TidyTuesday: Coastal Ocean Temperature - How Warming Reaches the Depths

TidyTuesday

Climate

Data Viz

7 years of daily ocean temperatures at 7 depths reveal how warming penetrates coastal waters differently by season and depth

Author

chokotto

Published

April 8, 2026

Overview

The ocean does not warm uniformly. Surface waters respond quickly to atmospheric changes, while deeper layers lag behind – buffered by density, mixing, and stratification. This week’s TidyTuesday data lets us see that structure directly: 7 years of daily temperatures at 7 depths from a single coastal monitoring station in Nova Scotia, Canada.

The data comes from the Centre for Marine Applied Research (CMAR) Coastal Monitoring Program at Birchy Head, Lunenburg County. Sensors at depths from 2m to 40m recorded temperatures from February 2018 through December 2025.

Data Source: TidyTuesday 2026-03-31 / CMAR Coastal Monitoring Program
Records: 19,165 daily depth-averaged observations
Depths: 2m, 5m, 10m, 15m, 20m, 30m, 40m
Period: 2018-02-20 to 2025-12-06
Angle: How temperature structure changes with depth, season, and over time

Data

Show code

library(tidyverse)
library(scales)
library(glue)
library(patchwork)

Show code

data_dir <- file.path(getwd(), "data")
ocean_temp <- read_csv(file.path(data_dir, "ocean_temperature.csv"),
                       show_col_types = FALSE)
deployments <- read_csv(file.path(data_dir, "ocean_temperature_deployments.csv"),
                        show_col_types = FALSE)

ocean_temp <- ocean_temp |>
  mutate(
    date = as.Date(date),
    year = year(date),
    month = month(date),
    month_name = month(date, label = TRUE),
    depth_label = paste0(sensor_depth_at_low_tide_m, "m")
  )

cat(glue(
  "Records: {format(nrow(ocean_temp), big.mark = ',')}\n",
  "Date range: {min(ocean_temp$date)} to {max(ocean_temp$date)}\n",
  "Depths: {paste(sort(unique(ocean_temp$sensor_depth_at_low_tide_m)), collapse = ', ')}m\n",
  "Years: {paste(sort(unique(ocean_temp$year)), collapse = ', ')}"
))

Records: 19,165
Date range: 2018-02-20 to 2025-12-06
Depths: 2, 5, 10, 15, 20, 30, 40m
Years: 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025

Show code

NOTE_TEXT   <- "Daily mean temperature at Birchy Head, Nova Scotia (2018-2025)"
SOURCE_TEXT <- "TidyTuesday 2026-03-31 / CMAR Coastal Monitoring Program"
CAPTION     <- glue("Note: {NOTE_TEXT}  |  Source: {SOURCE_TEXT}  |  \u00A9 2026 chokotto")

theme_fm <- theme_minimal(base_size = 12) +
  theme(
    plot.background = element_rect(fill = "white", color = NA),
    panel.background = element_rect(fill = "#f8fafc", color = NA),
    panel.grid.major = element_line(color = "#e2e8f0", linewidth = 0.3),
    panel.grid.minor = element_blank(),
    text = element_text(color = "#334155"),
    axis.text = element_text(color = "#475569"),
    plot.title = element_text(color = "#1e293b", face = "bold", size = 14),
    plot.subtitle = element_text(color = "#64748b", size = 10),
    plot.caption = element_text(
      face = "italic", color = "#94a3b8", size = 9,
      hjust = 0, margin = margin(t = 12)
    ),
    plot.caption.position = "plot",
    strip.text = element_text(color = "#1e293b", face = "bold"),
    legend.background = element_rect(fill = "white", color = NA),
    legend.text = element_text(color = "#475569"),
    plot.margin = margin(15, 15, 15, 15)
  )

ACCENT   <- "#e63946"
BLUE     <- "#3b82f6"
AMBER    <- "#f59e0b"
GREEN    <- "#10b981"
MUTED    <- "#94a3b8"

Visualizations

1. Temperature Structure: The Full Picture

The heatmap below shows every day and every depth. The seasonal pulse is unmistakable at the surface, but it fades and compresses with depth – the ocean’s thermal inertia made visible.

