Canadian Arctic Bathymetry Data: Compilation
Transcription
Canadian Arctic Bathymetry Data: Compilation
Transport Canada’s Northern Transportation Adaptation Initiative (NTAI) Canadian Arctic Bathymetry Data: Compilation, Assessment and Prioritization Client Transport Canada March 2014 -1- Avis : Les opinions exprimées dans ce rapport sont celles des auteurs et ne reflètent pas nécessairement celles du Gouvernement du Canada ou de Transports Canada, et sont propres au CIDCO, prestataire et auteur de ce rapport. Le CIDCO se réserve les droits de propriété intellectuelle de ce rapport. Disclaimer: The opinions expressed in this report reflect the views of the authors and not necessarily the official views or policies of the Government of Canada or Transport Canada, and are specific to CIDCO, author of the report. CIDCO reserves the intellectual property rights of this report. Ce rapport est disponible en ligne sur le site internet du CIDCO : http://www.cidco.ca/liste_publication.php This report is available online on the CIDCO website: http://www.cidco.ca/liste_publication.php?langue=en Equipe de recherche / Project team: - Nicolas Seube, Directeur Scientifique du CIDCO Mathieu Rondeau, géomaticien Jean-Guy Nistad, géomaticien Sylvain Gautier, géomaticien -2- Sommaire: Ce rapport tente d’expliquer le développement de la cartographie marine dans le Nord Canadien ème au vu des différents moyens technologiques disponibles du début du 20 siècle à nos jours. On s’intéresse ensuite aux différentes routes maritimes existantes ou à venir dans le Nord Canadien, et on explique les besoins associés de cartographie afin d’améliorer la sécurité de navigation. On explique la faible qualité des données hydrographiques et produits cartographiques dans certaines zones à un manque d’infrastructures de positionnement, à une connaissance imparfaite de la géodésie et des niveaux d’eau, qui sont des préalables essentiels pour une cartographie de qualité. On passe ensuite en revue différentes solutions envisagées à l’heure actuelle (déploiement de systèmes autonomes, bathymétrie collaborative) pour combler le déficit de données hydrographiques pour assurer la sécurité de navigation, et on propose une analyse critique de l’emploi de ces systèmes dans le contexte du Nord Canadien. On synthétise le point de vue de transporteurs opérant des navires dans le Nord Canadien, en terme de besoins cartographiques à des fins de sécurité de la navigation. Enfin, on compare deux approches de priorisation de campagnes de levés hydrographique : l’une a été développée par le Service Hydrographique du Canada, et l’autre est une adaptation du schéma développé par le Land Information New Zealand aux spécificités du Nord Canadien. Summary: This reports aims to explain the dynamics of North Canada marine charting development, in relation with the availability of technological resources, from the early 20th century. We detail the development of maritime corridors, routes and the associated requirements in terms of marine charts, taking into account the specificity of arctic waters navigation constraints. We explain the weakness of historical hydrographic data quality and related marine charting products in terms of lack of positioning infrastructures and accurate knowledge of both geodesy and water levels, which are essential pre-requisites for meeting modern hydrographic and cartographic quality standards. We summarize the different interviews of Northern Canada transport companies, which describe their needs in terms of availability of nautical charting products. Then, we compare two different approaches for the prioritization of hydrographic surveys. The first one has been proposed by the Canadian Hydrographic Service, and the other one is an adaptation of the Land Information New Zealand prioritization scheme to Canada’s North areas. -3- TABLE OF CONTENTS TABLE OF CONTENTS -4- TABLE OF FIGURES -6- GENERAL INFORMATION -8- 1. INTRODUCTION -9- 2. CONTEXT -9- 2.1 EXISTING AND POTENTIAL SHIPPING ROUTES - 10 - 2.2 OVERALL STATE OF CARTOGRAPHY - 14 - 3. HISTORICAL DEVELOPMENT OF HYDROGRAPHY IN NORTHERN CANADA - 16 - 3.1 PRECONDITIONS OF QUALITY HYDROGRAPHY GEOREFERENCING KNOWLEDGE OF TIDal Fluctuation SUMMARY - 16 - 16 - 21 - 22 - 3.2 - 22 - BATHYMETRIC METHODS AND EQUIPMENT 3.3 DEPLOYMENT OF HYDROGRAPHIC EQUIPMENT POSSIBLE CONVENTIONAL SOLUTIONS UNCONVENTIONAL SOLUTIONS CROWDSOURCED BATHYMETRY AUTONOMOUS SYSTEMS METHODOLOGICal ASPECTS HYDROGRAPHIC AND CARTOGRAPHic DATA - 23 - 23 - 24 - 24 - 27 - 29 - 30 - 4. - 39 - 43 - 5. PROBLEMS RELATED TO THE LACK OF CARTOGRAPHY IN NORTHERN CANADA SUMMARY PRIORITIZATION METHODS SUITABLE FOR NORTHERN CANADA 5.1 PRIORITIZATION AS defined BY THE CANADIAN HYDROGRAPHIC SERVICE ANALYSIS OF THIS METHOD -4- - 43 - 43 - 46 - 5.2 “PARTICIPATORY” prioritization PRIORITIZATION CONCEPTS STEP 1a: DATA COMPILATION STEP 1b: HAZARD IDENTIFICATION STEP 2: RISK ASSESSMENT STEP 3: ECONOMIC ANALYSIS STEP 4: HYDROGRAPHIC TECHNICAL VISIT STEP 5: DEFINITION OF PRIORITY HYDROGRAPHIC SURVEYS SUMMARY Conclusion - 48 - 48 - 49 - 51 - 52 - 53 - 54 - 54 - 55 - 56 - REFERENCES - 57 - APPENDIX 1: EXAMPLE OF IDENTIFYING THE CONSEQUENCES OF A MARINE ACCIDENT INVOLVING A SOLAS PASSENGER VESSEL - 59 - APPENDIX 2: EXAMPLE OF IDENTIFYING THE CONSEQUENCES OF A MARINE ACCIDENT INVOLVING A CARGO SOLAS VESSEL - 60 - -5- TABLE OF FIGURES Figure 1: Canadian Arctic: The red dots represent the locations of communities and the purple lines represent marine traffic density (2010 data). Map available on the geographic portal http://geoportal.gc.ca (Arctic Voyage planning guide) ......................................................... - 9 Figure 2: Shipping routes according to AIS data in Northern Canada (Source ArkGIS)........... - 10 Figure 3: Leyzack, Economic Benefits of Hydrography in the Canadian ArcticA case Study, Lighthouse Ed 77, 2011 .............................................................................. - 11 Figure 4: Potential projects for deep water harbours, in relation to mining projects (source: St. Lawrence Shipoperators, 2011 CIDCO Symposium) ........................................................ - 11 Figure 5: Map of the major mineral projects in the Canadian Arctic in March 2012. Source: Aboriginal Affairs and Northern Development Canada. ..................................................... - 12 Figure 6: List and map of mining projects in development. Source: St. Lawrence Shipoperators, 2011 CIDCO Symposium ................................................................................................... - 13 Figure 7: Points of gravity measurements used for the CGDV28. We can see that Northern Canada is relatively uncovered and, therefore, not modeled............................................. - 17 Figure 8: Distant Early Warning in the Arctic. We find that only very partial coverage can be achieved when using this system, for given the linear configuration of the network, you have to be situated between two radar beacons in order to obtain an accurate position. .......... - 19 Figure 9: Principle of horizontal positioning using two radar beacons. ..................................... - 19 Figure 10: Permanent GNSS network (CACS). Yellow indicates the regional network stations, green indicates the national network stations. It is clear that the network density is too low to allow for positioning via quality PPP. ................................................................................. - 20 Figure 11: Permanent tide gauges, paired with a permanent GNSS station operated by the CHS. Note that only 4 stations were installed. ............................................................................. - 22 Figure 12: Several hydrographic vessels, having different consistent measurement uncertainties. The blue ship has a positioning uncertainty greater than the vertical uncertainty; the reverse is true for the green ship. The red ship is the least effective. The surveys are consistent if every datum (centre of the cross) belongs to the intersection of uncertainty domains of every other sounding. In classical statistical theory of measurement error, the crosses correspond to the small and large axes of an ellipse (in 2D) or an uncertainty ellipsoid (in 3D). ......... - 25 Figure 13: Critical situation where the measurement uncertainties are inconsistent between different hydrographic surveys conducted by different vessels. The blue and green vessels are inconsistent, for their most probable measurements (centre of the crosses) are located beyond their respective uncertainty ellipses. ..................................................................... - 26 Figure 14: In crowdsourced bathymetry, no uncertainty can be assessed, for the set-ups are not rigorously studied and integrated. Therefore, it is impossible to define a notion of consistency -6- between hydrographic measurements. In the case of measurements repeated by a large number of small crafts, it is possible to define a statistic in relation to a purely descriptive statistic of the seafloor, but this would most certainly be impossible for scattered and infrequent measurements, as would be the case in Northern Canada. ............................. - 26 Figure 15: Drifter from the Seafloor Sounding for Polar and Remote Regions (SSPARR) project .. 28 Figure 16: Approach maps; above, electronic charts, below, paper charts. The locations of communities are indicated in red. The ENCs are indicated by the blue rectangles, the paper charts by the pink rectangles.............................................................................................. - 32 Figure 17: Coastal charts; above, electronic charts, below, paper charts. The locations of communities are indicated in red. The ENCs are indicated by the blue rectangles, the paper charts by the pink rectangles.............................................................................................. - 33 Figure 18: General maps; above, electronic charts, below, paper charts. The locations of communities are indicated in red. The ENCs are indicated by the blue rectangles, the paper charts by the pink rectangles.............................................................................................. - 34 Figure 19: ENC ports: Northern coast of Baffin Island and the Dease Strait. ........................... - 35 Figure 20: ENC ports: Vicinity of Prince Charles Island. ........................................................... - 36 Figure 21: Paper port maps: Only Churchill and a few ports on the northern coast of Baffin Island are indicated. ...................................................................................................................... - 36 Figure 22: Every ENC in Northern Canada ............................................................................... - 37 Figure 23: Areas for which the CHS possesses modern hydrographic data, Source Christopher Wright, Navigability of the Canadian Arctic, CHS 2012, Niagara Falls .............................. - 37 Figure 24: Correlation between the DEW line (black line on the map above) and coastal paper charts (pink rectangles on the map below). ....................................................................... - 38 - -7- GENERAL INFORM ATION CLIENT Company/ministry Transport Canada Person responsible Contact information Frédéric Sirois Tel.: (613) 949-8793 Email: [email protected] CIDCO Project leader Jean-Guy Nistad Information compiled by Mathieu Rondeau Sylvain Gautier Jean Laflamme Report drafted by Nicolas Seube DELIVERABLES - 1 electronic version of the final report (in Word format) for the client -8- 1. INTRODUCTION Acquiring hydrographic data poses difficult problems in Arctic regions. In fact, the possibility of compiling hydrographic data of cartographic quality does not depend exclusively on deploying hydrographic ships equipped with sophisticated sounders, but largely on having a) prior knowledge in the field of geodesy, b) precise positioning equipment requiring a land-based infrastructure, c) excellent knowledge of the coastline, d) measurement equipment, and e) tidal models. In the current state of Arctic cartography, we demonstrate why these methods are cost-prohibitive and not well suited in Northern Canada, not only due to the extent of the region, but also to the presence of complex geophysical phenomena. First, we discuss the various methodologies employed in Northern Canada, providing a brief historical perspective in order to understand why charting and mapping the North poses numerous problems. Then, we provide an overview of the hydrographic methods and equipment that have been used in Northern Canada since the 1930s, and discuss the current status of cartographic products for mariners. Finally, in order to coordinate the cartographic development of Northern Canada with its economic development potential, we present a prioritization method used by the Canadian Hydrographic Service (CHS). We analyze this method, before proposing a method we believe is more suitable for the North. 