE) was not effective, as the simulated ice ratio in the freeze-up and melt period is too large (Senneville, 2018). • Simulate the entire Hudson Bay area, including all the coasts of Nunavik, at the proper resolution for landfast ice to be visible (2-km resolution minimum). • Develop high resolution ice simulators (1- to 10-m resolution has been reached in an area of a few km2 in the St. Lawrence estuary) requiring ice observations with high resolution satellite images (RADARSAT, MODIS and LANDSAT) for results validation and the improvement of high-resolution ice models. • Expand camera network use for ice model validation. It would be useful for research and communities to add new cameras and maintain the activity of those already installed in the CAIMAN network in order to ensure local monitoring at the regional level. Consider coastal processes in studies of extreme water levels Recent studies have shown that submersion assessment through the addition of extreme water levels and the anticipated variation in global sea levels provided by the IPCC, combined with topographic data or a digital elevation model, is not sufficient to estimate the damage caused by submersion. The actual levels responsible for the most significant damage are higher than traditional assessments that consider water level frequency and the rise in global sea level, leaving out wave setup (Didier et al., 2019). The wave setup reaches a higher level than positive storm surges. It would therefore be useful to also assess wave run-up in estimating the total water level (Dumont, personal communication) in subsequent studies. Further develop the study of certain geomorphological hazards and conditions L’Hérault et al. (2017b) suggest further studies on ice pushes, as their impact on morphology and marine infrastructure is still poorly understood. They also suggest carrying out underwater mapping surveys in estuaries, bays and fjords that play an important role in indigenous food collection or are potential sites for

the creation or expansion of port facilities. Boisson (2019) proposes the further study of subaquatic landslides affecting coastal areas, as they could impact the marine infrastructure established on sea clay. Improve the accuracy of Hudson Bay, Hudson Strait and Ungava Bay bathymetry Little is known about the bathymetry of the study area, yet this data is essential for the simulation of waves and water levels. For example, the GEBCO, SHC and DTU10 bathymetric data used by Massé and Villeneuve (2013) each contain approximations that contribute to tidal wave errors (tidal wave propagation is too fast in Hudson Bay or too slow in Hudson Strait and Hudson Bay). Improve the portion of the Nunavik coastline in the CanCoast 2.0 database CanCoast 2.0 is based on so-called “open source” data (i.e. open access, free data). It can be updated when new data sources become available. Changes can be made to different layers of information, such as the post-glacial rebound rate, coastal soil types, or ice and permafrost conditions. However, as Cancoast is a national program, data quality and precision for local studies in Inuit communities are too coarse. Towards a better understanding of Nunavik’s coastal dynamics Currently, past and recent aerial photographs taken between 1960 and 2020 are studied by Laval University and Didier team (UQAR_ISMER) (Boisson et al. 2020). Those Local and more precise studies could help to improve coastal dynamics of several Nunavik’s communities understanding.

3.4.2 IMPROVING KNOWLEDGE FOR A MORE EFFICIENT AND ROBUST ADAPTATION PROCESS

Improve mapping of at-risk coastal areas Carbonneau et al. (2015) have mapped the areas at risk for storm surges, floods, landslides and avalanches for eight of the Nunavik communities (Umiujaq, Inukjuak, Ivujivik, Kangiqsujuaq, Quaqtaq, Aupaluk, Kuujjuaq and Kangiqsualujjuaq). Their analysis is based on field observations and photo-interpretation work (photographs from 2015 and LiDAR surveys from 2010). An update on the report by Carbonneau et al. (2015) was carried out by Boisson (2019) using the 100-year maximum levels modeled by Massé and Gallant (2016). Several community hazard maps were designed in 2020 for the MSP and by CEN (Allard et al., 2020a. 2020b and 2020c). Improve mapping of infrastructure Beyond the analysis of at-risk areas, it is also important to locate critical infrastructure in these areas. These different infrastructures should be classified according to their age, current condition, maintenance and repair history, foundation type, and private or public strategic interest, considering that the strategic value may be greater than the property value (Larrivée, 2010). It is also important to locate transport networks and categorize them according to size, number and type (road, rail, port, electricity, aqueduct, storm drainage, etc.), as well as infrastructures with social utility (aesthetic value, recreational tourism, utility for health or biodiversity, etc.), in order to avoid losing access to structures that play an important role in connecting people together. Evaluate the relevance of adaptation measures Not all adaptation measures are applicable in every location, and the choice of the most appropriate measures depends on several factors. Measures can be more or less restrictive and more or less expensive. But it is important to check that the most applicable measures are being used in the context of the particularities of each coast. Some adaptation solutions may generate additional impacts, or different impacts from the ones that are trying to be avoided.

