Introduction

Marine ecosystems are strongly affected by climate variables such as sea surface temperature (SST),
pH levels, and marine heatwaves. Corals are especially sensitive, showing stress through bleaching events.

This project uses a Kaggle dataset containing:

Research Questions

  1. How do SST and pH levels vary across regions and time?
  2. How are SST and pH related to coral bleaching severity?
  3. Are species counts affected by climate variables?
  4. How do marine heatwaves influence bleaching?
  5. Can these relationships be visualized in 2D and 3D?

Loading Data

data <- read_csv("realistic_ocean_climate_dataset.csv")
## Rows: 500 Columns: 9
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr  (2): Location, Bleaching Severity
## dbl  (5): Latitude, Longitude, SST (°C), pH Level, Species Observed
## lgl  (1): Marine Heatwave
## date (1): Date
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
head(data)
## # A tibble: 6 × 9
##   Date       Location           Latitude Longitude `SST (°C)` `pH Level`
##   <date>     <chr>                 <dbl>     <dbl>      <dbl>      <dbl>
## 1 2015-01-01 Red Sea              20.0        38.5       29.5       8.11
## 2 2015-01-07 Great Barrier Reef  -18.3       148.        29.6       8.00
## 3 2015-01-14 Caribbean Sea        15.0       -75.0       28.9       7.95
## 4 2015-01-20 Great Barrier Reef  -18.3       148.        29.0       8.00
## 5 2015-01-27 Galápagos            -0.880     -91.0       28.6       7.98
## 6 2015-02-02 Red Sea              20.0        38.4       29.1       8.01
## # ℹ 3 more variables: `Bleaching Severity` <chr>, `Species Observed` <dbl>,
## #   `Marine Heatwave` <lgl>

Data Cleaning + Preparation

data_clean <- data %>%
  mutate(
    Date = ymd(Date),
    BleachingSeverity = factor(`Bleaching Severity`,
                               levels = c("None","Low","Medium","High")),
    MarineHeatwave = as.logical(`Marine Heatwave`)
  ) %>%
  drop_na()
summary(data_clean)
##       Date              Location            Latitude          Longitude      
##  Min.   :2015-01-01   Length:500         Min.   :-18.3830   Min.   :-155.72  
##  1st Qu.:2017-04-01   Class :character   1st Qu.: -0.9329   1st Qu.: -90.93  
##  Median :2019-07-01   Mode  :character   Median : 10.0188   Median :  38.52  
##  Mean   :2019-07-01                      Mean   :  6.7284   Mean   :  11.93  
##  3rd Qu.:2021-09-29                      3rd Qu.: 19.8910   3rd Qu.: 114.99  
##  Max.   :2023-12-31                      Max.   : 20.1220   Max.   : 147.83  
##     SST (°C)        pH Level     Bleaching Severity Species Observed
##  Min.   :23.64   Min.   :7.872   Length:500         Min.   : 54.0   
##  1st Qu.:27.53   1st Qu.:8.011   Class :character   1st Qu.:107.8   
##  Median :28.52   Median :8.052   Mode  :character   Median :120.0   
##  Mean   :28.54   Mean   :8.050                      Mean   :120.5   
##  3rd Qu.:29.45   3rd Qu.:8.085                      3rd Qu.:133.2   
##  Max.   :33.21   Max.   :8.195                      Max.   :171.0   
##  Marine Heatwave BleachingSeverity MarineHeatwave 
##  Mode :logical   None  :150        Mode :logical  
##  FALSE:427       Low   :132        FALSE:427      
##  TRUE :73        Medium:130        TRUE :73       
##                  High  : 88                       
##                                                   
##

Exploratory Data Analysis

1. SST Over Time (Line Plot)

ggplot(data_clean, aes(Date, `SST (°C)`, color = Location)) +
  geom_line(linewidth = 1) +
  theme_minimal() +
  labs(title = "Sea Surface Temperature Over Time")

2. pH Level Over Time

ggplot(data_clean, aes(Date, `pH Level`, color = Location)) +
  geom_line(linewidth = 1) +
  theme_minimal() +
  labs(title = "Ocean pH Over Time")