2. CONTEXT Figure 1: Canadian Arctic: The red dots represent the locations of communities and the purple lines represent marine traffic density (2010 data). Map available on the geographic portal http://geoportal.gc.ca (Arctic Voyage planning guide) -9- Figure 2: Shipping routes according to AIS data in Northern Canada (Source ArcGIS) 2.1 EXISTING AND POTENTIAL SHIPPING ROUTES According to two separate sources, the figures above represent the most frequently used navigation routes in the Arctic. It is relatively clear that the main routes link the city of Churchill (Manitoba), the only port in the Arctic connected to the Canadian rail network, and otherwise include the Northwest Passage. The chart in Figure 1 seems to underestimate the traffic in the Foxe Basin. It is also apparent that the Northwest Passage is now increasingly used, and its economic potential (less than 4,000 km for a Europe-Asia crossing compared to the Panama Canal) is very significant. As an approach route to the Northwest Passage, the eastern coast of Baffin Island also experiences relatively significant marine traffic. It should be noted that the Port of Churchill, the only deep water port in this region, can receive Panamax vessels (vessels with draughts of up to 12 meters). An oil terminal project in Churchill is currently under study and the security of the navigation route for eventual oil exports is certainly an important question. The development of mining projects also serves as motivation for opening or securing commercial shipping routes. Finally, tourism development, mentioned in Publication C-55 of the International Hydrographic Organization (IHO), is a factor in marine traffic development presenting a significant risk. 100 cruise ships visit the Canadian Arctic each year. 69 boats took the Northwest Passage between 1906 and 2009. In 2010, 24 ships passed through in seven months, a high proportion of which were cruise ships. - 10 - Figure 3: Leyzack (2011), Economic Benefits of Hydrography in the Canadian Arctic A case Study, Lighthouse Ed 77, 2011 Traffic density in the Northwest Passage has increased from 7 vessels in 2007 to 16 in 2010, 13 of which were cruise ships. Currently, no cargo vessel takes the Northwest Passage. The traffic in Northern Canada has increased from 119 commercial vessels in 2005 to 220 in 2010, which can be partly explained by the population increase in the North (100,000 people today), with a high growth rate (+20 % per year). The number of fishing vessels increased from 30 in 2005 to 220 in 2010. Figure 4: Potential projects for deep water harbours, in relation to mining projects (source: St. Lawrence Shipoperators, 2011 CIDCO Symposium) - 11 - Figure 5: Map of the major mineral projects in the Canadian Arctic in March 2012. Source: Aboriginal Affairs and Northern Development Canada - 12 - Figure 6: List and map of mining projects in development. Source: St. Lawrence Shipoperators, 2011 CIDCO Symposium - 13 - It seems relatively clear that marine traffic in Northern Canada is related to: Procurement activities (supplying communities that have a growing population). This traffic uses a network of secondary shipping routes and is currently the most extensive; The very rapid development of mining projects in Northern Canada, which is creating a great need for deep water ports, particularly along the eastern coast of Baffin Island and the western Hudson Bay; Commercial traffic from terminal ports (such as Churchill, MB); Northwest Passage routes representing high international traffic (avoiding the Panama Canal), the dynamic of which could be very strong in the medium term. 2.2 OVERALL STATE OF CARTOGRAPHY In September 2013, according to IHO Publication C-55, the aim of which is to provide an estimation of areas covered by modern hydrographic surveys, Canada supplied the following information concerning international region A: Regarding the status of hydrographic surveys (hydrographic data serving as a basis for cartographic production): Status of hydrographic surveys Percentage A1 30 Adequately surveyed 0-200m A2 15 Adequately surveyed >200m B1 10 Requiring re-survey 0-200m B2 10 Requiring re-survey >200m C1 30 Never been systematically surveyed 0-200m C2 25 Never been systematically surveyed >200m Areas - 14 - Comments: 1. “The high proportion of inadequately surveyed waters is predominantly due to the large areas of Arctic waters that are unsurveyed or covered by frontier surveys only.” 2. “Ecotourism, climate change and resource development are increasing demand for surveys in Arctic and frontier areas.” Concerning the status of nautical charts available to mariners: Status of marine cartography Percentage small scale, offshore Percentage medium scale, coastal passage Percentage large scale, ports, approaches A 100 75 75 Paper (INT) B 100 75 100 Digitized from paper charts (RNC) C 100 100 100 Electronic charts (ENC) Areas covered by charts in format Comment: “Arctic and other frontier areas are not covered by electronic chart products due to lack of demand.” A few comments concerning these data: 1. The percentage of INT charts is inferior or equal to the percentage of ENCs, which is most surprising. In fact, international regulation requires the presence of paper charts onboard vessels. 2. The percentage of available charts is very high relative to the quality of hydrographic data. For example, concerning the coastal areas (0-200m), the A1/B1/C1 percentages are 30%, 10% and 30% respectively. This means that 30% of coastal areas are considered adequately surveyed, 10% of these require new hydrographic campaigns, and 30% have never been surveyed. 3. Therefore, we can conclude that a large percentage of data assimilated into nautical charts is of poor quality, and that charts contain areas devoid of any hydrographic data. - 15 - Concerning region N (Arctic Regional Commission), the member countries of which are Canada, Denmark, Norway, Russia and the United States, Canada provides no aggregate data. However, the 2012 report from Canada indicates that 18 ENCs, 1 INT chart and one RNC chart were produced in the Arctic region. In August, 2011, the CHS conducted a campaign in the Victoria Strait, in Nunavut. 520 km² were surveyed using modern methods and equipment. Therefore, it seems clear that the Canadian Arctic has been subject to inadequate hydrographic and cartographic coverage in view of the economic growth in the North, the opening of the Northwest Passage, and the relatively high increase in tourism in the Arctic regions and territories. According to the Canadian Hydrographic Service (CHS), 10% of waters in Northern Canada and 40% of waters in the Northwest Passage have been accurately surveyed. The following sections detail the history of hydrography and the status of marine cartography in the Canadian Arctic. 3. HISTORICAL CANAD A DEVELOPMENT OF HYDROGRAPHY IN NORTHERN 3.1 PRECONDITIONS OF QUALITY HYDROGRAPHY As we mentioned in the introduction, a hydrographic survey does not consist solely of submerging a water depth measurement device in water to obtain a sounding. A significant infrastructure is required to georeference each sounding so as to allow for consistent cartographic production. Georeferencing is done by referencing the so-called “horizontal” and “vertical” positioning of the sounding. In other words, it has to be referenced in geodetic coordinates (longitude, latitude) and height. In order to understand cartographic development in the Arctic, we must return to these basic notions and examine how various historical campaigns, using limited technology, produced certain approximations leading to the current situation. Needless to say, we do not provide an exhaustive description of each campaign (which would be pointless); rather, we examine the fundamental reasons as to why the production of hydrographic data in the Arctic is difficult and limited. It is equally important to mention that some fundamental data, such as coastline data, were incorporated into charts relatively early and based on old data. This is not purely hydrographic data, but geographic, which is quite difficult to update. Hence, given the complexity and effort involved in updating nautical charts, poor-quality historical data are ever present and constitute significant error sources. Therefore, understanding the origin of the data is fundamental to better recognizing their impact on the nautical documentation available today. GEOREFERENCING - 16 - Geodetic systems: Before the advent of satellite positioning, local geodetic models were approximated by reference ellipsoids, which had the advantage of approximating the geoid surface. Therefore, differences between normal (according to local gravity) and ellipsoidal (according to the ellipsoidal normal chosen) altitudes were somewhat negligible. Historically, geographical and geodetic data measured in the Canadian Arctic were referenced to different datums. For planimetric data, the NAD27 was used. It relies on the Clarke ellipsoid of 1866 which adapts particularly well to North America. For altimetric data, the CGVD28 system presented a 60 cm error relative to the mean level of seas from the East to the West of Canada. The NAD27 datum underwent many updates and is now known under the name NAD83. Numerous nautical charts reference this datum (Craymer, 2006). The Northern Horizontal Network (in 2D) is a network referenced to the NAD83 system which includes a benchmark every 20 or 100 km, but has a positioning uncertainty of 1 m (at minimum). Today, two geoid models prevail: the CGG 2010 and the EGM08. In certain areas, such as the eastern coastline of Baffin Island, they present differences of 0.8 m. Therefore, it remains relatively tricky to accurately georeference hydrographic data and connecting marine and land data continues to be problematic. Figure 7: Points of gravity measurements used for the CGDV28. We can see that Northern Canada is relatively uncovered and, therefore, not modeled. Chart datum: Every hydrographic (or geographic) measurement is relative and the origin of altitudes and planimetric coordinates must be specified. A change of geodetic systems can produce planimetric variations, since the reference ellipsoids do not necessarily have the same - 17 - center. Therefore, when using historical data, this must be taken into account in order to accurately locate data in cartographic projection. Nevertheless, the planimetric variations are simple to manage if we use metadata related to historical surveys, which allow us to perform accurate transformations. Altimetric references are related to a geoid (or rather a geoid model). In Northern Canada, in addition to the lack of dense gravity data, the relatively rapid nature of the uplifting of the Earth’s crust would require significant geodetic campaigns so as to refine a reliable geoid model. Hydrographic datum is more problematic, for it requires an intimate and historical knowledge of the tide. Hydrographic datum is defined at the national level: it is the Lower Low Water Large Tide (LLWLT). Today, some countries invest in models “separating” (separation models) chart datum and an ellipsoid reference (generally ITRF). These models relieve hydrographers from relating data to geodetic references, which can be complex and repetitive, thus costly. Therefore, a simple GNSS ellipsoid height measurement (if the former is presumably of minimal quality) allows the user to relate each sounding to a physical geoid model or nautical chart datum. Unfortunately, no such model exists in the Arctic, and the complexity and vastness of the territory would require a substantial investment in such models. We will see the consequences that the lack of separation models has on the practice of hydrography in Northern Canada. Positioning: The positioning equipment used in the Arctic was, successively, astronomical positioning equipment (hydrographic circle), radio-electric or electromagnetic (RADAR) positioning systems, such as the Distance Early Warning (DEW) system used in the Northern Arctic to detect Soviet aerial intrusions during the Cold War era. This type of positioning allowed a number of bathymetric measurement campaigns to be conducted, notably those by the USS Tanner and USS Baffin. The system was decommissioned in 1963. - 18 - Figure 8: Distant Early Warning in the Arctic. We find that only very partial coverage can be achieved when using this system. Given the linear configuration of the network, it is necessary to be situated between two radar beacons in order to achieve an accurate positioning. Between the two world wars and up until 1957, radio navigation was used vessels and aerial reconnaissance planes flying aerial photography missions charting the coastline. The positioning methods and equipment, other than traditional aerial navigation aids of the time, used radar beacons that allowed between two beacons (similar to the principle of the DEW). Figure 9: Principle of horizontal positioning using two radar beacons. - 19 - by hydrographic with the goal of gyroscopes and for a positioning This system, used for the aerial reconnaissance of Canadian coasts with the goal of charting the coastline, is the source of numerous positioning errors. In fact, because airplanes in the 1930s were poorly positioned when flying these missions, the aerial photographs are poorly georeferenced and as a result, the coastline can suffer from georeferencing errors of 3 or 4 nautical miles. The result is that many nautical charts are still drawn with an inaccurately positioned coastline. This is also true for numerous islands in the North, which are very poorly georeferenced on nautical charts. Currently, satellite positioning systems (GNSS) are widely used and referenced to the geodetic system defined by the ITRS. However, the level of precision depends on the infrastructure available on land so that real-time differential or real-time kinematic solutions, or even PPP (Precise Point Positioning) post-processing solutions, can be used. Real-time GNSS positioning requires a land base, which must be positioned relative to a known geodetic point or connected to the geodetic network. Post-processing solutions require the use of permanent GNSS stations that are part of a data dissemination infrastructure and allow for a post-processed correction by merging data from the hydrographic instrumentation and the station data. Figure 10 clearly shows that it is very difficult to achieve precise GNSS positioning in the Arctic given the lack of a permanent GNSS infrastructure. Indeed, there would need to be a permanent station approximately every 100 km to allow for correction of raw GNSS data, allowing for an absolute-centimeter positioning. Figure 10: Permanent GNSS network (CACS). Yellow indicates the regional network stations; green indicates the national network stations. It is clear that the network density is too low to allow for a positioning via quality PPP. - 20 - However, positioning solutions requiring fewer infrastructures can be implemented by connecting GNSS land based stations to the geodetic network and with knowledge of vertical references (nautical chart datum, therefore knowledge of the tide). The current methodology consists of installing a differential GNSS base station connected to the geodetic network and using an RTK (Real Time Kinematic) positioning technique to retrieve data referenced precisely to the ellipsoid. However, given the rapid advancement of PPP positioning, one would think this positioning technique, heretofore reserved for offshore applications (beyond the range of land-based stations) would allow for a sub-decimeter positioning according to the vertical component anywhere in the world. In September 2013, Fugro announced a vertical absolute positioning in ellipsoidal height precision of less than 8 cm. KNOW LEDGE OF TIDAL FLUCTUATION The Canadian Arctic was the subject of study in the field of large-scale harmonic tide modeling. On the regional scale, certain tide prediction models achieve performances of 20 cm for the M2 component and 5 cm for the N2, S2, K1 and O1 components. However, certain regions, like the south-central and southeast regions, Frobisher Bay and Boothia Bay, cannot be accurately modeled. Furthermore, these models are sensitive to the frictional drag of the tide wave with the ice and, therefore, their performance varies according to the season. The main issue concerning the tide in the Arctic is the overall nature of the studies that have been dedicated to it. It is perfectly clear that, given the complexity of the Canadian Arctic archipelago, the friction phenomena caused by coasts and estuaries play an important role. Therefore, it is rather hazardous to rely on a harmonic tide model to reduce hydrographic soundings to nautical chart datum. Furthermore, the lack of a dense network of tide gauges poses an infrastructure problem, for it is impossible to reduce a set of soundings relative to measurements that are too remote, or to a model that seems very approximate in certain regions. This poses another hydrographic methodology issue, for soundings achieved by GNSS vertical positioning are to be excluded (due to the lack of a separation model, as previously discussed). Consequently, there must be a systematic use of old tidal reduction methods, taking into account the settling of the vessel in relation to the water level, heave, and non-tidal water level variations, which are uncertainties that add to the total uncertainty of soundings. - 21 - Figure 11: Permanent tide gauges, paired with a permanent GNSS station operated by the CHS. Note that only 4 stations are installed. SUMMARY To summarize this section, we can conclude that hydrographic surveys in the Canadian Arctic require an effective positioning infrastructure and some basic geodetic knowledge in order for data acquired by a hydrographic platform to be adequately georeferenced. Additionally, knowledge of tidal fluctuation is essential to conducting hydrographic surveys with a view to navigation safety in coastal areas. From these viewpoints, Northern Canada does not currently have such an infrastructure; therefore, it should be a priority to implement one before conducting systematic or opportunistic surveys. 3.2 BATHYMETRIC METHODS AND EQUIPMENT At the beginning of the century, soundings were carried out using lead lines, in accordance with well-known historical sounding line and optical or astronomical positioning methods. These unproductive methods were replaced by the use of echosounders in the 1940s and up until 1990 the majority of missions were undertaken using single-beam echosounders. Therefore, the cartographic data do not represent an exhaustive coverage of the seafloor like the multi-beam sounders of today do. Nevertheless, they allow for the compilation of useful and reliable data (within the limits of positioning errors, as we previously discussed). Beginning in the 1960s, under the Polar Continental Shelf Program, an echosounder capable of operating through the ice sheet was developed. Such equipment allowed scientists to avoid cutting sounding holes and not be limited to survey in ice-free waters. From the 1930s to the 1970s, a large quantity of data acquired by the CHS came from KelvinHughes sounders, for which we provide the precision characteristics according to Hare, 1997 (see below): - 22 - The bathymetric surveys conducted today combine data from single-beam and multi-beam sounders. They can be deployed either from large vessels or from small surveyor launches carried by a larger hydrographic vessel. Vessels navigating in the Arctic need first and foremost to be adapted to constraints due to the immensity of the territory and particular navigation conditions. Vessel adaptation for mounting of acoustic sounders is of secondary importance. 3.3 DEPLOYMENT OF HYDROGRAPHIC EQUIPMENT Every hydrographic platform must, at minimum, integrate positioning equipment (GNSS), orientation measurement (inertial sensor), depth measurement (echosounder) and acquire and process this data in quasi real-time. Below, we provide a list of possible solutions and their applicability in Northern Canada. POSSIBLE CONVENTIONA L SOLUTIONS Today, we distinguish between manned platforms (vessels, boats) and autonomous platforms (AUVs, drifters). As for airborne methods, LiDAR equipment can be implemented, but we must consider the fact that it is limited by relatively shallow water penetration (10-30 metres maximum). However, it is perfectly suited to coastal areas and covers large spaces very quickly when sea conditions are good. In rough seas, the precision of the water surface measurement by optical means decreased. Manned vessels can support the integration of multi-beam sounders, even when small in size, given the current miniaturization of sounders. Multi-beam sounders installation and operation is well understood, even if it requires a certain expertise in hydrography. Autonomous equipment (AUV) is still in the exploratory stage in hydrography, the problem being (yet again) the long-term underwater positioning of these machines. Either they require long baseline acoustic positioning systems, which are very expensive to install and that limit the range of operability of AUVs, or their integrated navigation system drifts and does not allow for georeferencing data to be acquired with a precision that is in accordance with current hydrographic standards. Drifters can be an interesting alternative, but their main inconvenience is that it is impossible to spatially plan a hydrographic survey using such tools, because it is obviously impossible to control their trajectory. Furthermore, soundings measured acoustically cannot be corrected using a water column sound speed measurement, as is necessary for any echosounding device. - 23 - We provide a more detailed comparison of various types of systems and their integration. We also discuss crowdsourced bathymetry, which involves using soundings of opportunity surveyed by mariners (not hydrographers). We will see how this solution is difficult to apply in Northern Canada. UNCONVENTIONAL SOLUT IONS CROWDSOURCED BATHYMETRY In theory, crowdsourced bathymetry is an alluring concept due to its simplicity and extremely low cost, but it raises a number of issues. Before analyzing its relevance to potential applications in the Arctic, let us briefly review its principle. In general, for navigation safety purposes, every vessel is equipped with acoustic sounding instrumentation. The idea of crowdsourced bathymetry is to allow mariners to disseminate their own water-depth data to organizations whose role is to assimilate this data in order to produce hydrographic information that could aid in cartographic production. Several projects exist all over the world: Opensea map (www.openSeaMap.org) Teamsurv (www.teamsurv.eu) ActiveCaptain (www.activecaptain.com) Navionics (www.navionics.com) GoogleOcean (http://earth.google.com/ocean) IceWatch (www.naturewatch.ca) Olex (www.olex.no) SURVICE Engineering (argus.survice.com) First, we should mention that contrary to common belief, crowdsourced bathymetry data are not free. Certain Internet communities (or professional communities) reserve usage (i.e. confidentiality) rights to these data. The most advanced of these projects is the Teamsurv project (European), whose objective is to chart coastal areas. A comparison was made between the data published by TeamSurv and the data from the UKHO in order to determine the reliability of crowdsourced bathymetry data. Two pilot regions were chosen and reference data were tracked along isobaths (contour lines). The isobaths from soundings provided by Teamsurv allow for easy detection of inconsistencies, by comparing sounding values to adjacent contour lines. The conclusion of this study shows that, for relatively high-traffic areas (the English Channel), notable gaps appear on one of the two sets of data. These errors are not isolated and reach 30 metres. Another issue raised by crowdsourced bathymetry is the statistical nature of hydrographic data. All hydrographic data taken separately contains an error relative to the true value. This error is modeled statistically as a total propagated uncertainty, resulting from the propagation of every error source along the measurement system employed. Therefore, an isolated piece of - 24 - hydrographic data (such as that supplied by a single-beam sounding along the sounding line) has no meaning unless the uncertainty measurement is known. Moreover, it is very rare for two hydrographic measurement systems (i.e. two hydrographic vessels) to produce systematically consistent results on the same seafloor. This can be explained by the complexity of hydrographic measurement systems, the complexity of their calibration, the precise knowledge of exact positions of sensors and the potential distortions of mounted mechanical equipment on which sounders and various antennas are fixed. However, two hydrographic vessels producing quality data will produce consistent measurement uncertainties with the bias observed between the two datasets. And that is what is important, as figure 12 illustrates. In comparison, inconsistent datasets are produced in the situation depicted in figure 13. Figure 12: Several hydrographic vessels, having different, yet consistent, measurement uncertainties. The blue ship has a positioning uncertainty greater than its vertical uncertainty; the reverse is true for the green ship. The red ship has the greatest horizontal and vertical unicerity. The surveys are consistent if every sounding value (centre of the cross) belongs to the intersection of uncertainty domains of every other sounding. In classical statistical theory of measurement error, the crosses correspond to the small and large axes of an ellipse (in 2D) or an uncertainty ellipsoid (in 3D). - 25 - Figure 13: Critical situation where the measurement uncertainties are