For example, linear rock piles contribute to wave reflection, which can in turn cause washouts and the disappearance of beaches (Savard et al., 2008) Use decision-support tools Financial economic or cost-benefit (CBA) analyses, for example, can be used to better establish the links between physical impacts and the projected socio-economic costs associated with countering these impacts. This type of analysis has been widely used in other coastal environments in Quebec. For example, the CBA of adaptation measures implemented in the Îles de la Madeleine coastal zone (Circé et al., 2016) could serve as inspiration for the North. In Salluit, a financial analysis of development in the context of climate change was carried out for the Ministère des Affaires municipales, de l’occupation du territoire et de l’habitation (MAMH). This financial analysis examines the land, existing buildings and land-use planning options in order to identify expansion zones while taking into account the impact of climate change, especially on the slopes of the village (Journeaux, 2011).

GENERAL CONCLUSION

The knowledge presented in this document forms the basis for the most up-to-date findings with respect to observations and future trends, all with the aim of better guiding planning in northern environments and decision making. Thanks to sustained efforts in research and development by the MTQ and its partners on the impact of climate change on the marine and coastal environment during the last decade, significant advances in northern research have been made in several areas of expertise. Visual mapping tools have been created to help the reader locate the measurement instruments installed in Nunavik and to see where regional information on the various hydro-climatic hazards has been collected (Chapter 1). These contributions show some disparities between communities with respect to coverage by the data acquisition equipment network. The results of the storm surge extreme water level simulations (positive and negative storm surges) for very rare maximum levels (once every 100 years) and their variations for the 2040-2069 and 2070-2099 periods are presented in Chapter 2. The maps of the projected monthly ice conditions for the bodies of water surrounding Nunavik are compiled in an ice condition atlas (Senneville, 2018a) available from the MTQ (Chapter 2). At this stage, the findings on changes in storm processes are only preliminary and will require special attention in order to identify clear trends. New findings in connection with ISMER’s work on waves will be available in 2021. Ultimately, sustained acquisition of hydro-climatic hazard data would provide the region with a more comprehensive picture of the current and future impacts and risks suggested by the data. In addition, continuing the efforts made by authorities thus far will help the region to better coordinate data acquisition and integration, which will in turn facilitate risk management in isolated areas and contribute to the implementation of more suitable, context-specific adaptation solutions. Elsewhere in the world, coastal areas must contend with the impacts caused by the rise in the relative sea level. In Nunavik, however, post-glacial rebound compensates for these effects on the marine environment. Indeed, a 40- to 90-cm fall in relative sea level could occur depending on the considered community and scenario. However, the beneficial effects of this compensation will only be felt at the end of the 21st century. That being said, the coastal environment will still have to deal with impacts related to submersion and erosion at the turn of the century. In addition, the region will also be affected by thawing permafrost and increased river flows. Here are a few highlights to consider in the context of improving environmental protection and coastal development in Nunavik. Nunavik Extreme storm surge water levels could rise or fall by as much as ± 1m in the 2040-2069 and 2070-2099 periods, which is very close to the tidal range in many communities. By the end of the 21st century, the frequency of these extreme water levels could increase if global warming continues to escalate. The period during which extreme water levels could occur more frequently could be longer. Ice could form later and later in the fall and earlier and earlier in the spring, leading to a period of partial ice cover more than six weeks longer by 2040-2070 and more than two months longer by 2070-2100. The likelihood of ice erosion would then be higher. Hudson Bay 56

The increases in temperature and total winter precipitation may be more pronounced on Hudson Bay. Extreme negative storm surges could lower the minimum water level by 70 cm to 1 m by 2040-2069 and by 80 cm to 1.1 m by 2070-2099. Winter negative storm surges could take place more often in early winter and in March. 100-year extreme positive storm surges could be 10 to 20 cm higher than from 1989 to 2009. They could occur even more often in the fall. Ice presence has already decreased by more than 10% between 1968 and 2010. In addition, the ice cover is expected to almost disappear (10%) between Ivujivik and Inukjuaq by 2040-2070. Hudson Strait and Ungava Bay It is anticipated that storm surges (both positive and negative storm surges) could be strictly less than 1 m. Thus, the risks associated with positive storm surges are low and those associated with negative storm surges are very low. Ice concentrations in December could decrease by 40 to 60% on the coasts from Ivujivik to Kangiqsualujjuaq by 2040-2070. The Arctic is warming two times faster than the rest of the planet. We must therefore expect the impacts predicted in the North, and in Nunavik especially, to occur sooner. One way to decrease the serious repercussions of these impacts in Nunavik is to reduce the vulnerability of the natural environment, infrastructure and communities. There are already a number of ways to improve stakeholder skills, as well as the power or capacity of stakeholders to take concrete action aimed at reducing vulnerabilities. For example, the application of construction standards specific to the realities of Nunavik (BNQ 2501-500 for building foundations in permafrost) or the protocol on the vulnerability of public infrastructure engineering (https://pievc.ca) makes it possible to estimate the climate change adaptative capacity of infrastructure components. Northern engineering programs (https://arquluk.gci.ulaval.ca/) and summary works produced by Arctinet, CEN, RNCan and Ouranos also contribute to planning and decision-making processes. While decision-makers are advised to make decisions with up-to-date scientific knowledge, it must also be recognized that our understanding is incomplete in Nunavik and decisions must still be made to protect coasts and people in spite of this fact. There are several promising research avenues for improving the scientific understanding of climate risks and making the adaptation process more efficient and robust. This synthesis tool can be supplemented by consulting the Technical Synthesis, which provides information on the data collection and modeling processes and methods. The knowledge presented in the previous chapters is also available in a more accessible flyer or as targeted conferences.

