DATA1001 Individual Project

Introduction

Client

My client is the Australian Climate Change Authority. It is part of the Australian government’s Department of Climate Change, Energy, the Environment and Water. It advises on climate change policy, assesses Australia’s progress toward its climate change goals, and advises the government on ways to reduce greenhouse gas emissions.

Problem Context

Australia has committed to reducing its greenhouse gas emissions to 43% below its 2005 levels by 2030 as they ratified the Paris agreement. However, Australia currently faces a difficulty in achieving this goal due to their growing population. Since, a growing population increases demand for electricity, transport, and energy, all of which make reductions in national emissions more difficult. Although, over the last few decades technology has become more efficient and cleaner, which would reduce per-capita CO2 emissions, any reductions could be offset by population growth. Therefore, this report will investigate how Australia’s total and per-capita emissions have changed between 2025 to 2023 as they have experienced population growth.

Data Context

The dataset used in this investigation is the CO2 and greenhouse gas emissions dataset with 50,191 observations and 79 variables, which was published by Our World in Data (OWID). This dataset contains historical data regarding carbon emissions, compiled from multiple statistical sources for multiple countries, is publicly available, and is updated regularly once new emissions statistics are available. Each row has a particular observation about a particular country in a particular year, while each column represents a variable like energy consumption, CO2 emissions, population, and per-capita emissions.

Data cleaning and integrity checks

For the cleaning and data validation section, firstly, the dataset was restricted to Australia from 2005 to 2023, as these years are the time period of the investigation. Then, the variables “year”, “CO2” measured in tonnes, “CO2 per capita” measured in tonnes per person, and “population”, measured in number of people, were selected, as they are required to investigate the changes in Australia’s annual CO2 and CO2 per capita emissions.

Next, the integrity checks were conducted. This involved checking for missing values, duplicate entries, consistent data types, and invalid data entries like negative values. This helps increase the accuracy of the analysis and the calculation of annual CO2 emissions.

Code for integrity checks

co2_data <- read.csv("~/Desktop/individual project/owid-co2-data.csv")

australia_co2 <- subset(
  co2_data,
  country == "Australia" & year >= 2005 & year <= 2023,
  select = c(year, population, co2, co2_per_capita)
)

australia_co2 <- australia_co2[order(australia_co2$year), ]

dim(australia_co2)

[1] 19  4

range(australia_co2$year)

[1] 2005 2023

colSums(is.na(australia_co2))

          year     population            co2 co2_per_capita 
             0              0              0              0 

sum(duplicated(australia_co2$year))

[1] 0

setdiff(2005:2023, australia_co2$year)

integer(0)

sum(australia_co2$population <= 0, na.rm = TRUE)

[1] 0

sum(australia_co2$co2 < 0, na.rm = TRUE)

[1] 0

sum(australia_co2$co2_per_capita < 0, na.rm = TRUE)

[1] 0

str(australia_co2)

'data.frame':   19 obs. of  4 variables:
 $ year          : int  2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 ...
 $ population    : num  20294490 20590565 20954545 21370595 21783014 ...
 $ co2           : num  386 392 399 405 407 ...
 $ co2_per_capita: num  19 19.1 19.1 18.9 18.7 ...

summary(australia_co2)

      year        population            co2        co2_per_capita 
 Min.   :2005   Min.   :20294490   Min.   :383.0   Min.   :14.48  
 1st Qu.:2010   1st Qu.:21962298   1st Qu.:392.5   1st Qu.:16.43  
 Median :2014   Median :23595599   Median :401.5   Median :16.87  
 Mean   :2014   Mean   :23540148   Mean   :400.5   Mean   :17.13  
 3rd Qu.:2018   3rd Qu.:25301207   3rd Qu.:406.6   3rd Qu.:18.49  
 Max.   :2023   Max.   :26451125   Max.   :415.8   Max.   :19.06  

summary(australia_co2)

      year        population            co2        co2_per_capita 
 Min.   :2005   Min.   :20294490   Min.   :383.0   Min.   :14.48  
 1st Qu.:2010   1st Qu.:21962298   1st Qu.:392.5   1st Qu.:16.43  
 Median :2014   Median :23595599   Median :401.5   Median :16.87  
 Mean   :2014   Mean   :23540148   Mean   :400.5   Mean   :17.13  
 3rd Qu.:2018   3rd Qu.:25301207   3rd Qu.:406.6   3rd Qu.:18.49  
 Max.   :2023   Max.   :26451125   Max.   :415.8   Max.   :19.06  

dim(co2_data)

[1] 50191    79

dim(australia_co2)

[1] 19  4

names(co2_data)

