Summary Statistics

About the Dataset

About This Dashboard

This dashboard presents a comprehensive analysis of climate variables recorded monthly from January 2020 to May 2024 (53 records).

Variables Included

• Temperature: Average, Maximum, and Minimum (°C)
• Precipitation (mm) and Humidity (%)
• Wind Speed (m/s) and Solar Irradiance (W/m²)
• CO₂ Concentration (ppm) and Particulate Matter (µg/m³)
• Sea Surface Temperature (°C)
• ENSO Index, Vegetation Index, Urbanization Index

Data Cleaning

Missing values appeared as blank cells, the strings Unknown and NAN, and the sentinel value 99999. All were converted to NA on import. Missing values were then imputed using the mean of the same calendar month across available years.

Monthly Average Temperature Over Time

Temperature Range (Max – Min) by Month

Seasonal Temperature Distribution

Monthly Precipitation Over Time

Humidity Distribution by Season

CO₂ Concentration Over Time

Particulate Matter Distribution

CO₂ vs Sea Surface Temperature

Temperature Findings

Overall Temperature Patterns

Monthly average temperatures fluctuate considerably across the 2020–2024 period, with no strong consistent seasonal signal, suggesting either high variability in the data or that this station spans a climate zone with irregular seasonality.

Temperature Range

Some calendar months show notably wider Max–Min ranges than others. October in particular stands out with a large average daily range, which may indicate transitional weather patterns typical of autumn months.

Seasonal Distributions

The boxplot reveals overlapping temperature distributions across seasons, with Summer showing the highest median average temperatures as expected. Winter shows the greatest spread, suggesting more volatile conditions during colder months.

Precipitation & Humidity Findings

Precipitation Trends

Monthly precipitation varies widely with no clear long-term upward or downward trend over the study period. The loess smoothing line remains relatively flat, suggesting precipitation levels have been broadly stable from 2020 to 2024.

Humidity by Season

The violin plots reveal that Summer and Autumn tend to have more concentrated humidity distributions, while Winter and Spring show greater variability. This aligns with the expectation that warmer months maintain more consistent moisture levels.

CO₂ & Air Quality Findings

CO₂ Concentration

CO₂ levels show a modest but visible upward trend over the study period, consistent with global observations of rising atmospheric carbon dioxide. The loess confidence band widens toward 2024, reflecting fewer data points and greater uncertainty at the tail of the series.

Particulate Matter by Season

The distribution of particulate matter differs notably across seasons. Winter months tend to show higher PM concentrations, which is consistent with increased heating activity and reduced atmospheric dispersion during colder, calmer conditions.

CO₂ vs Sea Surface Temperature

The scatter plot suggests a weak positive relationship between CO₂ concentration and sea surface temperature, as CO₂ increases, SST shows a slight tendency to rise as well. This is consistent with the broader climate science understanding of CO₂ as a driver of ocean warming, though the short time span and limited sample size mean this relationship should be interpreted cautiously.

General Caveats

Data Quality

A meaningful number of missing values were present across several variables, appearing as blank cells, the strings Unknown and NAN, and the sentinel value 99999. These were all converted to NA and imputed using monthly means, but one that smooths over true variability and may understate extremes.

Sample Size

With only 53 monthly records spanning just over four years, all interpretations should be treated as exploratory rather than confirmatory. Longer time series would be needed to draw statistically robust conclusions about trends.

Causality

Associations visible in the plots, such as CO₂ vs SST, do not imply causation. Many variables in this dataset are likely correlated with common external drivers (e.g. season, year) rather than directly influencing one another.