GLOSSARY

Adaptation: Adaptation is a change in behaviors and characteristics of a system to be able to deal with a new specific situation. Adaptation to climate change refers to specific adjustments rely to a rapid climate evolution. Adaptative capacity: The capacity of systems (all possibilities, resources and institutions of a country, a district, a collectivity or a group) to implement efficient adaptation measures to climate change (including climate variability and extremes), to adjust to potential damage, to take advantage of opportunities, or to respond to consequences. Climate projection: A simulated response of the climate system to a scenario of future emission of greenhouse gases (GHGs) and aerosols, derived using climate models. Climate projections depend on the emission, concentration, radiative forcing scenario used, which is based on assumptions concerning, for example, future socio-economic and technological developments that may or may not be realized. Then, a climate projection is not a climate prediction. Ebb current: The horizontal movement of water associated with the falling tide. Ebb currents generally set seaward, or in the opposite direction to the progression. Also called ebb, ebb current or outgoing Extreme water levels: Water levels that have a very small recurrence. They happen when a spring tide occurs simultaneously with a positive storm surge (due to a storm surge) and winds blow towards the coast. They are a lot higher than the relative sea level annual average. Fetch: Free water horizontal distance over which wind does not encountering obstacles and blows continuously generating waves. Its intensity changes with the wind direction. The longer the fetch, the more important is the energy transferred to waves. The term is also used as a synonym for fetch length. Flood current: The horizontal movement of water associated with the rising tide. Flood streams generally set toward the shore, or in the direction of the tide progression. Also called flood, flood current or in going stream. Frequency: It is a probability (or a number of repetitions), of one phenomenon occurs or is surpassed in a given year. A 0,01 recurrence means that the considered phenomenon has 0,01 chance to be repeated each year. Freeze-thaw weathering: A process of erosion that happens in cold regions where ice forms. A crack in a rock can fill with water which then freezes as the temperature drops. As the ice expands, it pushes the crack apart, making it larger. This process continues, following freeze/thaw cycles, until, over the long term, the rock breaks and is transported elsewhere (wind, water). Hazard: The potential occurrence of a natural or human-induced event or trend that may case loss of life, injury, or other health impacts, as well as damage and loss to property, infrastructure, livelihoods, service provision, ecosystems and environmental resources. Inverted barometer effect: Response of the ocean to atmospheric pressure loads resulting by either a rising of level sea in the center of a Low pressure (storm) or a decrease of level in the center of a High pressure (anticyclone). Impact: Effects on natural ecosystems (physical impacts: floods, droughts and sea level rise) and human systems (human impacts: effects on lives, livelihoods, health, economies, societies, cultures, services and infrastructure) of extreme weather and climate events and climate change within a specific time period. Impacts are also referred to as consequences and outcomes.

Landfast ice: Sea ice that is anchored to the shore or ocean bottom and does not move with wind or currents. Neap tide (tidal currents): The tides of decreased range or tidal currents of decreased speed occurring near the times of the first and last quarters of the Moon. Negative storm surge: Locally measured or observed water height which is below the predicted mean water height (or level 0 m) at the same moment due to a storm. It is also called “reverse surge”. Positive storm surge: Locally measured or observed water level above the daily tide (or level 0 m) at the same moment, due to a storm activity nearby. Post-glacial rebound: Over most of Canada, it is the dominant source of vertical land motion. The large ice sheets thaw, during and following deglaciation, whom the weight depressed the land surface of Canada where they were located, took of the pressure and land surfaces began to rise. The time scale of Earth’s reaction is thousands of years, and it is still occurring today, thousands of years after deglaciation. It is also called glacial isostatic adjustment (GIA). Reanalysis: Atmospheric and oceanic analyses of temperature, wind, current, and other meteorological and oceanographic quantities, created by processing past meteorological and oceanographic data using fixed state-of-the-art weather forecasting models and data assimilation techniques. The reanalyse temporal continuity is improved but spatial uncertainties at global scale are still present due to changing coverage and biases in the observing systems. Relative sea level: On the coast, the relative sea level is measured on a fixed shore point. In Canada and Nunavik in particular, the postglacial rebound plays an essential role in the changes to relative sea levels because it compensates the mean sea level rise. Representative Concentration Pathways (RCP’s): Climate scenarios that include time series of emissions and concentrations of the full suite of greenhouse gases and aerosols as well as chemically active gases, and land use/ land cover. The word representative signifies that each RCP provides only one of many possible scenarios that would lead to the specific radiative forcing characteristics. The term pathway emphasizes that not only the long-term concentration levels are of interest, but also the trajectory taken over time to reach that outcome. Resilience: It is the capacity of social, economic and environmental systems to reorganize and adapt themselves in the wake of a hazardous event or disturbance, stress or shock factors, by recovering from damages and returning to their first equilibrium state, in ways that maintain their essential function, identity and structure. Return period: Period of time between two occurrences of a phenomenon. A return period of 100 years (or centennial) means that the considered phenomenon has 1 chance out of 100, each year to occur. If the return period decreases (50 or 30 years) then the phenomenon occurs more frequently. Also called recurrence interval. Risk: It is a probability of a dangerous event (hazard) has consequences (impacts) on something of value. It is more or less important due to hazards’ recurrence, and vulnerability of the exposed elements which multiply the impacts if these hazards. Sea ice: Any form of ice found at sea that has originated from the freezing of sea water. Spring tide: The tides of increased range occurring near the times of full moon and new moon. Storm surge: Local and abnormal rise in sea level, lasting from a few hours to a few days, caused by a violent action of the wind on the sea surface and / or by the drop in atmospheric pressure (when storm occurs). A storm surge can be more severe if it occurs in conjunction with a high tide. Also called storm tide, storm wave, tidal wave. A storm surge is devised in Positive and negative storm surge.