Show code

p1 <- ocean_temp |>
  ggplot(aes(x = date, y = factor(sensor_depth_at_low_tide_m),
             fill = mean_temperature_degree_c)) +
  geom_tile(width = 1) +
  scale_fill_viridis_c(
    option = "turbo",
    name = "Temp (\u00B0C)",
    limits = c(-1, 22),
    breaks = seq(0, 20, by = 5)
  ) +
  scale_x_date(
    date_breaks = "1 year",
    date_labels = "%Y",
    expand = c(0.01, 0)
  ) +
  scale_y_discrete(limits = rev(c("2", "5", "10", "15", "20", "30", "40")),
                   labels = function(x) paste0(x, "m")) +
  labs(
    title = "Surface waters swing 20\u00B0C across seasons -- deep layers barely move",
    subtitle = "Daily mean temperature by depth at Birchy Head, Nova Scotia (2018-2025)",
    caption = CAPTION,
    x = NULL, y = "Depth"
  ) +
  theme_fm +
  theme(
    legend.position = "right",
    panel.grid.major = element_blank()
  )

p1

2. Seasonal Profiles: How Deep Does Summer Reach?

By averaging across years, we can isolate the seasonal cycle at each depth. Surface waters (2m) reach 18-20 degrees C in summer, while the 40m layer stays below 8 degrees C year-round.

Show code

seasonal <- ocean_temp |>
  group_by(month, sensor_depth_at_low_tide_m) |>
  summarise(
    mean_temp = mean(mean_temperature_degree_c, na.rm = TRUE),
    sd_temp = sd(mean_temperature_degree_c, na.rm = TRUE),
    .groups = "drop"
  )

depth_colors <- setNames(
  viridisLite::viridis(7, option = "viridis"),
  c("2", "5", "10", "15", "20", "30", "40")
)

p2 <- seasonal |>
  ggplot(aes(x = month, y = mean_temp,
             color = factor(sensor_depth_at_low_tide_m),
             group = factor(sensor_depth_at_low_tide_m))) +
  geom_ribbon(aes(ymin = mean_temp - sd_temp, ymax = mean_temp + sd_temp,
                  fill = factor(sensor_depth_at_low_tide_m)),
              alpha = 0.1, color = NA) +
  geom_line(linewidth = 1.2) +
  geom_point(size = 1.5) +
  scale_x_continuous(breaks = 1:12, labels = month.abb) +
  scale_color_manual(values = depth_colors, name = "Depth",
                     labels = function(x) paste0(x, "m")) +
  scale_fill_manual(values = depth_colors, guide = "none") +
  labs(
    title = "Summer warmth barely penetrates below 20m -- the deep stays cold year-round",
    subtitle = "Monthly mean temperature by depth (ribbon = +/- 1 SD across years)",
    caption = CAPTION,
    x = NULL, y = "Temperature (\u00B0C)"
  ) +
  theme_fm +
  theme(legend.position = "right")

p2

3. Year-over-Year Trend: Is the Surface Warming Faster?

Comparing annual means by depth reveals whether warming is uniform or concentrated at the surface.

Show code

annual <- ocean_temp |>
  filter(month >= 3 & month <= 11) |>
  group_by(year, sensor_depth_at_low_tide_m) |>
  summarise(mean_temp = mean(mean_temperature_degree_c, na.rm = TRUE),
            .groups = "drop")

highlight_depths <- c(2, 20, 40)
annual_highlight <- annual |>
  filter(sensor_depth_at_low_tide_m %in% highlight_depths)

p3 <- annual |>
  ggplot(aes(x = year, y = mean_temp,
             group = factor(sensor_depth_at_low_tide_m))) +
  geom_line(color = MUTED, alpha = 0.3, linewidth = 0.8) +
  geom_line(data = annual_highlight,
            aes(color = factor(sensor_depth_at_low_tide_m)),
            linewidth = 1.4) +
  geom_point(data = annual_highlight,
             aes(color = factor(sensor_depth_at_low_tide_m)),
             size = 2.5) +
  scale_color_manual(
    values = c("2" = ACCENT, "20" = BLUE, "40" = "#440154"),
    name = "Depth",
    labels = function(x) paste0(x, "m")
  ) +
  scale_x_continuous(breaks = 2018:2025) +
  labs(
    title = "Surface and mid-depth temperatures show more year-to-year variability than the deep",
    subtitle = "Annual mean temperature (Mar-Nov) by depth | Grey = other depths",
    caption = CAPTION,
    x = NULL, y = "Temperature (\u00B0C)"
  ) +
  theme_fm +
  theme(legend.position = "right")

p3

4. Variability by Depth: The Surface is Noisy, the Deep is Stable

Standard deviation of daily temperatures reveals where the ocean is most variable – and where it is remarkably stable.

