In EffetsCumulatifsNavigation/ceanav: Research compendium for project on the assessment cumulative effects of marine vessel activities in the St. Lawrence

Shipping {#navigation}

We thank Clément Chion and Jean-François Sénécal of the Université du Québec en Outaouais, Jeff Campagnola of Transport Canada and Samuel Turgeon of the Saguenay-St. Lawrence Marine Park for the discussions and support.

meta <- read_yaml("../data/data-metadata/int_st_navigation.yml")
show_source(meta$rawData, lang = "en")

The intensity of shipping within the study area was assessed using satellite data from the Automatic Identification System (AIS) developed by Maerospace Corporation and acquired for the Canadian Space Agency [@tc2020; @tc2020b]. This database contains r meta$dataDescription$observations$total vessel locations (r meta$dataDescription$observations$moyenne $\pm$ r meta$dataDescription$observations$sd observations per year) for r sum(meta$dataDescription$categories$boats, na.rm = TRUE) vessels between r meta$dataDescription$temporal$start and r meta$dataDescription$temporal$end. It is important to note that this number does not indicate the number of transits, but rather the number of geographic locations identified by the AIS.

Individual location data were transformed into linear shipping segments representing individual and continuous vessel transits, assuming a straight line transit between successive locations. Vessels were identified by their Maritime Mobile Service Identity (MMSI) number. All available locations were used for each vessel to plot their movements within the study area. Since vessels may enter and exit the study area, or stay in berth for longer periods of time, vessel movements were segmented into a number of individual linear segments. Segmentation was performed when two consecutive locations were over 5 hours or more than 80.5 km apart. Linear segments overlapping the land environment at a distance of over 2 km were removed from the database. This arbitrary distance of 2 km was used to remove outlier segments from the database, while still keeping segments describing coastal shipping or vessels frequenting port facilities. This segmentation resulted in r meta$dataDescription$segments linear shipping segments. Since a number of vessel types navigate the waters of the St. Lawrence and Saguenay rivers, we considered r length(meta$dataDescription$categories$accronyme)-2 different vessel categories to complete the characterization of shipping intensity in the study area (Table \@ref(tab:navCat)). Among the available categories, we removed the Other category due to its lack of specificity and the Fishing category since few transits were available in the database.

To characterize shipping, we used an intensity index characterizing shipping in terms of single use -- or transits. For each vessel category, we assessed shipping intensity within the study area by the number of shipping segments that intersect each 1 $km^2$ cell of the study grid between r meta$dataDescription$temporal$start and r meta$dataDescription$temporal$end. This index results in an assessment of shipping intensity expressed as number of vessels ($n$) per $km^2$ ($n$ $vessels$ $*$ $km^{-2}$)

Despite the effort to segment and remove impossible transits, some impossible transits may still remain. For example, a very small number of transits is observed north of Île d’Orléans for cargo shipping (Figure \@ref(fig:CARGO)); these transits could not be removed by the automatic segmentation that was done and the total number of transits makes it impractical to remove them manually. However, there are very few of the impossible transits in the database, so this does not affect the final results of the cumulative effects assessment of marine vessel activities in the study area.

tab <- as.data.frame(meta$dataDescription$categories)
tab <- arrange(tab, english)
tab$transit <- as.character(round(tab$transit, 2))
tab <- tab[!tab$accronyme %in% c("Observation","navigation_portuaire"), ] # Remove observation mm and ports
tab <- select(tab, english, boats, observations, segments)


tab %>%
  knitr::kable(
    col.names = c("Ship type", "Ships", "Locations", "Tracks"),
    row.names = FALSE,
    caption = "Description of the types of ships considered for the cumulative effects assessment of marine vessel activities in the St. Lawrence and the Saguenay river and their representation in the Automatic Identification System data used [@tc2020; @tc2020b]. "
  )

knitr::include_graphics("./figures/figures-integrated/navigation-DRY.BULK.png")

knitr::include_graphics("./figures/figures-integrated/navigation-CARGO.png")

knitr::include_graphics("./figures/figures-integrated/navigation-GOVERNMENT.RESEARCH.png")

knitr::include_graphics("./figures/figures-integrated/navigation-PLEASURE.VESSELS.png")

knitr::include_graphics("./figures/figures-integrated/navigation-SPECIAL.SHIPS.png")

knitr::include_graphics("./figures/figures-integrated/navigation-PASSENGER.FERRY.RO.RO.png")

knitr::include_graphics("./figures/figures-integrated/navigation-TANKER.png")

knitr::include_graphics("./figures/figures-integrated/navigation-CONTAINER.png")

knitr::include_graphics("./figures/figures-integrated/navigation-TUGS.PORT.png")

We have completed the description of shipping intensity with two additional shipping categories. The first deals with the intensity of marine mammal watching activities at sea on craft with a class 1 permit in 2017 [@turgeon2019]. The data available from @turgeon2019 provides a characterization of the number of observation vessels that went through a regular 500m x 500m grid. We resampled this grid to distribute the number of transits recorded in it within our 1 $km^2$ study grid. Although this exercise is limited to the Saguenay-St. Lawrence Marine Park, this is where most of the marine mammal watching excursions take place in the study area.

knitr::include_graphics("./figures/figures-integrated/navigation-Observation.png")

The second category deals with shipping in port areas, i.e. activities that take place near port facilities such as ports and marinas. Using the assumption that shipping activities are more prevalent in port areas, AIS data were used to quantify the number of transits through the study area port areas. The port areas were characterized using data on the location of port facilities [@mtq2021] and industrial port areas [@mei2021] in Quebec. A port area zone of influence was plotted by applying a buffer zone of 200 $m$. This zone is deliberately narrow in order to primarily capture vessels that frequent port areas rather than vessels transiting the St. Lawrence River waterway, especially for facilities in the fluvial environment. The number of transits through these port areas was then calculated using the AIS data. According to this approach, a total of r meta$dataDescription$navigationPortuaire$transits transits were identified in port areas between r meta$dataDescription$temporal$start and r meta$dataDescription$temporal$end. These transits were then plotted in the 1 $km^2$ study grid by assigning to each cell the number of transits of the most heavily frequented overlapping port area. For example, if a grid cell was affected by a first port area with 100 transits and a second area with 200 transits, that cell was assigned a value of 200.

knitr::include_graphics("./figures/figures-integrated/navigation-navigation_portuaire.png")

EffetsCumulatifsNavigation/ceanav documentation built on April 17, 2023, 1:02 p.m.