Température de l’eau au cap d’Antibes

Introduction

Distribution des relevés de température de l’eau au cap d’Antibes (Celsius), par mois et par année.

Résultats

En été l’ouest du cap est légèrement plus chaud que l’est du cap.
En hiver l’est du cap semble légèrement moins froid que l’ouest du cap.
Limite: ces différences ne sont pas très significatives.

Visualisation

Code

Import

library(tidyverse)
library(ggridges)
library(viridis)
library(slider)
library(patchwork)
library(paletteer)

date_caption <- "31 décembre 2024"
source("tools/themes.R") # themes

source_est <- read_csv("posts/2025-01-23/data/antibes-eau-temperature-est-cap.csv")
source_ouest <- read_csv("posts/2025-01-23/data/antibes-eau-temperature-ouest-cap.csv")

df_est <-source_est
df_ouest <- source_ouest

Inspect

head(df_est)
names(df_est)
glimpse(df_est)
summary(df_ouest)

Clean

# standardize and bind datasets
fn_clean <- function(df, name) {
  # standardize cols
  df <-
    df %>% 
    select(-1, -2) %>%                  # remove useless cols
    rename_with(tolower) %>%                      # lower case variable names
    rename_with(~ str_squish(.)) %>%             
    rename_with(~ str_replace_all(.,"annee ", ""))
  
  # adding useful cols
  df <-
    df %>%
    mutate(region = name)
  
  # pivot longer
  df<-
    df %>%
    pivot_longer(cols = 2:6, names_to = "annee", values_to = "temperature")
  
  return(df)
}

df_ouest_clean <- fn_clean(df_ouest, "ouest")
df_est_clean <- fn_clean(df_est, "est")

# bind
df_bind <- bind_rows(df_est_clean, df_ouest_clean)

# Manage dates
df_bind_clean <-
  df_bind %>% 
    separate(jour, c("day", "month"), "-") %>%
    mutate(month = str_replace_all(month, "\\.", ""),
           month = str_to_title(month)) # Need to escape the dot (.) here
  
# check  
unique(df_bind_clean$month) # month col still needs cleaning 

df_bind_clean <-
  df_bind_clean %>%
    mutate(month = recode(month,
                          "Janv" = "Jan",
                          "Fevr" = "Feb",
                          "Febr" = "Feb",
                          "Févr" = "Feb",
                          "Mars" = "Mar", 
                          "Avr" = "Apr",
                          "Mai" = "May", 
                          "Juin" = "Jun",
                          "Juil" = "Jul",
                          "Août" = "Aug",
                          "Sept" = "Sep",
                          "Oct" = "Oct",
                          "Déc" = "Dec"
            ))            
 
# add date col 
df_bind_clean <-
  df_bind_clean %>% 
    mutate(date = dmy(paste(day, month, annee, sep = "-"))) %>%
    select(region, date, temperature, year = annee, month, day) %>% # reorder
    drop_na() %>%                                                   # drop NA
    arrange(date)

Plot

# Line graph 
plot_region_year <-
df_bind_clean %>%
  drop_na() %>%
  mutate(date = update(date, year = 2024)) %>% # plot on synchronous x-axis, /!\ chose a leap year
  ggplot(aes(x = date, y = temperature, group = year)) + 
  geom_line(aes(color = region)) +
  scale_x_date(date_breaks = "4 month", date_labels = "%b") +
  facet_wrap(~region+year, nrow = 2) 
 plot_region_year

     df_bind_clean_plot <-
     df_bind_clean %>%
       filter(year > 2020) %>%
       mutate(region = recode(region, "est" = "Est du Cap",
                                      "ouest" = "Ouest du Cap")) %>%
       mutate(month = factor(month, 
                             levels = c("Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec")))
     
     
    # Ridges temperatures per year
    plot_ridges_year <-
      df_bind_clean_plot %>%
      group_by(year) %>%
      ggplot(aes(x = temperature, y = year, fill = after_stat(x))) +
      geom_density_ridges_gradient(color = darkblue) +  
      scale_fill_gradientn(colors = custom_palette_water_temp) +
      coord_cartesian(xlim=c(5,34)) +
      scale_x_continuous(labels = scales::label_dollar(prefix = "", suffix = "°c"))  
    plot_ridges_year

    # Ridges for temperature by month
    plot_ridges_month <-
      df_bind_clean_plot %>%
      group_by(month) %>%
      ggplot(aes(x = temperature, y = fct_rev(month), fill = after_stat(x))) +
      geom_density_ridges_gradient(color = darkblue) +  
      scale_fill_gradientn(colors = custom_palette_water_temp) +
      coord_cartesian(xlim=c(9,31)) +
      scale_x_continuous(labels = scales::label_dollar(prefix = "", suffix = "°c"))
    plot_ridges_month

Render

plot_ridges_year_final <-  
plot_ridges_year + 
    facet_wrap(~region, ncol =1, strip.position = "top", scales = "free") +
   labs(
     title = "Baignade : l'année 2024 légèrement plus chaude que les précédentes",
     subtitle = "Distribution des revelés de température de l'eau au cap d'Antibes (Celsius), par années",
     x = "Température (Celsius)",
     caption = social_caption2) + tt8
  
plot_ridges_month_final <-    
plot_ridges_month + 
  facet_wrap(~region, ncol =1, strip.position = "top", scales = "free") +
  labs(
    title = "Antibes Juan-Les-Pins baignable une grande partie de l'année",
    subtitle = "Distribution des revelés de température de l'eau au cap d'Antibes (Celsius), par mois",
    x = "Température (Celsius)",
    caption = social_caption2) + tt8

plot_ridges_month_final  

plot_ridges_year_final

Note

Smoothing with a moving average method using slide_dbl() will not plot correctly because there are irregular missing dates in the data. 
Fixing this would require filling missing values with tidyr::complete() and then filling upwards the NA values or computing them, or using library(zoo) for something like rollmean(df_bind_clean$temperature, 3, fill = "extend"). 
We are chosing to work with actual values only for this analysis.

Source

Température de l’eau de mer à Antibes Juan-les-Pins, disponible sur data.gouv.fr. Mise à jour du 31 décembre 2024.