Show code

variability <- ocean_temp |>
  group_by(sensor_depth_at_low_tide_m) |>
  summarise(
    mean_sd = mean(sd_temperature_degree_c, na.rm = TRUE),
    overall_sd = sd(mean_temperature_degree_c, na.rm = TRUE),
    .groups = "drop"
  ) |>
  mutate(
    depth_label = paste0(sensor_depth_at_low_tide_m, "m"),
    fill_color = if_else(overall_sd == max(overall_sd), ACCENT, BLUE)
  )

p4 <- variability |>
  ggplot(aes(x = fct_reorder(depth_label, sensor_depth_at_low_tide_m),
             y = overall_sd, fill = fill_color)) +
  geom_col(width = 0.6) +
  geom_text(aes(label = sprintf("%.1f\u00B0C", overall_sd)),
            vjust = -0.5, size = 4, color = "#475569") +
  scale_fill_identity() +
  scale_y_continuous(expand = expansion(mult = c(0, 0.15))) +
  labs(
    title = "Surface temperature varies 4x more than the deepest layer",
    subtitle = "Standard deviation of daily mean temperature across the full observation period",
    caption = CAPTION,
    x = "Depth", y = "Temperature SD (\u00B0C)"
  ) +
  theme_fm +
  theme(legend.position = "none")

p4

Key Findings

The ocean has a thermal memory. Surface waters (2m) swing 20+ degrees C across seasons, but the 40m layer remains nearly constant, buffered by thermal stratification.
Summer warmth barely reaches 20m. The seasonal pulse attenuates sharply with depth. Below 20m, the annual temperature range compresses to just a few degrees.
Variability decreases monotonically with depth. The standard deviation of daily temperatures at 2m is roughly 4x that at 40m, confirming that deeper waters act as thermal stabilizers.
Year-to-year variation is most visible at the surface. Inter-annual differences in annual mean temperature are larger at shallower depths, suggesting surface layers are more responsive to atmospheric forcing.

This post is part of the TidyTuesday weekly data visualization project.

Disclaimer

This analysis is for educational and practice purposes only. Data visualizations and interpretations are based on the provided dataset and may not represent complete or current information. Single-station observations cannot be generalized to broader ocean warming trends.