Submersion: It is a seawater flood of a coastal area higher than the high astronomical tide, in a natural or artificial area, above and beyond the height of the first line of the coastal protections. Tidal range: For a given day, the tidal range is the difference between vertical distance between consecutive low and high sea water limits. It included between the spring tide position and the neap tide position. Tide: The periodic rise (high tide) and fall (low tide) of the surface of oceans, bays, etc., due principally to the gravitational interactions between the Moon, Sun and Earth. Time series: A data time series is a succession of observed or simulated data that allow the calculation of a trend. To drive an RCM (or to nest an RCM): To supply regional climate model (RCM), which has spatially limited grid cover (45 km or less) with values of variables from global climate model (GCM) simulation or reanalyses at the lateral boundaries and ocean surfaces. The values at the boundaries are gradually combined from a large scale to a finer-scale at each of its time steps. Then the simulation produced by the RCM over a region stays coherent with the planetary climate. Total water level: The total observed seawater level resulting from the combination of storm surge, setup, run-up and the astronomical tide. Vulnerability: The degree to which a system is susceptible to, and unable to cope with adverse effects of climate change. It is a function of the system location, of the character, magnitude and rate of change to which the system is exposed and its sensitivity and adaptive capacity. Wave setup or swash: Water mass projected on an upper shore by wave surges. Its maximum height depends on the period and significant height of the waves and the beach slope on which those unfurl. The maximum elevation reached by the wave setup is called the run-up.

REFERENCES

Akperov M., Rinke A., Mokhov I. I., Semenov V. A., Parfenova M. R., Matthes H., Adakudlu M., Boberg F., Christensen J. H. Demitskaya M. A. Dethloff K., Fettweis X., Gutjahr O., Heinemann G., Koenigk T., Koldunov N. V. Laprise R., Mottram R., Nikiéma O., Sein D., Sobolowsi S., Winger K, Zhang W. (2019). « Future projections of cyclone activity in the Arctic for the 21s century from regional climate models (Arctic-CORDEX) ». Global and Planetary Change. 182. 103005. Allard M., Aubé-Michaud, S., L’Hérault, E., Mathon-Dufour, V., Deslauriers, C. et Chiasson, A., (2020a). « Identification des risques actuels et appréhendés sur le territoire des communautés du Nunavik en fonction des changements climatiques – phase 2 : Document synthèse, communauté d’Akulivik ». Rapport final réalisé pour le compte du ministère de la Sécurité publique (gouvernement du Québec), Centre d’études nordiques, Université Laval, 63 p. Allard M., Aubé-Michaud, S., L’Hérault, E., Mathon-Dufour, V., Deslauriers, C. et Chiasson, A., (2020b). « Identification des risques actuels et appréhendés sur le territoire des communautés du Nunavik en fonction des changements climatiques – phase 2: Document synthèse, communauté de Kuujjuaq ». Rapport final réalisé pour le compte du ministère de la Sécurité publique (gouvernement du Québec), Centre d’études nordiques, Université Laval, 67 p. Allard M., Aubé-Michaud, S., L’Hérault, E., Mathon-Dufour, V., Deslauriers, C. et Chiasson, A., (2020c). « Identification des risques actuels et appréhendés sur le territoire des communautés du Nunavik en fonction des changements climatiques – phase 2 : Document synthèse, communauté de Salluit ». Rapport final réalisé pour le compte du Ministère de la Sécurité publique (gouvernement du Québec), Centre d’études nordiques, Université Laval, 74 p. Allard M., Lemay M., Barrett M., Sheldon T. et Brown R. (2102). « De la science aux politiques publiques au Nunavik et au Nunatsiavut : synthèse et recommandations – une étude intégrée d’impact régional des changements climatiques et de la modernisation ». Dans Allard M. et M. Lemay (Éd), Nunavik and Nunatsiavut : From science to policy. An Integrated Regional Impact Study (IRIS) of climate change and modernization. ArcticNet Inc., Québec, Cana, 72 pages. Allard M., L’Hérault E., Gibéryen T. et Barrette C. (2010). « L’impact des changements climatiques sur la problématique de la fonte de pergélisol au communauté de Salluit, Nunavik ». Rapport final : S’adapter et croître. Allard, M. et Lemay, M. (2013). « Le Nunavik et le Nunatsiavut : de la science aux politiques publiques ».