inconsistent between different hydrographic surveys conducted by different vessels. The blue and green vessels are inconsistent, for their most probable measurements (centre of the crosses) are located beyond their respective uncertainty ellipses. In crowdsourced bathymetry, it is impossible to control the uncertainty of multiple measurement systems onboard ships of opportunity and it is unlikely that mariners calibrate and correct their acoustic data accurately, and reliably reduce their soundings to a common vertical reference (in short, executing the complex work of a hydrographer). Figure 14: In crowdsourced bathymetry, no uncertainty can be assessed, for the set-ups are not rigorously studied and integrated. Therefore, it is impossible to define a notion of consistency between hydrographic measurements. In the case of measurements repeated by a large number of small crafts, it is possible to calculate purely descriptive statistics of the seafloor, but this would most certainly be impossible for scattered and infrequent measurements, as would be the case in Northern Canada. Basically, the principle of crowdsourced bathymetry is to conduct a statistical estimation of a seafloor from a multitude of soundings measured by different measurement systems, thus submitted to different metrological biases. Needless to say, if several measurements are taken at the same location, and if we consider that measurement system biases have zero-mean (which cannot be justified), the sounding will be accurately estimated. However, if a sounding is determined by an insufficient number of measurements, and it is evident that measurement system biases have a very low probability of being zero. Thus, the sounding will be error ridden. - 26 - It should be noted that the above mentioned errors of 30 m concern seafloors measuring 30 m in depth in relation to nautical chart datum in the Teamsurv-UKHO comparative example. Therefore, the sounding error is 100%! It is thus not possible to apply this technique to the Arctic, for it goes without saying that marine traffic, including intercoastal vessels operated by communities (e.g. fishermen), does not produce sufficient statistics for preventing measurement biases. Therefore, data collected via these means in Northern Canada would evidently be very dangerous. Given the navigation hazard, marine traffic remains confined to narrow waterways, making crowdsourced bathymetry irrelevant, as this technique would only be productive if a large number of vessels were to follow routes covering vast territories, which will not be the case in the medium-term. Therefore, we can conclude that crowdsourced bathymetry is not a suitable solution for cartography in the Arctic, since it does not guarantee the absence of significant sounding biases in the case of low sample measurement (from a statistical point of view). AUTONOMOUS SYSTEMS Autonomous systems may seem like viable alternative solutions, allowing for large spaces to be covered without any human intervention. Below, we discuss the case of AUV (Autonomous Underwater Vehicles) and drifters in greater detail. As we previously mentioned, every water depth measurement system must be positioned horizontally and vertically in order to accurately georeference the sounding data after reducing the tidal fluctuation. AUVs do not escape this universal rule; therefore, conducting an AUV-based hydrographic survey requires the machine to be well-positioned underwater. However, there is a possible alternative: 1. Use an acoustic network of Long Base Lines (LBL) mounted on the seafloor, whose position is determined once for every surface vessel. 2. Position the AUV on an Ultra-Short Base Line (USBL), carried by a surface vessel following the AUV. In both cases, we see that it is the positioning infrastructure that creates the difficulty because it either requires the range of operation to be reduced to the envelope of long base line positions (which are very expensive to install), or the presence of a surface vessel positioning the AUV. In fact, AUV-based solutions are only being developed in the oil and gas industry, for conducting ultra-high-resolution surveying of areas that are limited in space and very deep (a sounding from the surface would be too low-resolution). This type of application has nothing in common with the task of providing extensive coverage of large areas in Northern Canada. The second tempting alternative is to use drifters. This has been done for about ten years in the Arctic. It is possible and relatively inexpensive to equip a buoy with an echosounder, positioning equipment and attitude compensation. We can cite the example of the SSPARR Project commissioned by the National Science Foundation (Rognstad et al., 2005); (Hall, 2006). Similarly, CIDCO developed a prototype for a hydrographic buoy fitted with an inclinometer, a - 27 - single-beam sounder and an L1/L2 GNSS system. For now, this buoy is reserved for shallow waters (< 80 m) and specifically designed for river hydrography. Figure 15: Drifter from the Seafloor Sounding for Polar and Remote Regions (SSPARR) project In the SSPARR Project, tests were performed on the Yermark Plateau in northern Spitzberg in 2010 and 2011. The data was used to improve the bathymetric grid produced by the IBCAO (International Bathymetric Chart of the Arctic Ocean) for this region. This autonomous buoy has a 5-year battery life and an Iridium modem allowing for data transfer via satellite. Another example is the SOFAR drifters, deployed under the European project DAMOCLES, fitted with low frequency acoustic emitters for the purpose of positioning a network of profiling floats. The use of such drifters poses problems on two fronts: survey planning (drifters are at the mercy of currents, winds and ice drifts); and acoustic measurement calibration, which relies heavily on the average water sound speed. Concerning the survey planning problem, without the use of an oceanographic circulation model, it is impossible to predict the trajectories of buoys and the spacing between the sounding lines of each buoy. With uncontrolled trajectories, the data can only be scattered; therefore, it will not meet international hydrographic standards in terms of sounding line spacing according to map scales. The second major problem is the futility of correcting data related to environmental conditions, particularly regarding the average water sound speed. Because oceanographic data and ocean circulation models in Northern Canada are as rare as hydrographic data, they do not serve as a reliable basis for predicting water stratification. Moreover, some straits experience major changes in current patterns, particularly in the sub-Arctic regions. Therefore, it would be impossible to - 28 - correct acoustic returns of the average water column sound speed value, rendering the hydrographic data relatively unreliable, even for medium depths (a 1% error in average sound speed produces an equivalent sounding error). We should also mention that automatic bottom detection algorithms (without any human intervention) require the implementation of algorithms allowing for distinction between multiple reflections caused by ice and icebergs. Moreover, thanks to swath bathymetry sondeurs, it is possible to cover a relatively significant seafloor swath (45 degrees, i.e. 0.8 times the water height). However, in shallow water, this benefit is somewhat lost. Finally, the deployment and recovery of these machines requires the use of aerial equipment, which would be relatively expensive to implement, given the extensive areal coverage in which a network of hydrographic buoys would be located. METHODOLOGICAL ASPECTS The lack of positioning and tide gauge infrastructure in the North, as well as the geoid modeling uncertainties, cause problems that go well beyond choosing a hydrographic platform. Indeed, the hydrographic methodology is greatly restricted. As there are practically no permanent tidal and GNSS measurement stations in the whole of Northern Canada, tide gauges must be installed before every hydrographic campaign; otherwise, soundings would have to be reduced to a global harmonic model that, as we have seen, is relatively inaccurate. Also, GNSS bases stations that enable accurate positioning for coastal surveys (0-200 m), allowing for a horizontal accuracy of soundings for depths less than 2 m, need to be installed prior to commencing a survey. This is not a problem for offshore surveys (>200 m, whereby a tidal reduction is not necessary for ensuring safe navigation). From this observation, it is relatively clear that the deployment of autonomous AUVs and drifters is a solution restricted by its range of operation, since it depends on accurate tidal and GNSS information. Therefore, these are very poor solutions in terms of deployment, for the autonomy of these machines is merely relative. Extending their autonomy significantly increases the complexity of their use while diminishing their reliability in extreme conditions (presence of ice). Mobilizing systems with high coverage capabilities, like multi-beam sounders on ship-based platforms (whether of-opportunity or not), remains the soundest solution. From a methodological 1 perspective, sounding reduction should be systematically done from tide gauges installed specifically for each hydrographic mission. Therefore, hydrographic surveying of Northern Canada must rely on a conventional methodology in order to be effective. Any use of autonomous systems (which is tempting, given the extreme conditions of operations), would incur significant costs for data of an inferior quality. 1 Preferably submerged and equipped with acoustic release systems allowed them to surface after use. - 29 - HYDROGRAPHIC AND CARTOGRAPHIC DATA It seems relatively clear that vessel traffic in Northern Canada is increasing, even if the intensity of the said increase is currently very limited. Moreover, it would be unrealistic to set a short-term goal of producing cartography of quality comparable to that of the large shipping routes in the South. Even the concept of quality cartography (or accurate hydrography) must be adapted to the unique Northern context. It should also be noted that the relationships between hydrography (in the sense of hydrographic surveying, i.e. the collection of georeferenced bathymetric data) and cartography are complex. For example, the use of a nautical chart does not ensure that bathymetric data are accurate, or even adequately cover the area delineated by the chart. A substantial number of nautical charts contain erroneous or outdated data, geographic areas without any data, or even data that is too scattered to actually indicate an isolated hazard. Take a map of the Red Sea, for example, where we can read a note indicating the “presence of blooming coral reefs” in the middle of an area where every sounding indicates a depth greater than 600 meters. In this case, the hydrographic th data dates back to the beginning of the 19 century. In the case of Northern Canada, there is not an abundance of old data; rather, it is the lack of data that poses a problem. Furthermore, the data acquired and integrated into nautical charts of the North can be considered as having been georeferenced rather inaccurately. In planimetry: We already mentioned the positioning problems that continue to pose a number of difficulties in collecting quality data. In altimetry: Tidal fluctuations are only modeled very generally; geoid models do not coincide. In order to distinguish between a sounding indicated on a nautical chart and its quality, the IHO introduced the concept of a “zone of confidence,” organized into several categories, called CATZOC. On the current charts, they are indicated by inverse triangles and the number of stars (maximum 6, minimum 0) indicates the zone of confidence category, codified from level A1 (very good) to level U (unassessed). According to resolution 3.1.8 of the IHO S-52 standard concerning the electronic chart display and information systems (ECDIS), it is indicated that: Chart data quality indicator: A bathymetric data quality indicator by zones of confidence (M_QUAL CATZOC) will cover the entire area of depth data or bathymetry for the ENC (although not all data will be assessed initially). The table of "CATZOC" values giving the meaning of each zone of confidence should be readily available to the mariner. - 30 - Thus, it is more important to study the CATZOCs than the presence of nautical charts. First, we use a graphic to present the portfolio of nautical charts available from the Canadian Hydrographic Service. We distinguish between paper charts in INT format and ENCs (electronic charts). - 31 - Figure 16: Approach charts. Electronic charts (above). Paper charts (below). The locations of communities are indicated in red. ENCs are indicated by the blue rectangles, the paper charts by the pink rectangles. - 32 - Figure 17: Coastal charts. Electronic charts (above). Paper charts (below). The locations of communities are indicated in red. ENCs are indicated by the blue rectangles, the paper charts by the pink rectangles. - 33 - Figure 18: General maps. Electronic charts (above). Paper charts (below). The locations of communities are indicated in red. The ENCs are indicated by the blue rectangles, the paper charts by the pink rectangles. - 34 - Figure 19: ENC ports: Northern coast of Baffin Island and Dease Strait. - 35 - Figure 20: ENC ports: Vicinity of Prince Charles Island. Figure 21: Paper port maps: Only Churchill and a few ports on the northern coast of Baffin Island are indicated. - 36 - According to the above graphics, it would seem the modern production of ENCs is more centered on the Northwest passages. There are also very few approach charts and port maps. However, these communities do not have deep water ports. This explains the lack of port maps. The presence of general maps tells us nothing about the presence of hydrographic data, which can be very scattered and come from a wide range of sources. In any case, they are unsuitable for coastal or approach navigation. Let us now compare the complete portfolio of ENCs with a diagram indicating the areas where there is quality “modern” hydrographic data: Figure 22: Every ENC in Northern Canada Figure 23: Areas for which the CHS possesses modern hydrographic data. Source Christopher Wright, Navigability of the Canadian Arctic, CHS 2012, Niagara Falls First, we notice that certain areas are accurately surveyed, like the northern coast of Baffin Island and certain areas in the Northwest passages, for example. We also notice there is no clear correlation between the availability of hydrographic data and the availability of ENCs, for incorporating hydrographic data into cartography is a very long process. Curiously, in the Dease Strait and Coronation Gulf areas, there is a significant density of coastal ENCs available, but no quality modern hydrographic data. Let us examine the portfolio of paper charts (based on historical data) in relation to the above mentioned DEW line. The correlation between the DEW line and a series of coastal charts is remarkable; it is apparent that the cartography of this area has developed thanks to the availability of relatively accurate positioning methods and equipment for the time, implemented for military purposes. Because the Northwest Passage was completely irrelevant during that time period, it was not charted. - 37 - Figure 24: Correlation between the DEW line (black line on the map above) and the paper coastal charts (pink rectangles on the map below). - 38 - 4. PROBLEMS RELATED TO THE LACK OF CARTOGRAPHY IN NORTHERN CANAD A This section provides an overview of navigation issues relating to the lack of quality nautical charts in Northern Canada, and summarizes a few recent accident reports relating to hydrography. Under this study, discussions were held with transportation companies operating vessels in the North. These are the following businesses and organizations that were consulted: St. Lawrence Shipoperators (Martin Fournier) Groupe Desgagnés (Captain Richard Perron) Transport Nanuk Inc. (Georges Tousignant) Below, we summarize the above mentioned discussions and provide a more detailed account of the analysis conducted by Groupe Desgagnés, which was based on an interview with mariners having in-depth knowledge of the Eastern Arctic and the Northwest Passage. Every captain has a guide approximately 600 pages in length, created from CHS paper charts (waterways, moorings, etc.) and updated with the group’s observations, consistent with the Groupe Desgagnés internal safety policy. Currently, no vessel in Canada is equipped to navigate with electronic charts. This information is somewhat surprising, but really, it seems more prudent to navigate with a paper chart, rather than with an ECDIS based on a RASTER chart that could generate excessive trust in data that are, in fact, erroneous or decaying. Electronic charts do not seem reliable for Arctic navigation and Groupe Desgagnés addresses a serious problem concerning the georeferencing of charts (positioning biases easily reach 3 nautical miles). Remember that if an accident is caused by a hazard indicated on a paper chart, and the said hazard is not indicated on electronic charts, the captain of the vessel will be held responsible in an inquiry. Captains will not use an updated version of an electronic chart to change their trajectory if this information is not available on paper charts. Therefore, according to Groupe Desgagnés, the priorities are: 1. Update Arctic paper charts with the bathymetric and positioning information that is currently available and concurrent with the electronic charts (paper charts have not been updated for a very long time and Groupe Desgagnés still use American military charts in certain areas). 2. Increase cartography and positioning efforts in the Arctic; however, human resource and financial issues at the CHS appear to be hindering the accomplishment of these two priorities. - 39 - 3. There lies a danger in creating seaways in the Arctic: it would restrict a captain’s initiative to deviate from a route in the event that a corridor is blocked by ice. In any case, boat captains should be consulted regarding any initiative establishing shipping corridors (which has never been done). Transport Nanuk made the following observations: Their vessels exclusively use CHS paper and electronic charts for navigating in the Arctic. They do not perform autonomous hydrographic data acquisition. The company has been a member of a consultation committee with the CHS Centre and Arctic in Burlington for several years, for the purpose of identifying priorities for acquiring hydrographic data (difficult bathymetry in certain locations), but mainly for identifying processing priorities. There appears to be a lot of data available, but a lack of resources at the CHS to process it and incorporate it into charts. It would also seem that validating hydrographic data poses a challenge (notably due to problems referencing water levels and inaccurate vertical geodetic references). Transport Nanuk is very satisfied with the collaboration with the CHS; communications seem to be transparent. The Burlington CHS office does not have an acquisition platform dedicated to hydrography in the Arctic; they perform acquisitions of opportunity on DFO ice-breakers when possible (when there is free time after other tasks). Acquiring hydrographic data is not a priority; the CHS does not perform continuous collection in conjunction with their other activities of assimilating hydrographic data in map databases. There are well-charted transportation routes in the Arctic and in proximity to communities. There is a lack of alternative routes allowing mariners to circumvent ice. The company has a keen interest in a more thorough charting of the Ungava Bay section, due to very strong tides. Other areas of interest are Hudson Bay, Foxe Basin, Lancaster Sound and the Northwest Passage route. Both professional organizations raise the issue of consistency between charts (paper or ENC) and modern positioning methods and equipment (GNSS). In fact, GNSS systems undeniably provide more accurate positions than those of old positioning systems that were used to chart coastlines on nautical charts. Therefore, it seems there is a very urgent need to correct the coastlines, using inexpensive current satellite data (at the very least). The two companies also mention the lack of vertical referencing of hydrographic data, which we also commented above. Let us now compare these opinions from professionals, who know the navigation conditions in Northern Canada, with the incident report from the Clipper Adventurer: The Clipper Adventurer (27 August 2010): The following passages are excerpts from the Marine Investigation Report “Grounding, Passenger vessel Clipper Adventurer, Coronation Gulf, Nunavut, 27 August 2010,” drafted by the - 40 - TSB, under the reference M10H0006. We have retained only the passages that relate to our study on cartography. Summary: On 27 August 2010 at approximately 1832 Mountain Daylight Time, the passenger vessel Clipper Adventurer ran aground in Coronation Gulf, Nunavut while on a 14-day Arctic cruise. On 29 August, all 128 passengers were transferred to the CCGS Amundsen and taken to Kugluktuk, Nunavut. The Clipper Adventurer was refloated on 14 September 2010 and escorted to Port Epworth, Nunavut. There was minor pollution and no injuries.” The Clipper Adventurer is a passenger vessel measuring 90.91 m in length, with a forward draught of 4.5 m and an aft draught of 4.6 m. “The Clipper Adventurer is also fitted with a forward looking sonar mounted on the head of the bulbous bow; however, it was unserviceable at the time of the occurrence. Since 1998, the Clipper Adventurer has been extensively used in adventure cruises.” “Before departing anchorage, the bridge team prepared courses from Port Epworth to Kugluktuk using Canadian Hydrographic Service (CHS) chart No. 7777.” “The bridge team used the vessel’s Electronic Chart System (ECS) to monitor the progress of the vessel as displayed on raster navigation chart (RNC) CHS No. 7777. The chief officer who was in charge of the watch monitored the vessel’s progress using parallel indexing on the starboard radar and monitored the water depth on the echo-sounder. The master monitored the portside radar when on the bridge. Once clear of Port Epworth and on course 300°gyro, the vessel was placed on autopilot and proceeded at 13.9 knots. The quartermaster remained on the bridge, to take over the steering when required. Shortly after departing Port Epworth, the chief officer marked a depth of 66 m on the chart in an area near where the chart indicated a depth of 40 m.” “CHS Central and Arctic is responsible for conducting surveys in the Arctic. According to CHS, less than 10% of the Canadian Arctic is surveyed to modern standards, and many charts include information that was obtained more than 50 years ago using less reliable technologies than are available today. The routes commonly used are those that have been surveyed more extensively.” “CHS accepts outside sources of data to issue a chart modification if they consider the information to be sufficiently accurate and if it will serve to improve safety for mariners, in accordance with International Hydrographic Organization (IHO), SOLAS Chapter V and CHS standards and processes.” “The shoal on which the Clipper Adventurer grounded had been previously discovered on 13 September 2007 by the CCGS Sir Wilfrid Laurier while conducting scientific research. The CCGS Sir Wilfrid Laurier reported the shoal to MCTS Iqaluit, who then broadcast a NOTSHIP for CHS o chart n 7777. The notice indicated that, “…a shoal was discovered between the Lawson Islands and the Home Islands in the Southern Coronation Gulf at position 67° 58.25′ N, 112° 40.39′ W. Charted depth in the area: 29 m. Least measured depth:3.3 m. Isolated rock Ref. to NAD 83 datum.” It was still in effect at the time of the grounding. When the CCG Sir Wilfrid Laurier’s crew first discovered the shoal in 2007 they were aware of the risks of crossing an island archipelago on a single line of soundings. They were transiting the area from the north at reduced speed. The bridge team was monitoring the depth sounder and lookouts scanning ahead for discoloration of the water indicative of shallower water. On 14 September 2007, CHS Central and Arctic regional office received the information about the shoal discovered by the CCGS Sir Wilfrid Laurier. It first established that there was a NOTSHIP - 41 - that had been issued. Then, based on the preliminary information received, it determined the location would require more extensive surveying prior to issuing a permanent chart correction. In late 2007, there was an exchange of information between the CCGS Sir Wilfrid Laurier and CHS Central and Arctic regarding the reported shoal. CHS Central and Arctic determined that the depth surveys conducted by the CCGS Sir Wilfrid Laurier were not up to CHS standards. CHS standards are based on those of the International Hydrographic Organization (IHO). In accordance with IHO Regulation B-611.9, 23 permanent chart updates should not be made based on a single vessel report, except in the following instances: They originate from recognized survey vessels, research ships or other vessels/masters known to be reliable; They are reports of shoal depths, preferably accompanied by supporting evidence, e.g., an unambiguous echo-sounder trace, for areas where it is unlikely that corroboration can be obtained; They are the sole source of information in a remote area; They are of particular significance to navigation; or The location is in an area where the level of information flow and lines of communication are poor. Notwithstanding the above, CHS Central and Arctic requires validated data with systematic coverage and a sufficient level of confidence