--- title: "TidyTuesday: Coastal Ocean Temperature - How Warming Reaches the Depths" description: "7 years of daily ocean temperatures at 7 depths reveal how warming penetrates coastal waters differently by season and depth" date: "2026-04-08" x-posted: true author: "chokotto" categories: - TidyTuesday - R - Climate - Data Viz engine: knitr source-topic: "Coastal Ocean Temperature by Depth" source-url: "https://github.com/rfordatascience/tidytuesday/blob/main/data/2026/2026-03-31/readme.md" image: "thumbnail.svg" code-fold: true code-tools: true code-summary: "Show code" twitter-card: card-type: summary_large_image image: "thumbnail.png" title: "TidyTuesday: Coastal Ocean Temperature by Depth" description: "7 years of daily temperatures at 7 depths - how warming reaches the sea floor" --- ## Overview The ocean does not warm uniformly. Surface waters respond quickly to atmospheric changes, while deeper layers lag behind -- buffered by density, mixing, and stratification. This week's TidyTuesday data lets us see that structure directly: **7 years of daily temperatures at 7 depths** from a single coastal monitoring station in Nova Scotia, Canada. The data comes from the Centre for Marine Applied Research (CMAR) Coastal Monitoring Program at Birchy Head, Lunenburg County. Sensors at depths from 2m to 40m recorded temperatures from February 2018 through December 2025. - **Data Source**: [TidyTuesday 2026-03-31](https://github.com/rfordatascience/tidytuesday/blob/main/data/2026/2026-03-31/readme.md) / CMAR Coastal Monitoring Program - **Records**: 19,165 daily depth-averaged observations - **Depths**: 2m, 5m, 10m, 15m, 20m, 30m, 40m - **Period**: 2018-02-20 to 2025-12-06 - **Angle**: How temperature structure changes with depth, season, and over time ## Data ```{r} #| label: load-packages #| message: false #| warning: false library(tidyverse) library(scales) library(glue) library(patchwork) ``` ```{r} #| label: load-data #| message: false data_dir <- file.path(getwd(), "data") ocean_temp <- read_csv(file.path(data_dir, "ocean_temperature.csv"), show_col_types = FALSE) deployments <- read_csv(file.path(data_dir, "ocean_temperature_deployments.csv"), show_col_types = FALSE) ocean_temp <- ocean_temp |> mutate( date = as.Date(date), year = year(date), month = month(date), month_name = month(date, label = TRUE), depth_label = paste0(sensor_depth_at_low_tide_m, "m") ) cat(glue( "Records: {format(nrow(ocean_temp), big.mark = ',')}\n", "Date range: {min(ocean_temp$date)} to {max(ocean_temp$date)}\n", "Depths: {paste(sort(unique(ocean_temp$sensor_depth_at_low_tide_m)), collapse = ', ')}m\n", "Years: {paste(sort(unique(ocean_temp$year)), collapse = ', ')}" )) ``` ```{r} #| label: setup-theme #| message: false NOTE_TEXT <- "Daily mean temperature at Birchy Head, Nova Scotia (2018-2025)" SOURCE_TEXT <- "TidyTuesday 2026-03-31 / CMAR Coastal Monitoring Program" CAPTION <- glue("Note: {NOTE_TEXT} | Source: {SOURCE_TEXT} | \u00A9 2026 chokotto") theme_fm <- theme_minimal(base_size = 12) + theme( plot.background = element_rect(fill = "white", color = NA), panel.background = element_rect(fill = "#f8fafc", color = NA), panel.grid.major = element_line(color = "#e2e8f0", linewidth = 0.3), panel.grid.minor = element_blank(), text = element_text(color = "#334155"), axis.text = element_text(color = "#475569"), plot.title = element_text(color = "#1e293b", face = "bold", size = 14), plot.subtitle = element_text(color = "#64748b", size = 10), plot.caption = element_text( face = "italic", color = "#94a3b8", size = 9, hjust = 0, margin = margin(t = 12) ), plot.caption.position = "plot", strip.text = element_text(color = "#1e293b", face = "bold"), legend.background = element_rect(fill = "white", color = NA), legend.text = element_text(color = "#475569"), plot.margin = margin(15, 15, 15, 15) ) ACCENT <- "#e63946" BLUE <- "#3b82f6" AMBER <- "#f59e0b" GREEN <- "#10b981" MUTED <- "#94a3b8" ``` ## Visualizations ### 1. Temperature Structure: The Full Picture The heatmap below shows every day and every depth. The seasonal pulse is unmistakable at the surface, but it fades and compresses with depth -- the ocean's thermal inertia made visible. ```{r} #| label: heatmap #| fig-width: 12 #| fig-height: 7 #| warning: false p1 <- ocean_temp |> ggplot(aes(x = date, y = factor(sensor_depth_at_low_tide_m), fill = mean_temperature_degree_c)) + geom_tile(width = 1) + scale_fill_viridis_c( option = "turbo", name = "Temp (\u00B0C)", limits = c(-1, 22), breaks = seq(0, 20, by = 5) ) + scale_x_date( date_breaks = "1 year", date_labels = "%Y", expand = c(0.01, 0) ) + scale_y_discrete(limits = rev(c("2", "5", "10", "15", "20", "30", "40")), labels = function(x) paste0(x, "m")) + labs( title = "Surface waters swing 20\u00B0C across seasons -- deep layers barely move", subtitle = "Daily mean temperature by depth at Birchy Head, Nova Scotia (2018-2025)", caption = CAPTION, x = NULL, y = "Depth" ) + theme_fm + theme( legend.position = "right", panel.grid.major = element_blank() ) p1 ``` ### 2. Seasonal Profiles: How Deep Does Summer Reach? By averaging across years, we can isolate the seasonal cycle at each depth. Surface waters (2m) reach 18-20 degrees C in summer, while the 40m layer stays below 