Atkinson, D.E., D.L. Forbes et T.S. James. (2016). « Dynamic coasts in a changing climate; in Canada’s Marine Coasts in a Changing Climate », (ed.) D.S. Lemmen, F.J. Warren, T.S. James and C.S.L. Mercer Clarke; Government of Canada, Ottawa, ON, p. 27-68.

Bernier, M., Gignac, C., Chokmani, K., Poulin, J. et Y. Gauthier (2016). « Un atlas interactif sur la probabilité de l’aléa glace à l’échelle des infrastructures maritimes et côtières dans un contexte de changements climatiques ». Rapport final. Québec, INRS, 62pages.

Bernier M., Poulin J., Ratsimbazafy T, Gauthier Y. et D. Éthier. (2019). « Suivi des conditions de glaces de rive et de vagues à l’aide de caméras et d’imagerie satellitaire à proximité d’infrastructures maritimes au Nunavik dans un contexte de changements climatiques : Kuujjuarapik, Umiujaq, Ivujivik, Baie Déception, Quaqtaq et Aupaluk (2017-2020) ». Rapport d’étape 2 remis au Bureau de la coordination du Nord-du-Québec, Ministère des Transports du Québec, INRS R1852, 28 février 2019, 72 pages.

Bernier, M., Poulin, J., Gauthier, Y., Gignac, C. (2017). « Suivi des conditions de glaces de rives à proximité d’infrastructures maritimes au Nunavik dans un contexte de changements climatiques : Kuujjuarapik, Umiujaq,

Ivujivik, Baie Déception, Quaqtaq et Aupaluk ». Rapport final réalisé pour le compte du Ministère des Transports du Québec. Institut national de la recherche scientifique, Québec, 155 pages.

Bleau, S. (2012). « Étude du comportement des glaces dans un environnement subarctique en régime macrotidal, estuaire de la rivière Koksoak, Nunavik ». Mémoire pour l’obtention du grade M.Sc. Sciences de l’eau. INRS, centre Eau, Terre, Environnement (M-1341). 235 pages. Directrice : M. Bernier, Codirecteur : M. Allard.

Boisson A., Didier D., Allard M., Bélanger A., Bernatchez P., (2020). Towards a better understanding of Nunavik’s coastal dynamics. Arctic Change 2020, Comes to you!, December 7-10, 2020

Boisson A. (2019). « Caractérisation et modèles d’évolution des environnements côtiers du Nunavik », Thèse de doctorat en sciences géographiques, Université Laval, Québec, Canada, 267 pages.

Boyer-Villemaire U. (2016). « Évaluation intégrée de la vulnérabilité des communautés côtières faisant face aux aléas naturels dans un contexte de changements climatiques : les cas d’Avignon (Canada), Kilkeel (Royaume-Uni) et Chipiona (Espagne) ». Thèse de doctorat en sciences de l’environnement, UQAR.

Boyer-Villemaire U., Savard J.-P. et Roy P. (2016) « Évaluation des niveaux d’eau extrêmes causant des dommages de submersion en zone côtière au Québec ». Ouranos, Montréal. 30 pages.

Brooks, H. (2019). « Quantitative Risk Analysis for Linear Infrastructure Supported by Permafrost: Methodology and Computer Program) ». Thèse de doctorat génie civil, Université Laval, 434 pages. Consulté sur https://arquluk.gci.ulaval.ca/publications_et_documents/theses_et_memoires/.

Bush, E. et Lemmen, D.S., éditeurs : « Canada’ Changing Climate Report », Government of Canada, Ottawa, Ontario, 2019, 446 pages. https://changingclimate.ca/CCCR2019/

Carbonneau, A.-S., L’Hérault, E., Aubé-Michaud, S., Taillefer, M., Ducharme, M.-A.,Pelletier, M., & Allard, M. (2015). « Production de cartes des caractéristiques du pergélisol afin de guider le développement de l'environnement bâti pour huit communautés du Nunavik ». Rapport final. Québec, Centre d'études nordiques, Université Laval. 127 pages.

Charron, I. (2015). « Élaboration du portrait climatique régional du Nunavik », Ouranos, Montréal, 86pages.

Charron, I. (2016). « Guide sur les scénarios climatiques : Utilisation de l’information climatique pour guider la recherche et la prise de décision en matière d’adaptation », Édition 2016. Ouranos, 94 pages.

