6 GLORYS Gridded Sea Bottom Temperature

Data Type: Gridded Data

Spatial Scope: Northwest Atlantic

Duration: 2020-2025

Source: CMEMS Global Ocean Physics Reanalysis

6.1 Introduction to Indicator

The GLORYS12V1 dataset is a product of the Copernicus Marine Environment Monitoring Service (CMEMS), providing spatiotemporal modeled estimates of environmental data for global oceans at 1/12° horizontal resolution and 50 vertical levels (Copernicus Marine Service 2023). Here, we use the bottom temperature estimate for the Northwest Atlantic, encompassing the Scotian Shelf region.

GLORYS bottom temperature estimates are generated using a numerical ocean model with assimilated satellite and in-situ observations, but see Chapter 5 for summarized empirical temperature measurements.

6.2 View Data

library(terra)
library(plotly)

# prepare sf shapes as coordinates
# NAFO regions
#sf <- readRDS("data/derived_data/azmp_area.rds") %>%
#  filter(area %in% c("4X","4Vn","4Vs","4W")) %>%
#  inner_join(global_cols2 %>% rename("area" = "region"))
#sf <- sf %>% st_transform(crs = crs(r_stack))
#sf_coords <- st_coordinates(st_cast(sf, "MULTILINESTRING") ) %>%
#  as.data.frame() %>% 
#  split(.,f = .$L2) %>%
#  purrr::map(.x = .,
#             .f = ~.x %>% rbind(c(NA,NA,NA,NA))) 
#

sf <- readRDS("data/derived_data/nafo_regions_correct.rds") %>%
  filter(region %in% c("4X","4Vn","4Vs","4W")) %>%
  inner_join(global_cols2) %>%
  rename(area = region)
sf <- sf %>% st_buffer(-.05)
sf <- sf %>% st_transform(crs = crs(r_stack)) 
# add NA after multipolygons so that plotly draws lines correctly
sf_coords <- st_coordinates(st_cast(sf, "MULTILINESTRING") ) %>%
  as.data.frame() %>% 
  split(.,f = interaction(.$L2,.$L1), drop = T) %>%
  purrr::map(.x = .,
             .f = ~.x %>% rbind(c(NA,NA,NA,NA))) 
# bind back together by region
name_ids <- sub("\\..*$", "",  names(sf_coords))
grouped <- split(sf_coords, name_ids)
grouped <- purrr::map(.x = grouped, 
                      .f = ~.x %>% bind_rows)

data("coastline")
coast_coords <- st_coordinates(coastline) %>%
    as.data.frame() %>% 
  split(.,f = interaction(.$L1,.$L2,.$L3, drop = T)) %>% 
  purrr::map(.x = .,
             .f = ~.x %>% rbind(c(NA,NA,NA,NA,NA))) %>%
  bind_rows()   


times <- names(r_stack)

# consistent color scale
zmin <- range(values(r_stack), na.rm = TRUE)[1]
zmax <- range(values(r_stack), na.rm = TRUE)[2]

x_vals <- xFromCol(r_stack, 1:ncol(r_stack))
y_vals <- yFromRow(r_stack, 1:nrow(r_stack))
ext <- ext(r_stack)

# Base plot
p <- plot_ly(
  x = x_vals, y = y_vals,
  z =as.matrix(r_stack[[1]], wide = TRUE),
  type = "heatmap",
  colorscale = "YlGnBu",
  hoverinfo = "none",
  colorbar = list(title = "Bottom<br>Temperature<br>(°C)"),
  zmin = zmin, zmax = zmax,
  showscale = TRUE
)

