To analyze historical data to investigate the impact and evidence of climate change across different geographical regions and time periods. With the data, I will perform:

• Trend Analysis
• Variability and Extremes
• Comparative Analysis
• 3d Visualizations

Data set from: https://www.kaggle.com/datasets/palinatx/mean-temperature-for-countries-by-year-2014-2022

• library(ggplot2)
• library(dplyr)
• library(plotly)
• library(readr)

The columns from the data set to work with include:

• Country
• Year
• Annual Mean
• 5-yr smooth
• Code

Initial data suggests that from 1901 to 2022, there is a positive linear relationship between Time and Temperature.

global_yearly_average <- data %>%
  group_by(Year) %>%
  summarize(Average_Temperature = mean(Annual.Mean, na.rm = TRUE))
  
a <- ggplot(global_yearly_average, aes(x = Year, y = Average_Temperature)) + 
  geom_line(color = "blue") + 
  labs(title = "Global Average Annual Mean Temperature (1901-2022)", 
    x = "Year", 
    y = "Temperature (°C)") + 
  geom_smooth(method = "lm", color = "red", se = FALSE) + 
  theme_minimal()

The thought process is now to check for anomalies. Picking countries to identify if there were any unusual temperature changes specific to where they may be located across the globe.

countries_of_interest <- c("USA", "RUS", "VNM", "BRA", "AUS", "SWE", "KEN")
filtered_data <- data %>%
  filter(Code %in% countries_of_interest)

# Calculate the average annual mean temperature
country_yearly_avg <- filtered_data %>%
  group_by(Year, Country) %>%
  summarise(Average_Temperature = mean(Annual.Mean, na.rm = TRUE), 
            .groups = 'drop')

# Plot the temperature trends
b = ggplot(country_yearly_avg, aes(x = Year, y = Average_Temperature, 
                                   color = Country)) +
  geom_line() +
  labs(title = "Comparison of Average Annual Mean Temperature",
       x = "Year",
       y = "Temperature (°C)") +
  theme_minimal() +
  scale_color_brewer(palette = "Dark2")  # palette

Dividing the data into two time periods, we can clearly see that there is a slight shift to the right for the later time period. This is to further illustrate how temperatures have risen over time.

## [1] "Correlation coefficient: 0.999112510324981"

## [1] "P-value: 0"

Threshold: 3 Standard Deviations

##               Country Year Annual.Mean X5.yr.smooth Code   z_score
## 1              Canada 1972       -6.81        -5.20  CAN -3.098106
## 2  Russian Federation 1902       -6.54        -5.54  RUS -3.065619
## 3  Russian Federation 1969       -6.50        -5.45  RUS -3.060806
## 4  Russian Federation 1912       -6.42        -5.73  RUS -3.051180
## 5              Canada 1917       -6.30        -5.55  CAN -3.036741
## 6  Russian Federation 1941       -6.22        -5.06  RUS -3.027115
## 7              Canada 1933       -6.16        -5.30  CAN -3.019896
## 8  Russian Federation 1966       -6.09        -5.33  RUS -3.011473
## 9              Canada 1950       -6.09        -5.23  CAN -3.011473
## 10             Canada 1907       -6.05        -5.50  CAN -3.006661
## 11 Russian Federation 1915       -6.03        -5.75  RUS -3.004254
## 12 Russian Federation 1987       -6.03        -4.64  RUS -3.004254
## 13 Russian Federation 1952       -6.00        -5.24  RUS -3.000644

The hypothesis is because Russia and Canada are closer to the North Pole, they had more extreme variations compared to a country closer to the equator, such as Mexico.

Showing the temperature by decade to paint a clear picture of the general trend occuring.

data$Decade <- floor(data$Year / 10) * 10  # Create new decade column

# Calculate average temperature for each decade
decadal_temps <- data %>%
  group_by(Decade) %>%
  summarise(Average_Temperature = mean(Annual.Mean, na.rm = TRUE)) 

# Plot temperature trends by decade
c = ggplot(decadal_temps, aes(x = Decade, y = Average_Temperature)) +
  geom_line() +
  geom_point() +
  theme_minimal() +
  labs(title = "Average Temperature by Decade",
       x = "Decade",
       y = "Average Temperature (°C)")

Using the Stanard Deviation by decade to determine variability as time progresses.

decadal_variability <- data %>%
  group_by(Decade) %>%
  summarise(Temperature_SD = sd(Annual.Mean, na.rm = TRUE))

# Plot the variability trends
d= ggplot(decadal_variability, aes(x = Decade, y = Temperature_SD)) +
  geom_line() +
  geom_point() +
  theme_minimal() +
  labs(title = "Temperature Variability by Decade",
       x = "Decade",
       y = "Standard Deviation of Temperature (°C)")

• From 1901-2022 the global average temperature has increased, as evidenced in mean, distribution analysis, and decade charts.
• Variability has decreased over time, with colder countries having the most variability
• The anomalies closer to the north pole have higher temperature fluctuations
• Since 1970, there has been an exponential increase in temperature with a decrease in variability.

The data suggests a sharp upward trend in global temperature as evidence of accelerated climate change, however, I must acknowledge that this is only about 120 years of data in Earth’s long lifespan. These trends highlight the urgent need for decisive action to mitigate the long term impacts of climate change.