before permanently modifying a chart. For example, it requires 3 types of hydrographic soundings in order to convey an accurate depiction of a hazard on charts: representational (periphery), significant (depths leading to peaks) and critical (peaks). In the summer of 2008, a CHS team of hydrographers on board the CCGS Sir Wilfrid Laurier evaluated the accuracy of the data collected the previous year and confirmed that it was not sufficient to produce a chart correction according to CHS Central and Arctic practice. On 4 September 2008, the passenger vessel Akademik Ioffe transited south into Port Epworth along the same line of soundings as the Clipper Adventurer was to later follow. The vessel’s logbook recorded a depth of 16 m when passing near the 29 m sounding on the CHS chart No. 7777 in proximity to the shoal at 67°58.4’ N, 112°40.0’ W. The vessel was not aware of NOTSHIP A102/07. At that time, the NOTSHIP was no longer being broadcast by radio but was available by other means (see list below). CHS Central and Arctic has a prioritized list of areas to be surveyed. While CHS does not have dedicated vessels for surveys, CHS Central and Arctic typically plans to have 1 or 2 teams conducting surveys in the Arctic for several weeks during the summer navigation season and takes advantage of situations where CCG vessel routes and activities coincide with planned survey site locations on an opportunity basis. In 2009, CHS had planned to survey the shoal based on their prioritized list. However, there was no opportunity at that time to survey the shoal using CCG vessels. Subsequent to the grounding of the Clipper Adventurer, a team of CHS hydrographers on board the CCGS Sir Wilfrid Laurier completed a survey of the area. On 8 October 2010, CHS chart No. 7777 was corrected by a permanent indication of the shoal and a NOTMAR was issued.” The analysis of this incident report reveals that a lack of cartography, rather than a lack hydrography, caused this accident. The seafloor sounding had been surveyed, but not validated o or reported on chart n 7777. However, mariners had been notified. - 42 - SUMMARY According to the information we were able to collect, there seems to be a significant amount of hydrographic or geographical data available (particularly relating to the coastline). But such data cannot be integrated into existing charts, due to a lack of sufficient cartographic production equipment. It would seem priorities are more hydrographic in nature (data collection, improvement of vertical references), rather than cartographic. 5. PRIORITIZATION METHO DS SUITABLE FOR NORTHERN CANADA According to the preceding analysis, cartography in Northern Canada is undeniably underdeveloped. This is essentially due to the lack of geodetic infrastructures, tidal stations and nautical equipment. Because marine cartography is a very expensive operation, it is highly unlikely that in the short-term, solutions regarding the great need of data and sound navigation equipment in the North will be developed. Consequently, there is a need to prioritize hydrographic operations in order to provide a better, more transparent solution to the real needs of mariners in Northern Canada. First, we present two methods of prioritizing hydrographic surveys. One is defined by the Canadian Hydrographic Service, the other was developed in the Southwest Pacific region (a region also suffering from a lack of hydrographic data). Finally, we discuss the relevance of these prioritization methods and make a few recommendations. 5.1 PRIORITIZATION SERVICE AS DEFINED BY THE CANADIAN HYDROGRAPHIC The approach adopted by the CHS was illustrated by its previous director S. Narayanan with the Bay of Fundy case, in the Atlantic region. This approach falls under the national program to prioritize hydrographic surveys, supporting SOLAS directives and pollution prevention in the Arctic. The approach involves reviewing the portfolio of CHS charts, developing a prioritization method, consolidating methods for managing hydrographic and cartographic data and establishing partnerships with a view to sharing hydrographic survey platforms and producing hydrographic and cartographic data. This approach to prioritization is as follows: in the 0 to 50 m band of soundings (sensitive areas), the type of seafloor (complex or relatively flat) and complex estuary zones (sediment movements, significant tidal fluctuation) are classified, and the required CATZOC classification is deducted from that. The required CATZOC is compared against the existing one; the CATZOC gap represents the areas needing attention. This is explained in the following graphic (the slides are from a CHS presentation on this subject). - 43 - Here, this method is illustrated for the Bay of Fundy (NB): We find that in order for this method to function correctly, it requires a priori data, particularly for determining so-called complex areas (morphological analysis of the seafloor). The required CATZOCs are created somewhat arbitrarily, by assigning complex areas to CATZOC A, non-complex areas (but with an a priori depth of less than 50 m) to CATZOC B, and the rest to CATZOC C. - 44 - The existing surveyed CATZOCs can be extracted from CHS databases, depending on what hydrographic data is available in the area. We then determine the gap between these areas, which reveals every area that would need to be re-surveyed. Therefore, the prioritization method analyzes existing marine traffic, by tracing historical AIS trajectories of vessels according to their identification. The incident response time for a marine accident is then analyzed. This allows for the consequences of a marine accident to be weighted. - 45 - Finally, these two data are projected onto the CATZOC gap prioritization chart and a weighting is produced to determine: The intensity and type of traffic; The consequence of an accident (measured exclusively by incident response time). Finally, the following chart is produced: ANALYSIS OF THIS METHOD - 46 - The method proposed by the CHS can be formulated as follows: STEP 1 (Identify the hazards from a hydrographic viewpoint and determine the regions lacking quality hydrographic data): a. Determine which areas are significant from a bathymetric perspective, classifying them according to several criteria: i. Class A: Complex seafloor (shoal, rocks), changing seafloor (subjected to sediment dynamics); ii. Class B: Complex waterway; iii. Class C: Depths of the seafloor greater than 50 m (these hold no concern for navigation safety). b. Assign a potential hydrographic hazard (i.e. a seafloor of type A or B) to a required CATZOC c. Search CATZOC data in databases and determine the gap between the required and existing CATZOC. STEP 2 Determine the risk associated with marine traffic: Analyze marine traffic intensity (frequency of passages) and the consequences of a marine accident related to a grounding. a. Classify and represent traffic according to: i. Dangerous cargo and passenger vessels (>50 people) ii. Other traffic iii. Lack of traffic b. Incident response time c. Weighting between the frequency of traffic and consequence of an accident STEP 3: Calculate the prioritization value by weighting the risk associated with marine traffic and the hydrographic hazard of a collision. This method requires a priori knowledge of bathymetry, for STEP 1 requires such information in order to categorize areas according to hydrographic hazard. The concept of hydrographic hazard is intimately linked to seafloor morphology (presence of shoals, rocks, etc.). We can formalize the notion of risk associated with an event by multiplying the product of a probability (or a frequency) by the cost incurred from the consequences of the said event. Due to a lack of documentation available on the risk assessment developed in STEP 2, we are unable to analyze the method allowing the CHS to estimate the risk associated with a consequence. In fact, STEP 2 appears to be an empirical method closer to risk analysis, but does not actually refer to an explicit, transparent calculation of risk. Nevertheless, for a community of stakeholders to consent to a prioritization method, it is imperative that the method for calculating risk (and hydrographic hazard) is transparent and impeccably documented. In fact, every prioritization method produces results that rely heavily on the risk calculation method, and these results must be obtained using weighting factors agreed upon by everybody. Another comment to be made, subsequent to the preceding comment, concerns the lack of reference to environmental phenomena that could result from collisions with the seafloor (black - 47 - tides, chemical pollution that could affect preserved natural areas). Therefore, the environmental cost of a marine accident does not seem to factor into this risk analysis method. The method’s applicability in Northern Canadian and Arctic waters is questionable. How can this approach be applied if there is a complete lack of hydrographic knowledge? How can the notion of hydrographic hazard be determined? How can marine traffic be analyzed if it is very low (which is effectively the case in Northern Canada). We were not able to analyze the simulation result of this prioritization tool in the case of areas in Northern Canada, but it is fairly evident that given the lack of data for traffic and hydrographic hazards, it cannot prove relevant. Nevertheless, an interesting aspect of this approach is the fact that it does not rely exclusively on past risk associated with the analysis of collisions and various marine accidents, but on the potential and intrinsic risk of a geographical area. Indeed, it would make no sense to conduct an analysis solely based on marine accident reports, which do not constitute a risk in regions with very low marine traffic, such as Northern Canada. The approach developed by the CHS also suffers from a lack of analysis of the economic impact of a hydrographic survey. It is obvious that opening and securing seaways optimizes ship loading, decreases transit insurance rates and, quite simply, allows for the start-up of development projects that require goods to be transported at the lowest possible cost. The approach discussed below is inspired by the IMO’s FSA (Formal Safety Assessment) approach, and takes into account a number of additional parameters for clearly prioritizing hydrographic surveys. It was originally developed by LINZ (Land Information New Zealand) and is currently under review by the International Hydrographic Organization. 5.2 “PARTICIPAT ORY” PRIORITIZATION Here, we present a prioritization method inspired by the one developed by LINZ for the Southwest Pacific region (a region also suffering from a lack of minimal quality hydrographic data) in order to ensure feasible marine traffic. We have adapted the case to Northern Canada, by way of proposing an effective prioritization method for this very unique region. In fact, the additional parameters to take into account are the presence of ice, the strong mining potential of the region, the strong dependence of communities on being resupplied and the impact of transportation costs on the prices of commodities consumed. PRIORITIZATION CONCEPTS The two fundamental parameters to consider in a prioritization approach are: - 48 - The current and potential level of economic activity; The risk of a marine accident and its consequences on the various stakeholders (transportation companies, communities, economic activity) resulting from loss of life at sea and environmental catastrophes. It is important to state that risk analysis in the context of hydrography and marine cartography differs from risk analysis in the field of marine transportation. In the field of marine transportation, the goal of risk analysis is to determine a) the cost incurred by a marine accident and b) the ship safety standards that would minimize this cost. In the field of hydrography, the goal is to estimate the new economic potential of securing a seaway, from the viewpoint of a navigation aid supplied by quality cartography. The proposed general methodology for hydrographic prioritization is as follows: STEP 1: Compile preparatory data (geographical areas, constraints due to the presence of ice, identify and consult with stakeholders, compile traffic data (SOLAS or non-SOLAS vessels), compile data relating to the economy of communities and economic development projects); identify hazards (marine traffic analysis (SOLAS or non-SOLAS vessel), areas of risk, areas devoid of any hydrographic data, areas where basic geographical information is lacking); compile data on the cultural and environmental impact of an accident and protected areas. STEP 2: Determine risk: define the risk criteria (frequency and nature of traffic, consequences of a marine accident and likelihood of these consequences). STEP 3: Economic analysis of each community or economic development project. STEP 4: Hydrographic technical visit, including national and