8 degrees C year-round. ```{r} #| label: seasonal-profiles #| fig-width: 12 #| fig-height: 6 #| warning: false seasonal <- ocean_temp |> group_by(month, sensor_depth_at_low_tide_m) |> summarise( mean_temp = mean(mean_temperature_degree_c, na.rm = TRUE), sd_temp = sd(mean_temperature_degree_c, na.rm = TRUE), .groups = "drop" ) depth_colors <- setNames( viridisLite::viridis(7, option = "viridis"), c("2", "5", "10", "15", "20", "30", "40") ) p2 <- seasonal |> ggplot(aes(x = month, y = mean_temp, color = factor(sensor_depth_at_low_tide_m), group = factor(sensor_depth_at_low_tide_m))) + geom_ribbon(aes(ymin = mean_temp - sd_temp, ymax = mean_temp + sd_temp, fill = factor(sensor_depth_at_low_tide_m)), alpha = 0.1, color = NA) + geom_line(linewidth = 1.2) + geom_point(size = 1.5) + scale_x_continuous(breaks = 1:12, labels = month.abb) + scale_color_manual(values = depth_colors, name = "Depth", labels = function(x) paste0(x, "m")) + scale_fill_manual(values = depth_colors, guide = "none") + labs( title = "Summer warmth barely penetrates below 20m -- the deep stays cold year-round", subtitle = "Monthly mean temperature by depth (ribbon = +/- 1 SD across years)", caption = CAPTION, x = NULL, y = "Temperature (\u00B0C)" ) + theme_fm + theme(legend.position = "right") p2 ``` ### 3. Year-over-Year Trend: Is the Surface Warming Faster? Comparing annual means by depth reveals whether warming is uniform or concentrated at the surface. ```{r} #| label: annual-trend #| fig-width: 12 #| fig-height: 6 #| warning: false annual <- ocean_temp |> filter(month >= 3 & month <= 11) |> group_by(year, sensor_depth_at_low_tide_m) |> summarise(mean_temp = mean(mean_temperature_degree_c, na.rm = TRUE), .groups = "drop") highlight_depths <- c(2, 20, 40) annual_highlight <- annual |> filter(sensor_depth_at_low_tide_m %in% highlight_depths) p3 <- annual |> ggplot(aes(x = year, y = mean_temp, group = factor(sensor_depth_at_low_tide_m))) + geom_line(color = MUTED, alpha = 0.3, linewidth = 0.8) + geom_line(data = annual_highlight, aes(color = factor(sensor_depth_at_low_tide_m)), linewidth = 1.4) + geom_point(data = annual_highlight, aes(color = factor(sensor_depth_at_low_tide_m)), size = 2.5) + scale_color_manual( values = c("2" = ACCENT, "20" = BLUE, "40" = "#440154"), name = "Depth", labels = function(x) paste0(x, "m") ) + scale_x_continuous(breaks = 2018:2025) + labs( title = "Surface and mid-depth temperatures show more year-to-year variability than the deep", subtitle = "Annual mean temperature (Mar-Nov) by depth | Grey = other depths", caption = CAPTION, x = NULL, y = "Temperature (\u00B0C)" ) + theme_fm + theme(legend.position = "right") p3 ``` ### 4. Variability by Depth: The Surface is Noisy, the Deep is Stable Standard deviation of daily temperatures reveals where the ocean is most variable -- and where it is remarkably stable. ```{r} #| label: variability #| fig-width: 12 #| fig-height: 5 #| warning: false variability <- ocean_temp |> group_by(sensor_depth_at_low_tide_m) |> summarise( mean_sd = mean(sd_temperature_degree_c, na.rm = TRUE), overall_sd = sd(mean_temperature_degree_c, na.rm = TRUE), .groups = "drop" ) |> mutate( depth_label = paste0(sensor_depth_at_low_tide_m, "m"), fill_color = if_else(overall_sd == max(overall_sd), ACCENT, BLUE) ) p4 <- variability |> ggplot(aes(x = fct_reorder(depth_label, sensor_depth_at_low_tide_m), y = overall_sd, fill = fill_color)) + geom_col(width = 0.6) + geom_text(aes(label = sprintf("%.1f\u00B0C", overall_sd)), vjust = -0.5, size = 4, color = "#475569") + scale_fill_identity() + scale_y_continuous(expand = expansion(mult = c(0, 0.15))) + labs( title = "Surface temperature varies 4x more than the deepest layer", subtitle = "Standard deviation of daily mean temperature across the full observation period", caption = CAPTION, x = "Depth", y = "Temperature SD (\u00B0C)" ) + theme_fm + theme(legend.position = "none") p4 ``` ## Key Findings 1. **The ocean has a thermal memory.** Surface waters (2m) swing 20+ degrees C across seasons, but the 40m layer remains nearly constant, buffered by thermal stratification. 2. **Summer warmth barely reaches 20m.** The seasonal pulse attenuates sharply with depth. Below 20m, the annual temperature range compresses to just a few degrees. 3. **Variability decreases monotonically with depth.** The standard deviation of daily temperatures at 2m is roughly 4x that at 40m, confirming that deeper waters act as thermal stabilizers. 4. **Year-to-year variation is most visible at the surface.** Inter-annual differences in annual mean temperature are larger at shallower depths, suggesting surface layers are more responsive to atmospheric forcing. *** _This post is part of the [TidyTuesday](https://github.com/rfordatascience/tidytuesday) weekly data visualization project._ :::{.callout-caution collapse="false" appearance="minimal" icon="false"} ## Disclaimer ::: {style="font-size: 0.85em; color: #64748b; line-height: 1.6;"} This analysis is for educational and practice purposes only. Data visualizations and interpretations are based on the provided dataset and may not represent complete or current information. Single-station observations cannot be generalized to broader ocean warming trends. ::: :::