Circé, M., Da Silva, L., Duff, G., Boyer-Villemaire, U., Corbeil, S., Desjarlais, C., Morneau F. (2016) « Analyse coûts- avantages des options d’adaptation en zone côtière aux Îles-de-laMadeleine ». Ouranos, Montréal. 174 pages et annexes

Clerc, C., Poulin, J., Gauthier, Y., Bernier, M., Bleau, S., Gignac, C., Bédard, J.S., & Duhamel-Beaudry, E. (2012). « Descripteurs et indicateurs de la couverture glacielle au Nunavik: Quaqtaq, Umiujaq et Kuujjuaq ». Rapport de recherche n° R1389 remis au ministère des Transports du Québec, au Consortium Ouranos et au ministère des Affaires autochtones et Développement du Nord Canada. Institut national de la recherche scientifique, 186 pages.

Clerc, C., M. Gagnon, K. Breton-Honeyman, M. Tremblay, S. Bleau, Y. Gauthier, S. Aloupa, A. Kasudluak, C. Furgal, M. Bernier et M. Barrett (2011). « Changements climatiques et infrastructures marines au Nunavik - Connaissances locales et point de vue des communautés de Quaqtaq, Umiujaq et Kuujjuaq ». Québec, INRS - Centre Eau Terre Environnement, ix, 144 pages. (rapport de recherche; 1273f).

Comtois, C., Slack, B., Champagne-Gélinas, A., Savard, S. et Lagacé, M. (2020) « Transport maritime au Nunavik: Vulnérabilité, opportunités et défis d’adaptation », Rapport interne au Bureau de la coordination du Nord-du-

Québec, Ministère des Transports du Québec, Rouyn-Noranda: Ministère des Transports du Québec, - non public, 332 p.

Corriveau, M., Fraser, C., Caron, T., Bernatchez, P., Buffin-Bélanger, T. Van-Wierts, S., (2016). « Étude de la dynamique morphosédimentaire des côtes basses sablonneuses en bordure de la route 138 sur la Côte-Nord du Saint- Laurent en contexte de changements climatiques » : Rapport final. Projet X016.1. Laboratoire de dynamique et de gestion intégrée des zones côtières, Université du Québec à Rimouski. Rapport remis au ministère des Transports du Québec, Mars 2016, 421 p. + annexes.

Derksen, C., Smith, S.L., Sharp, M., Brown, L., Howell, S., Copland, L., Mueller, D.R., Gauthier, Y., Fletcher, C.G., Tivy, A., Bernier, M., Bourgeois, J., Brown, R., Burn, C.R., Duguay, C., Kushner, P., Langlois, A., Lewkowicz, A.G., Royer, A. et Walker, A. (2012). « Variability and change in the Canadian cryosphere ». Climatic Change, 115(1), 59–88. doi:10.1007/s10584-012-0470-0.

Didier D., Bandet M., Bernatchez P., Dumont D. (2019). « Modelling Coastal flood propagation under sea level rise: a case study in Maria, Eastern Canada ». Geosciences, 9, 76, doi:10.3390/geosciences9020076.

Ford, J.D., T. Bell et N.J. Couture. (2016) « Perspectives relatives à la région de la côte Nord du Canada », Chapitre 5 dans Le littoral maritime du Canada face à l’évolution du climat, D.S. Lemmen, F.J. Warren, T.S. James et C.S.L. Mercer Clarke (éd.); Gouvernement du Canada, Ottawa (Ontario), p. 153–208.

Hill P.R., Meulé S. et Longuépée H. (2003). « Combined-flow processes and sedimentary structures on the shoreface of the wave-dominated Grande-rivière-de-la-Baleine delta. »Journal of SedimentaryResearch, 73(2) : 217-226.

Ingénieurs Canada, Bureau canadien des conditions d’admission en génie, (s/d). « Principes d’adaptation aux changements climatiques à l’intention des ingénieurs », 40 pages. https://engineerscanada.ca/sites/default/files/Principes-adaptation-changement-climatique.pdf

Institut de la statistique du Québec (ISQ). (2019). « Le bilan démographique du Québec, Édition 2019 », [En ligne], Québec, L’Institut, 180 pages. https://www.stat.gouv.qc.ca/statistiques/population- demographie/bilan2019.pdf

Intergovernmental Panel on Climate Change (IPCC), (2019). « Summary for policymakers ». In : IPCC Special Report on the Ocean and Cryosphere in a Changing Climate. [H. O. Pörtner, D. C. Roberts, V. Masson-Delmotte, P. Zhai, M” Tignor, E” Poloczanska, K. Mintenbeck, M” Nicolai, A. Okem, J” Petzold, B” Rama, M” Weyer (eds.)]. In Press.

Intergovernmental Panel on Climate Change (IPCC). (2014). Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, 151 pp

Intergovernmental Panel on Climate Change (IPCC). (2014). « Annex II: Glossary [Mach, K.J., S. Planton and C. von Stechow (eds.)]. In: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change » [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, pp. 117-130.