international representatives; review the available hydrographic documentation. STEP 5: Define the priorities of a hydrographic survey, review the costs/benefits of hydrographic surveys, and define a hydrographic and cartographic production plan. As we see here, the economic (economic feasibility and development) and environmental aspects are taken into account from the very beginning of the process. The stakeholders are also included from STEP 1 and in STEP 5. Below is a more detailed description of the five steps. In every step, the use of a GIS is necessary. STEP 1A: DATA COMPILATION In this initial stage, the issues requiring resolution should be clearly defined, by making a list of the constraints, geographic extents, areas grouped according to their community of economic interests and areas targeted exclusively for exploitation projects. It will be necessary to distinguish between areas where traffic must be improved for the viability of communities, and areas where traffic is prospective and exclusively related to industrial and mining development. The groups responsible for implementing a prioritization schema should also be selected. They should be representatives from marine authorities (coast guard), the field of hydrography, the economic world (communities, mining companies) and marine transportation. - 49 - Finally, it is necessary to identify the marine transportation stakeholders that are involved in its growth and security. These stakeholders may include port authorities and representatives from: Transport Canada, communities, provincial and territory governments, transportation companies, Environment Canada, and fisheries. It is very important to compile information regarding SOLAS vessel traffic, as well as the movements of local vessels (fishermen) before developing questionnaires, for questionnaires depend on such information. A questionnaire for marine transportation stakeholders could include the following: 1. Available nautical information a. Type of nautical charts used, CATZOC identification b. Date of last hydrographic survey, technology used, coverage and scale of survey c. Vertical references used and, in the case of data referenced to the WGS84 geodetic system, is the data recent or was old data modified? 2. Charts used for: a. Navigation in the Northwest passages (international traffic) b. Inter-archipelago navigation of Northern Canada c. Port approaches d. Community resupplying e. Mining traffic f. Tourist traffic 3. Volume transported a. AIS data (land and satellite-based) b. Data for ports and communities c. Mining company data 4. Type of vessel a. Passenger vessels (tourism) b. International traffic c. Cargo for community revitalization d. Oil transportation e. Fishing vessels f. Others 5. Frequency of passage by ship type, indicating dimensions (length, height, tonnage) 6. Marine accidents a. Official information b. Proof of patent risk of accident or near-accident c. Issues raised by local communities concerning marine accidents 7. Infrastructures a. Deep water ports b. Project ports (mining) c. Loading docks d. Anchorage areas 8. Nature of access and shore-based infrastructures a. Access channels and ice coverage b. Depth of anchorage areas c. Presence of navigation aids (night/day) and trust in these aids according to season (drift from beacons due to ice melts) - 50 - d. Possibility of navigation aids (presence of ice) e. How are approach areas navigated? f. Radio communications g. Presence of areas protected by environmental standards h. Marine cargo unloading equipment i. Presence of radar approach j. Presence of tows, the nearest ice-breaker ships 9. Possible impact of a marine accident a. Presence of an environmentally protected site b. Pollution with impact on fishery resources 10. Economic impact (possible development related to quality marine cartography) a. Arctic tourism (current, potential) b. Exports (minerals, oil) c. Impact of a loss of: i. Cargo, cargo vehicle ii. Fishing vessel iii. Environmental interest area Site visits allow for meetings to be organized with stakeholders and additional information to be gathered for the questionnaire. Information on the economic gains/losses related to marine safety or the lack thereof can be collected independently, for the economic stakeholders (mining companies and marine transportation companies) are rarely based in the communities. Evidently, the communities themselves supply data on their own economic development related to the securing (or not) of marine traffic. The assessment of economic potential generated by hydrography must be realistic and include an analysis of changes in this economic growth should a major marine accident occur. STEP 1B: HAZARD IDENTIFICATIO N This second facet of STEP 1 is fundamental and is more in line with the CHS prioritization method. It consists of analyzing data collected during the above-described step, geographically analyzing marine traffic density (using a GIS), identifying hazardous marine traffic and, subsequently, analyzing the risk (frequency x consequences) of a marine accident. For this, it is helpful to quantify risk using a risk matrix accepted by every stakeholder involved in the process. Indeed, this is the core of the prioritization process, which involves establishing a consensus as to how to model risk. Below, we briefly describe one possible method of determining risk. It also consists of identifying navigation hazards (or the supposed hazards) and the likely consequences of an accident (specifically, a collision with the seafloor). In the appendix, we provide an example of a decision tree that could be used to determine the consequences of ship grounding. This decision tree is inspired by the marine accident involving passenger vessel CLIPPER ADVENTURER in the Coronation Gulf (Nunavut), on August 27, 2010. The - 51 - consequences of this accident sit astride the two scenarios described in Appendix 1. Appendix 2 applies the same type of information to a cargo vessel supplying a Northern community. The assessment of navigation hazards also includes elements relating to bathymetry, when they can be estimated. Known seafloor morphologies can help actually estimate the likelihood of potential and uncharted hazards being present. Finally, the environmental interest areas under study can be assessed according to their classification. It is also necessary to indicate how these areas would be affected by each type of accident. STEP 2: RISK ASSESSMENT Risk can generally be modeled as the product of a frequency (or probability) of an event by a measure of the cost of its consequences. A transparent and documented methodology for risk measurement should be implemented and accepted by every stakeholder. One method currently used is: - Analyze marine traffic using AIS data; Identify the probable causes of a marine accident; Identify the cost of each consequence (environmental, loss of human life, cost incurred by material damages and costs resulting from service interruption). Each risk factor can be modeled spatially using a layer-based GIS, and the final risk analysis will involve weighting each layer for every cell of the GIS. Analysis of marine traffic As we previously saw, it is imperative to have knowledge of marine traffic in order to determine the level of risk in a given area. This can be obtained from AIS data, the coast guard, or even local communities (for non-SOLAS traffic). The data needs to be manually post-processed in order to be fully and reliably reconstituted, so it can be represented on a GIS. Every type of vessel possesses an intrinsic risk factor, dependent upon its cargo and tonnage. It is helpful to use a linear weighting of the tonnage in order to represent the risk potentials associated with each vessel. In this way, according to the characteristics of the vessel, we can model and calculate its risk potential. Environmental risk factor components Each risk factor component must be quantified according to a scale (from 1 to 5, for example), indicating the frequency (or probability) for each source of risk. a. Meteorological and oceanic conditions (winds, visibility, presence and size of icebergs, sea state) contribute to marine accidents. These conditions can be modeled spatially according to the coastal exposure to the various prevailing wind sectors in each Northern region. Tidal conditions also generate risks (the stronger the tide, the stronger the currents). - 52 - b. Complexity of waterway: Here, we indicate the complexity of channels and approach areas, which require precision manœuvres. This is contrary to ocean navigation, in which mariners can deviate from a route without incurring the risk of a collision. c. Navigation aids: These include nautical charts and radar. We already discussed the notion of CATZOC as defined in the IHO S-57 standard. The equivalent for an ENC is the M_QUAL, which includes a CATZOC attribute. d. Bathymetry plays a role in vessel manoeuvrability (in shallow waters, vessels lose a large part of their propulsion ability). It is frequently noted that the 15 m isobath limits the navigable area. Therefore, it is possible to determine a risk factor according to the distance to the 15 m isobaths of each geographical cell. The type of seafloor (mentioned in the CHS approach) also plays a role. Its hardness and morphology represent a risk factor (soft, flat seabed or a hard, erratic seabed) e. It is possible to make a list of navigation hazards (except for those which are, unfortunately, inaccurately charted) for each area of study. It is necessary to distinguish surface hazards (ice, icebergs) that could cause a vessel to reroute, subsequent to which a natural seaway expansion would have to be considered in order to account for these eventualities; seafloor hazards (irregular seafloors, presence of shoal). Types of consequences a. b. c. d. e. Environmental (proximity to protected areas, fishing areas) Impact on re-rupplying for communities and the local economy Impact on port access or anchorage area Impact on tourism development Impact on economic activity or mining development With the use of a GIS, risk can be assessed by conducting a spatial analysis of each contribution to overall risk, taking into account the frequency of an event, by calculating its consequences and weighting each information layer. It should also be noted that certain criteria could be dispersed spatially according to the distance of a site (cell analysis) to a potential hazard (for example: accidental pollution in proximity to a protected area or the presence of navigational hazards). Every risk factor can be weighted by several coefficients: an inter-class weighting coefficient (risk factors as mentioned above and consequences) and one coefficient per category (a, b, c, d, e). This allows for the definition of a transparent, documented method in which each weighting is clearly defined in relation to other criteria, even if we must concede that all risk quantification is somewhat arbitrary. The previous weightings allow for the different GIS layers to be fused and allow us to obtain an overall risk map, integrating every factor. STEP 3: ECONOMIC ANALYSIS We have seen that risk is defined as the product of the frequency of an event and the measure of the cost of its consequences. Evidently, the consequences of an accident are systematically negative in terms of cost. This is not the case for economic development potential generated by - 53 - the opening or securing of seaways, which could have a significant impact on the growth of communities in Northern Canada. The goal of prioritizing hydrographic surveys (which are aiming to secure seaways) is to foster economic development and decrease marine risk. Therefore, economic development potential should be incorporated into risk in order to define a final prioritization value. For example, an area presenting an elevated level of risk due to uncertain cartography and the presence of significant hazards, but presenting economic development potential related to the securement of a seaway, would be assigned a high level of priority. In the study conducted by Transport Canada “Multi-Purpose Marine Facilities for Cambridge Bay, Pond Inlet, and Rankin Inlet” in 2011, it is clearly established that the construction of deep water ports could have an impact on the development and sustainable economic growth of communities. It is also reported that mining activity, which requires transportation infrastructures, could not generate sustainable growth, given its ephemeral nature. Tourism development in Northern Canada should also be taken into account, for it seems to be experiencing a rather strong increase. The economic analysis in connection with cartographic prioritization could be conducted locally for each coherent group of communities. The important points to be studied are: a. In-depth analysis of commercial traffic, passenger vessels, tourist vessels b. Feasibility and growth of community revitalization, study