3. Species Observed Over Time

ggplot(data_clean, aes(Date, `Species Observed`, color = Location)) +
  geom_line(linewidth = 1) +
  theme_minimal() +
  labs(title = "Species Richness Trends Over Time")

4. Bleaching Severity Distribution

ggplot(data_clean, aes(BleachingSeverity, fill = Location)) +
  geom_bar(position = "dodge") +
  theme_minimal() +
  labs(title = "Distribution of Coral Bleaching Severity")

Relationship Between Climate & Bleaching

5. SST vs pH by Bleaching Severity (colored)

ggplot(data_clean, aes(`SST (°C)`, `pH Level`, color = BleachingSeverity)) +
  geom_point(size = 3, alpha = 0.8) +
  theme_minimal() +
  labs(title = "SST vs pH Level Colored by Bleaching Severity")

6. Three-Variable Bubble Plot (SST–pH–Species)

ggplot(data_clean, aes(
  x = `SST (°C)`,
  y = `pH Level`,
  size = `Species Observed`,
  color = BleachingSeverity
)) +
  geom_point(alpha = 0.6) +
  scale_size(range = c(2, 12)) +
  theme_minimal() +
  labs(title = "3D Bubble Plot: SST vs pH vs Species Observed")

Advanced Visualizations

7. 3D Scatter Plot (SST, pH, Species)

plot_ly(
  data_clean,
  x = ~`SST (°C)`,
  y = ~`pH Level`,
  z = ~`Species Observed`,
  type = "scatter3d",
  mode = "markers",
  color = ~BleachingSeverity,
  marker = list(size = 5)
)

8. 3D Geographic Map (Lat, Long, Species, SST Color)

plot_ly(
  data_clean,
  x = ~Longitude,
  y = ~Latitude,
  z = ~`Species Observed`,
  type = "scatter3d",
  mode = "markers",
  color = ~`SST (°C)`,
  marker = list(size = 5),
  hoverinfo = "text",
  text = ~paste(
    "Location:", Location,
    "<br>SST:", `SST (°C)`,"°C",
    "<br>pH:", `pH Level`,
    "<br>Bleaching:", BleachingSeverity
  )
) %>%
  layout(
    title = "3D Geographic Map of Marine Conditions",
    scene = list(
      xaxis=list(title="Longitude"),
      yaxis=list(title="Latitude"),
      zaxis=list(title="Species Observed")
    )
  )

Marine Heatwave Effects

9. Heatwaves vs Bleaching Severity

ggplot(data_clean, aes(MarineHeatwave, fill = BleachingSeverity)) +
  geom_bar(position = "fill") +
  theme_minimal() +
  labs(title = "Impact of Marine Heatwaves on Coral Bleaching",
       y = "Proportion")

Correlation Matrix

10. Numeric variable correlations

num_data <- data_clean %>%
  select(`SST (°C)`, `pH Level`, `Species Observed`)

corrplot(cor(num_data), method = "color", addCoef.col = "black")

Simple Regression: Predicting Species Count from SST

model <- lm(`Species Observed` ~ `SST (°C)`, data = data_clean)
summary(model)
## 
## Call:
## lm(formula = `Species Observed` ~ `SST (°C)`, data = data_clean)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -46.867 -10.198  -0.615  10.159  56.701 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 399.8114    13.4665   29.69   <2e-16 ***
## `SST (°C)`   -9.7886     0.4713  -20.77   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 15 on 498 degrees of freedom
## Multiple R-squared:  0.4641, Adjusted R-squared:  0.4631 
## F-statistic: 431.4 on 1 and 498 DF,  p-value: < 2.2e-16
ggplot(data_clean, aes(`SST (°C)`, `Species Observed`)) +
  geom_point(color = "blue") +
  geom_smooth(method = "lm", se = FALSE, color = "red") +
  theme_minimal() +
  labs(title = "Regression: SST vs Species Observed")
## `geom_smooth()` using formula = 'y ~ x'

Conclusion

From the analysis:

This study supports well-documented scientific evidence that warming and acidification are the largest threats to coral ecosystems globally.