# add polygons
p <- p %>%
    add_polygons(
    x = coast_coords[,1],
    y = coast_coords[,2],
    fillcolor = "#dcdedc",
    showlegend = F,
    line = list(color = "black", width= .2),
    hoverinfo = "none",
    inherit = FALSE
  ) %>% 
  add_paths(x = grouped[[1]][,1],
            y = grouped[[1]][,2],
            line = list(color = unique(sf$color[1]),
                        width = 3),
            hoverinfo = "none",
            inherit = F,
            name = unique(sf$area[1])) %>%
    add_paths(x = grouped[[2]][,1],
            y = grouped[[2]][,2],
            line = list(color = unique(sf$color[2]),
                        width = 3),
            hoverinfo = "none",
            inherit = F,
            name = unique(sf$area[2])) %>%
    add_paths(x = grouped[[3]][,1],
            y = grouped[[3]][,2],
            line = list(color = unique(sf$color[3]),
                        width = 3),
            hoverinfo = "none",
            inherit = F,
            name = unique(sf$area[3])) %>%
    add_paths(x = grouped[[4]][,1],
            y = grouped[[4]][,2],
            line = list(color = unique(sf$color[4]),
                        width = 3),
            hoverinfo = "none",
            inherit = F,
            name = unique(sf$area[4]))


# loop through raster layers to make frames
frames <- purrr::map(
  .x = c(1:nlyr(r_stack)),
  .f = ~list(
    name = times[.x],
    data = list(list(
      z = as.matrix(r_stack[[.x]], wide = TRUE),
      type = "heatmap",
      zmin = zmin,
      zmax = zmax,
      hoverinfo = "none"
    ))
  )
)

# add frames to map
p$x$frames <- frames

# make animation list
steps <- purrr::map(
  .x = c(1:nlyr(r_stack)),
  .f = ~list(
    method = "animate",
    label = times[.x],
    args = list(
      list(times[.x]),
      list(
        mode = "immediate",
        frame = list(duration = 0, redraw = T),
        transition = list(duration = 0)
      )
    )
  )
)

# Slider steps (scrub)
steps <- lapply(seq_along(times), function(i) {
  list(
    method = "animate",
    label  = times[i],
    args   = list(
      list(times[i]),
      list(
        mode = "immediate",
        frame = list(duration = 0, redraw = TRUE),
        transition = list(duration = 0)
      )
    )
  )
})

# Layout + aspect lock + slider + play/pause
p <- p %>%
  layout(
    xaxis = list(
      title = "",
      showgrid = FALSE, zeroline = FALSE,
      range = c(xmin(ext), xmax(ext)),
      constrain = "domain"
    ),
    yaxis = list(
      title = "",
      showgrid = FALSE, zeroline = FALSE,
      range = c(ymin(ext), ymax(ext)),
      scaleanchor = "x",
      scaleratio = 1.5
    ),
    margin = list(l = 0, r = 0, t = 50, b = 70),

    sliders = list(list(
      active = 0,
      y = -.05,
      currentvalue = list(prefix = "Time: "),
      pad = list(t = 10),
      steps = steps
    )),

    updatemenus = list(list(
      type = "buttons",
      direction = "left",
      x = 0.02, y = 1.15,
      xanchor = "left", yanchor = "top",
      pad = list(r = 10, t = 0),
      buttons = list(
        # PLAY: explicitly animate through *all frame names*
        list(
          label = "Play",
          method = "animate",
          args = list(
            as.list(times),
            list(
              mode = "immediate",
              fromcurrent = TRUE,
              frame = list(duration = 250, redraw = TRUE),
              transition = list(duration = 0)
            )
          )
        ),
        # PAUSE
        list(
          label = "Pause",
          method = "animate",
          args = list(
            list(NULL),
            list(
              mode = "immediate",
              frame = list(duration = 0, redraw = FALSE),
              transition = list(duration = 0)
            )
          )
        )
      )
    ))
  )

p %>% config(displayModeBar = F)

Figure 6.1: GLORYS bottom temperature data for Scotian Shelf region; 2020-2025. Use pause and play button or slider bar to filter through month/years, and click legend to isolate target NAFO divisions.

6.3 Summary and Trends

Trend and summary values are automatically generated; data were last updated on marea package install on 2026-02-10

Sea bottom temperatures in the GLORYS dataset are highly heterogenous across space andtime in the Scotian Shelf region, with higher temperatures and greater temporal variability in coastal regions (Fig. 6.1).