Intergovernmental Panel on Climate Change (IPCC), (2013). « Climate Change 2013 – The Physical Science Basis- Summary for policymakers ». Working Group I. [Edited by Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex et P.M. Midgley]. Cambridge University Press, Cambridge, United Kingdom and New York, USA.

James T. S.,Henton J. A., Leonard L. J., Darlington A., Forbes D. L., et Craymer, M. (2014) « Relative Sea-level Projections in Canada and the Adjacent Mainland United States », GEOLOGICAL SURVEY OF CANADA - OPEN FILE 7737-. doi: 10.4095/295574.

Joly, S., S. Senneville, D. Caya et F. Saucier (2011).« Sensitivity of Hudson Bay Sea ice and ocean climate to atmospheric temperature forcing ».Climate Dynamics, 36(9) :1835-1849. DOI: 10.1007/s00382-009-0731-4

Journeaux N. (2011). « Analyse technique et financière du plan d’aménagement du village de Salluit et recommandations sur la réalisation d’études géotechniques au Nunavik. ». Rapport noL-10-1398-Rév.4. Affaires municipales, Régions et Occupation du territoire. Québec. 56 pages.

Larrivée C., (2010). « Élaborer un plan d’adaptation aux changements climatiques. Guide destiné au milieu municipal québécois », Ouranos, Montréal (Québec), 48 pages.

Lemmen, D.S., Warren, F.J. and Mercer Clarke, C.S.L. (2016): Introduction; in Canada’s Marine Coasts in a Changing Climate, (ed.), D.S. Lemmen, F.J. Warren, T.S. James and C.S.L. Mercer Clarke; Government of Canada, Ottawa, ON, p. 17-26.

L'Hérault E., Allard M., Fortier D., Carbonneau A-S., Doyon-Robitaille J., Lachance M-P., Ducharme M-A., Larrivée K., Grandmont K., Lemieux C. (2013). « Production de cartes prédictives des caractéristiques du pergélisol afin de guider le développement de l'environnement bâti pour quatre communautés du Nunavik ». Rapport final. Québec, Centre d'études nordiques, Université Laval, 90 pages.

L’Hérault, E., Boisson, A., Allard, M., Aubé-Michaud, S, Sarrazin, D., Roger, J. et C. Barrette (2017a). « Détermination et analyse des vulnérabilités du Nunavik en fonction des composantes environnementales et des processus physiques naturels liés au climat ». Rapport final. Réalisé pour le compte du Ministère des Forêts, de la Faune et des Parcs, Gouvernement du Québec. Centre d’études nordiques, Université Laval, 160 pages.

L’Hérault E., Boisson A., Allard M., Aubé-Michaud S., Sarrazin D., Roger J. et Barette C. (2017b). « Répartition spatiale des aléas naturels recensés au Nunavik ». Centre d’études nordiques, échelle 1 : 3 350 000.

Logan, T., I. Charron, D. Chaumont et D. Houle. (2011). « Atlas de scénarios climatiques pour la forêt québécoise ». Ouranos et MRNF. 55pages +annexes.

Massé, A. et M. Villeneuve (2013). « Infrastructures maritimes du Nunavik – Modélisation numérique des marées et des ondes de tempête dans la baie d’Hudson, le détroit d’Hudson et la baie d’Ungava ». Rapport présenté à Ouranos, Le Groupe-Conseil LaSalle Inc., 45 pages. + annexes.

Massé, A. et N. Gallant (2016). « Marée et ondes de tempêtes dans la baie d’Hudson, la baie James, le détroit d’Hudson et la baie d’Ungava – Modélisation numérique des niveaux d’eau actuels et futurs dus aux changements climatiques ». Rapport présenté à Ouranos, LaSalle NHC, Montréal, 31 pages. + annexes.

Ministère des Transports du Québec (MTQ) (2011). « Projet de recherche visant à évaluer l’impact des changements climatiques sur les infrastructures maritimes du Nunavik et à déterminer les solutions d’adaptation ». Programme de travail – Transports Québec. 24 pages.

Morin, M., (2008a). « Approches et principes en sécurité civile. Ministère de la Sécurité publique ». ISBN 978-2-550- 54256-8. 70 pages

Morin, M., (2008b). « Concepts de base en sécurité civile. Ministère de la Sécurité publique ». ISBN 978-2-550-54254- 4. 60 pages.

Neumeier U., Joly S., Collin S., (2019), « Suivi, analyse et modélisation des conditions de vagues en milieu côtier au Nunavik en fonction des conditions de glace dans un contexte de changements climatiques ». Projet de recherche CC16.1. Rapport d'étape 1 : Rapport produit pour le compte du Bureau de la coordination du Nord- du-Québec, Ministère des Transports du Québec. Février 2019. Uqar- ISMER. 45 pages.

Pêches et océans (2019) Tables de marées et courants du Canada. Service hydrographique du Canada, Ottawa.