concerning the impact of deeperdraught vessels on the transportation cost c. Sustainability of mining development and economic impact d. Tourism potential in Northern Canada and impact on communities STEP 4: HYDROGRAPHIC TECHNICAL VISIT This step involves compiling various analyses, organizing technical visits by consulting with hydrographic experts, and relying on national and international authorities (IHO, OMI). The goal of these visits is to review prioritization documentation, study the adequacy of cartographic and hydrographic documentation, as well as infrastructures for disseminating nautical information. International organizations play a fundamental role in this process, as they guarantee the independence of the assessment. STEP 5: DEFINITION OF PRIORITY HYDROGRAPHIC SURVEYS This step concludes the process and specifies the hydrographic surveys to be conducted, as well as their methodology according to the environment, positioning infrastructure, knowledge of tidal fluctuations and geodetic knowledge of the sites. The type of hydrographic survey (reconnaissance, regular, control, obstructions search) must be defined for each area, according to the objectives explicitly defined by the preceding step and depending upon the available budgets. Needless to say, this conclusive phase of a prioritization process requires the participation of every stakeholder (every stakeholder should participate from the outset). - 54 - The following points seem to be particularly important: a. b. c. d. Provide an overview of risk analysis and economic analysis Specify revision/cartographic production schema Specify associated hydrographic surveys Establish charting priorities according to available budgets, risk analysis and economic analysis (cost/benefit) e. Planning hydrographic surveys according to environmental constraints (ice), budget constraints The overview of information collected during steps 1, 2 and 3 can be rather complex if we seek to do more than find an empirical solution to the problem. Let us summarize the situation. Before conducting an analysis, it is necessary to distinguish between the types of information collected and summarized, according to whether they are: Spatial and temporal statistics (type of seafloor, site’s exposure to meteo-oceanic conditions), which is the case for the majority of geographic data; Influenced by the decision to do a major cartographic update or hydrographic survey (marine traffic intensity, economic development). At this stage, we have data for economic development potential (sustainable with regard to communities, and eventually ephemeral, since it consists of exploiting natural resources that are limited in quantity). This data is related to certain locations. We also have specialized data of incurred risk, obtained from step 2, and marine traffic data. All that remains is to model the impact of a decision to conduct a hydrographic survey or do a cartographic update of a given area on all of the economic development parameters. In fact, the decision to conduct a hydrographic survey in a given area would decrease the level of potential risk to navigation on a route, or a portion of the shipping route. As for the route of a vessel, the risk incurred during a leg of the journey would be decreased, which would automatically decrease the transportation cost (part of said cost being directly proportional to technological risk). Therefore, the destination of the said marine route could experience more intense traffic, at lower risk. According to the economic development factor of the destination, the growth of effective economic activity of the destination can be modeled in relation to the decision to prioritize a hydrographic survey. The complexity of the issue arises from the fact that prioritizing a given area can have a cascading effect on numerous destinations. Several destinations can have a common shipping route, or a part of a shipping route, downstream needed to connect them. Therefore, the prioritization of one area can impact several destinations. But the reverse situation exists as well: an approach to a section of route opening up a reliable approach to a given destination can decrease the risk incurred when linking destinations situated downstream. SUMMARY - 55 - To conclude this section, several prioritization methods were analyzed. It would appear the method used by the CHS does not take hydrographic risk into account and ignores the impact on environmental and economic development factors. The second method, inspired by a recent study undertaken by LINZ, takes these factors into account and proposes a completely transparent and participatory approach. We tried adapting it to the Northern Canadian context, accounting for the presence of ice and the unique geographical, sociological and economic traits of this region. CONCLUSION In order to develop a marine transportation network in anticipation of climate changes, it is necessary to secure routes and maritime approaches using cartography of a minimum quality. The unique context of the North and its lack of infrastructure allowing for quality hydrographic surveys to be conducted in recent history explain the deficit of quality nautical information and the cause of certain serious marine accidents. Hydrographic surveying, which requires positioning equipment, accurate geodetic knowledge and tide models allowing for the accurate reduction of soundings, is a major challenge in Northern Canada. Several alternative solutions to so-called “classical” hydrographic methods were considered (the use of autonomous systems and crowdsourced bathymetry in particular). We demonstrated the limits to the application of these two approaches which, unfortunately, cannot provide a reliable and economic solution to the lack of quality data in the North. Marine cartography (and hydrography) is necessary for all economic and human development of a territory. It would appear that, given the immensity of the region and the difficulty of conducting hydrographic surveys in the North, a prioritization phase that a) harnesses communities, provincial governments and concerned federal authorities, and b) unites marine experts, economic stakeholders and experts in hydrography, can serve as a solution leading to transparent and documented strategies, allowing for sensible planning of hydrographic campaigns. Northern Canada, like other regions in the world (the Southwest Pacific, for example), can adopt an effective prioritization method incorporating economic development factors, as well as marinerelated risk factors (which are the result of a lack of nautical information). If adopted, the approach we have proposed here should allow for hydrographic survey planning to be coordinated in conjunction with sustainable economic development, in anticipation of certain consequences due to climate change. - 56 - REFERENCES Chayes, D. N., J. Ardai, R. Anderson, S. Goemmer, B. J. Coakley, M. R. Rognstad, R. B. Davis and M. Edwards (1999). Seafloor Characterization And Mapping Pods (SCAMP): Submarine-mounted Geophysical Mapping. OCEANS-CONFERENCE-, INSTITUTE OF ELECTRICAL & ELECTRONICS ENGINEERS. Comtois, C. (2010). Bathymetry au travers la glace en Arctic : les aspects humains. Conférence Hydrographic du Canada. Québec. Church et al. (2009) Developing Strategies to Facilitate Long Term Seabed Monitoring in the Canadian Arctic using Post Processed GPS and Tidal Models, Porceeding of the US HYDRO Conf Dawson, J., Stewart, E.J., Pearce, T., and Ford, J. 2011. Emerging cruise tourism economies in Canadian Arctic: Ulukhaktok case study. Ottawa: Indian and Northern Affairs Canada Edwards, M. H. and B. J. Coakley (2003). "SCICEX investigations of the Arctic Ocean System." Chemie der Erde-Geochemistry 63(4): 281-328. Fadaie, K. (2012). "Need for Integrated Bathymetric and Topographic Charts in Nothern Communities." Lighthouse 80: 2. Forbes, S., J. Verhoef and D. Mosher (2012). "The Clock is Ticking - The Journey for Canada's Submission to the United Nations Commission on the Limits of the Continental Shelf." Lighthouse(79): 10. Hall, J. K. (2006). "Autonomous Drifting Echo-sounding Buoys." Hydro International 10(5): 3. Hare, R.M. (1997). Procedures for Evaluating and Reporting Hydrographic Data Quality, Internal report, Canadian Hydrographic Service Hare, R., D. R. Peyton and J. Conyon (2012). "Remote Processing of Ship Based Hydrographic Multi-beam Data." Lighthouse 80: 7. Hopkin, D., E. MacNeil, R. Pederson, J. Manning, C. Kaminski, T. Crees and A. Forrest (2010). The Application of Autonomous Underwater Vehicles in Support of the Canadian UNCLOS Submission. Canadian Hydrographic Conference. Québec. Jakobsson, M., N. Cherkis, J. Woodward, R. Macnab and B. Coakley (2000). "New grid of Arctic bathymetry aids scientists and mapmakers." EOS, Transactions American Geophysical Union 81(9): 89-96. Knudsen, P., O. Andersen, R. Forsberg, R. Saldo and H. Skriver (2012). Satellite bathymetry and other satellite derived data: 36. Lehmenhecker, S. and T. Wulff (2013). "Flying Drone for AUV Under-Ice Missions." Sea Technology 54(2): 4. Muggah, J., I. Church, J. Beaudoin and J. H. Clarke (2010). Seamless Online Distribution of Amundsen Multi-beam Data. Canadian Hydrographic Conference. Québec. - 57 - Savi Narayanan (2012) CHS Chart-related Operational Services Update, Mariner’s Workshop February 21-22, Montreal Rognstad, M., R. Anderson, D. Chayes and L. Mayer (2005). Seafloor Sounding in Polar and Remote Regions (SSPARR) Project-Initial Field Trials. AGU Fall Meeting Abstracts. Rondeau, M. and S. Roche (2010). Le WikiSIG : un outil de cartography participative appliqué à l’espace maritime. Conférence Hydrographic du Canada. Québec. Stewart, E.J., Howell, S.E.L., Draper, D., Yackel, J., and Tivy, A. 2007. Sea ice in Canada’s Arctic: Implications for cruise tourism. Arctic 60(4): 370–380 Weber, J. (1983). "Charts of the Arctic Basin seafloor: A history of bathymetry and its interpretation." Arctic: 121-142. - 58 - APPENDIX 1: EXAM PLE OF IDENTIFYING THE CONSEQUENCES OF A M ARINE ACCIDENT INVOLVING A SOLAS PASSENGER VESSEL Most likely scenario: Low-speed collision approaching an anchorage area: - Too much trust in the ECDIS, faulty marine chart Route deviated from the regular channel due to the presence of ice or small icebergs undetectable on the radar Breakdown or no navigation aid Positioning problem Refloating the ship using the machine / ballasting the ship Most likely consequences - Human: minor injuries Vessel: minor damage, no immobilization or towing required Environmental: no damage Stakeholders: loss of tourists’ trust, loss of trust in the port/destination if outdated charts are the cause. International media attention. Worst-case scenario: High-speed collision at sea - Too much trust in the ECDIS, faulty chart Route deviated from regular channel due to the presence of ice, icebergs or small icebergs undetectable on the radar Breakdown or no navigation aid Positioning problem Significant damage, waterway, vessel sliding from the shoal and sinking, abandonment of vessel with weather becoming increasingly inclement. Consequences - Human: loss of human life during abandonment of vessel Vessel: serious consequences or total loss Environmental: pollution (hundreds of tonnes of fuel oil). Fuel tanks needing to be refloated. Stakeholders: loss of company’s reputation. Significant damages, loss of trust in the region, major international media coverage. - 59 - APPENDIX 2: EXAM PLE OF IDENTIFYING THE CONSEQUENCES OF A M ARINE ACCIDENT INVOLVING A CARGO SOLAS VESSEL Most likely scenario: Collision with the seafloor at low speed, approaching an anchorage area: - Too much trust in the ECDIS, faulty nautical chart Route deviated from the regular channel due to the presence of ice, icebergs or small icebergs undetectable by radar Breakdown or no navigation aid Positioning problem Refloating the ship using the machine / ballasting the ship Most likely consequences - Human: minor injuries Vessel: minor damage, no grounding or towing Environmental: loss of fuel oil, coastal damage Stakeholders: considerable delays in re-supplying a community, loss of reputation at a shipyard or anchorage area, consequences for local fishing businesses Worst-case scenario: High-speed collision at sea - Too much trust in the ECDIS, faulty chart Route deviated from the regular channel due to the presence of ice or small icebergs undetectable on the radar Breakdown or no navigation aid Positioning problem Significant damage: waterway, partial loss of containers due to excessive listing, very difficult refloating requiring a large-scale operation, with weather becoming increasingly inclement. Consequences - Human: minor injuries Vessel: serious consequences and major repairs needed Environment: pollution (hundreds of tonnes of fuel oil). Fuel tanks needing to be refloated. Stakeholders: serious problem revitalizing a community, loss of containers. Major damages, major international media coverage. - 60 -