Monthly mean values in this dataset span only a short time period (2020-2025). Among NAFO divisions within the Scotian Shelf, region 4Vs had the lowest mean temperature (2.76°C), and region 4X had the highest mean temperature (5.66°C) (average of individual cell means over time). Due to the short sampling duration and high seasonal variability in this dataset, trends of temperature change within regions are displayed here for summer (August) and winter (February) separately, with target months chosen to maximize data in each time series. In all cases, temperature decreased over time (Fig. 6.2).

Figure 6.2: Coefficients of GLORYS bottom temperature trends over time (2020-2025) for summer (August) and winter (February) withing NAFO divisions in the Eastern and Western Scotian Shelf regions. Error bars indicate ± 1 standard deviation of individual raster cell trends within each region.

Across the greater Scotian Shelf and Northwest Atlantic region, both mean temperature and temperature trend (here, shown as the slope of August temperatures over time) were highly variable across space, with coastal shelf regions cooling on average, and outer shelf regions experiencing little change (Fig. 6.3).

Figure 6.3: Map of mean monthly temperatures (left), and summer (month of August) temperature change trends over time (right) for the Scotian Shelf/Northwest Atlantic region. Hover over NAFO divisions to view summary values per region.

6.3.1 Summary Table by Region

Summary values for the GLORYS Sea Bottom Temperature within each individual NAFO division in the Eastern and Western Scotian Shelf are shown below (Table 6.1)

Table 6.1: Summary values of GLORYS Sea Bottom Temperature for NAFO divisions in the Eastern Scotian Shelf (4Vn, 4Vs, 4W), and Western Scotian Shelf (4X).

Metric	Eastern Scotian Shelf	Western Scotian Shelf
Overall Mean	4Vn: 5.28°C 4Vs: 2.76°C 4W: 3.94°C	4X: 5.66°C
Warming Trend (August 2020-2025)	4Vn: -0.07 ± 0.13°C/year 4Vs: -0.08 ± 0.26°C/year 4W: -0.14 ± 0.25°C/year	4X: -0.17 ± 0.23°C/year
Warming Trend (Winter 2020-2025)	4Vn: -0.13 ± 0.08°C/year 4Vs: -0.06 ± 0.21°C/year 4W: -0.16 ± 0.28°C/year	4Vn: -0.13 ± 0.08°C/year 4Vs: -0.06 ± 0.21°C/year 4W: -0.16 ± 0.28°C/year
Most Recent Value (2025-5)	4Vn: 4.66 ± 1.9°C/year 4Vs: 2.38 ± 0.94°C/year 4W: 3.31 ± 2.1°C/year	4X: 4.8 ± 2.52°C/year

6.4 Relevance to Research and Stock Assessments

Sea bottom temperature can strongly affect the populations, distribtuions, and productivity of marine species. See Chapter 5 for an overview of known effects in Atlantic Canada.

6.5 Variable Definitions

GLORYS bottom temperature is stored as a spatial dataframe in marea. See variable names and definitions in the table below (Table 6.2)

Table 6.2: Column names and definitions in the glorys bottom temperature dataset.

variable	description	unit
geometry	Spatial polygon denoting the boundaries of 12° grid cells	Spatial Polygon
time_descriptor	Year and month of modeled value
value	Modeled sea bottom temperature	°C
year	Year of modeled value
month	Month of modeled value

6.6 Additional Data

Data in the GLORYS Bottom Temperature dataset are gridded with no regional groupings. Summary values displayed on this page were generated, not explicitly provided in the marea package.

6.7 Get the Data

library(marea)
data('glorys_bottom_temperature')
plot(glorys_bottom_temperature, style = "fill")

References

Copernicus Marine Service. 2023. “Global Ocean Physics Reanalysis (GLOBAL_MULTIYEAR_PHY_001_030).” https://data.marine.copernicus.eu/product/GLOBAL_MULTIYEAR_PHY_001_030/description.