Poulin, J., Bernier, M., Ratsimbazafy, T. (2018). « Suivi des conditions des glaces de rive et de vagues à l’aide de caméras et d’imagerie satellitaire à proximité d’infrastructures maritimes au Nunavik dans un contexte de changements climatiques: Kuujjuarapik, Umiujaq, Ivujivik, Baie Déception, Quaqtaq et Aupaluk (2017-2020) ». Rapport d’étape no. 1 réalisé pour le compte du Ministère des Transports, de la Mobilité durable et de l’Électrification des transports du Québec. Institut national de la recherche scientifique, Québec, 59 pages.

Ropars, Y. (2014). « Étude sur la vulnérabilité des infrastructures maritimes du Nunavik aux niveaux d’eau extrêmes en conditions de changement climatique ». Rapport remis au Ministère des Transports du Québec, CIMA +, 45 pages. + annexes.

Ouranos (2015a). « Vers l’adaptation. Synthèse des connaissances sur les changements climatiques au Québec. Partie 1 : Évolution climatique au Québec ». Edition 2015. Montréal, Québec : Ouranos, 114 pages.

Ouranos (2015b). « Vers l’adaptation. Synthèse des connaissances sur les changements climatiques au Québec. Partie 2 : Vulnérabilités, impacts et adaptation aux changements climatiques ». Édition 2015. Montréal, Québec :Ouranos, 234 pages.

Ouranos (2015c). « Vers l’adaptation. Synthèse des connaissances sur les changements climatiques auQuébec. Partie 3 : Vers la mise en œuvre de l’adaptation ». Édition 2015. Montréal, Québec : Ouranos. 49 pages.

Savard, J.-P. (2016). « Impacts des changements climatiques sur le régime des tempêtes et les niveaux d’eau dans la baie d’Hudson, la baie James et le détroit d’Hudson ». Rapport présenté à la Division des impacts et de l’adaptation liés aux changements climatiques, Ressources naturelles Canada, Montréal, Ouranos.

Savard, J.-P., van Proosdij, D. et O’Carroll, S. (2016). « Perspectives relatives à la région de la côte Est du Canada », dans Le littoral maritime du Canada face à l’évolution du climat, D.S. Lemmen, F.J. Warren, T.S. James et C.S.L. Mercer Clarke (éd.), Gouvernement du Canada, Ottawa (Ontario) p. 99–152.

Savard, J.-P., Gachon P., Rosu C., Aider R., Martin P., et C. Saad. (2014) « Impact des changements climatiques sur le régime des tempêtes, les niveaux d’eau et les vagues dans le Nunavik », Consortium Ouranos. Rapport d’étude pour le MTQ. Octobre 2014, 109 pages.

Savard J-P, Bernatchez P., Morneau F., Saucier F., Gachon P., Senneville S., Fraser C., Jolivet Y. (2008). « Étude de la sensibilité des côtes et de la vulnérabilité des communautés du golfe du Saint-Laurent aux impacts des changements climatiques ». Rapport synthèse. Éd : Ouranos. 58 pages. ISBN : 978-2-923292-02-1.

Senneville, S. (2018a). « Atlas mensuel climatique de la concentration et épaisseur de glace au Nunavik ». (Annexe rapport finale MTQ, projet CC05.1) 74 pages.

Senneville, S. (2018b). « Modélisaton des glaces de rive à fine échelle à proximité d’infrastructures maritimes au Nunavik en contexte de changements climatiques : Kuujjuarapik, Umiujaq, Ivujivik, Baie Déception Quaqtaq et Aupaluk ». Rapport final remis au Bureau de la coordination du Nord du Québec, Ministère des Transports du Québec. Projet CC05.1. 15 octobre 2018. 67 pages.

Senneville, S. et S. St-Onge Drouin (2013), Étude de la variation des glaces dans le système couplé océan-glace de mer de la baie d’Hudson, Rapport final remis au Ministère des Transports du Québec et au Consortium Ouranos, 109 pages.

Shah, C., Ford J., Labbé J., Flynn M. (2018) Rapport final : Adaptation policy and practice in Nunavik, Rapport final remis au Consortium Ouranos, Université McGill, 80 pages.

APPENDIX – WATER LEVEL AND COASTAL SENSITIVITY MAPS

DISCLAMER ON THE FOLLOWING MAPS The Coastal sensitivity Index used here shouldn’t considered to take landing and managing decisions. Indeed, the data used to create the CSI have a very small and inaccurate scale for local analysis – community scale (See chapter 3). The following maps are given only to show the improving direction possible to allow new coastal environment knowledges.

Figure 1. Umiujaq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods

Figure 2. Inukjuak coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods

Figure 3. Puvirnituq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods

Figure 5. Ivujivik coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods

Figure 6. Deception Bay coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods

Figure 7. Kangiqsujuaq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods

Figure 8. Quaqtaq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods

Figure 9. Kuujjuaq coastal climate change sensitivity, 100-year surge levels for the recent past and variations in 100-year extreme negative and positive storm surges for the 2049-2069 and 2070-2099 periods