 [1] "country"                                  
 [2] "year"                                     
 [3] "iso_code"                                 
 [4] "population"                               
 [5] "gdp"                                      
 [6] "cement_co2"                               
 [7] "cement_co2_per_capita"                    
 [8] "co2"                                      
 [9] "co2_growth_abs"                           
[10] "co2_growth_prct"                          
[11] "co2_including_luc"                        
[12] "co2_including_luc_growth_abs"             
[13] "co2_including_luc_growth_prct"            
[14] "co2_including_luc_per_capita"             
[15] "co2_including_luc_per_gdp"                
[16] "co2_including_luc_per_unit_energy"        
[17] "co2_per_capita"                           
[18] "co2_per_gdp"                              
[19] "co2_per_unit_energy"                      
[20] "coal_co2"                                 
[21] "coal_co2_per_capita"                      
[22] "consumption_co2"                          
[23] "consumption_co2_per_capita"               
[24] "consumption_co2_per_gdp"                  
[25] "cumulative_cement_co2"                    
[26] "cumulative_co2"                           
[27] "cumulative_co2_including_luc"             
[28] "cumulative_coal_co2"                      
[29] "cumulative_flaring_co2"                   
[30] "cumulative_gas_co2"                       
[31] "cumulative_luc_co2"                       
[32] "cumulative_oil_co2"                       
[33] "cumulative_other_co2"                     
[34] "energy_per_capita"                        
[35] "energy_per_gdp"                           
[36] "flaring_co2"                              
[37] "flaring_co2_per_capita"                   
[38] "gas_co2"                                  
[39] "gas_co2_per_capita"                       
[40] "ghg_excluding_lucf_per_capita"            
[41] "ghg_per_capita"                           
[42] "land_use_change_co2"                      
[43] "land_use_change_co2_per_capita"           
[44] "methane"                                  
[45] "methane_per_capita"                       
[46] "nitrous_oxide"                            
[47] "nitrous_oxide_per_capita"                 
[48] "oil_co2"                                  
[49] "oil_co2_per_capita"                       
[50] "other_co2_per_capita"                     
[51] "other_industry_co2"                       
[52] "primary_energy_consumption"               
[53] "share_global_cement_co2"                  
[54] "share_global_co2"                         
[55] "share_global_co2_including_luc"           
[56] "share_global_coal_co2"                    
[57] "share_global_cumulative_cement_co2"       
[58] "share_global_cumulative_co2"              
[59] "share_global_cumulative_co2_including_luc"
[60] "share_global_cumulative_coal_co2"         
[61] "share_global_cumulative_flaring_co2"      
[62] "share_global_cumulative_gas_co2"          
[63] "share_global_cumulative_luc_co2"          
[64] "share_global_cumulative_oil_co2"          
[65] "share_global_cumulative_other_co2"        
[66] "share_global_flaring_co2"                 
[67] "share_global_gas_co2"                     
[68] "share_global_luc_co2"                     
[69] "share_global_oil_co2"                     
[70] "share_global_other_co2"                   
[71] "share_of_temperature_change_from_ghg"     
[72] "temperature_change_from_ch4"              
[73] "temperature_change_from_co2"              
[74] "temperature_change_from_ghg"              
[75] "temperature_change_from_n2o"              
[76] "total_ghg"                                
[77] "total_ghg_excluding_lucf"                 
[78] "trade_co2"                                
[79] "trade_co2_share"                          

range(australia_co2$year)

[1] 2005 2023

colSums(is.na(australia_co2))

          year     population            co2 co2_per_capita 
             0              0              0              0 

sum(duplicated(australia_co2$year))

[1] 0

subset(australia_co2, co2 < 0)

[1] year           population     co2            co2_per_capita
<0 rows> (or 0-length row.names)

These checks confirm that there are no missing, duplicate or negative values. Therefore, no observations have to be omitted.

Evidence

plot(australia_co2$year,
     australia_co2$population / 1000000,
     type = "l",
     col = "steelblue",
     lwd = 2,
     xlab = "Year",
     ylab = "Population (Millions)",
     main = "Graph 1: Australia's Population (2005–2023)")

Graph 1 shows that Australia’s population increased by 30.34%, from 20.29 million to 26.45 million, during the period of the investigation. This allows us to analyse the changes in CO2 emissions against the population growth.

australia_co2$total_change <- 
  (australia_co2$co2 / australia_co2$co2[1] - 1) * 100

australia_co2$per_capita_change <-
  (australia_co2$co2_per_capita /
     australia_co2$co2_per_capita[1] - 1) * 100

plot(australia_co2$year,
     australia_co2$total_change,
     type = "l",
     col = "steelblue",
     lwd = 2,
     ylim = range(c(australia_co2$total_change,
                    australia_co2$per_capita_change)),
     xlab = "Year",
     ylab = "Change Since 2005 (%)",
     main = "Graph 2: Changes in Australia's CO2 Emissions")

lines(australia_co2$year,
      australia_co2$per_capita_change,
      col = "tomato",
      lwd = 2)

abline(h = 0, lty = 2)

legend("topright",
       legend = c("Total CO2", "CO2 Per Capita"),
       col = c("steelblue", "tomato"),
       lwd = 2,
       bty = "n")

Graph 2 visualises the percentage change in total and per-capita CO2 emissions from 2005 to 2023. Per-capita emissions fell by 23.92%, from 19.029 tonnes to 14.478 tonnes per person, while total CO2 emissions fell by only 0.83%, from 386.176 million tonnes to 382.972 million tonnes.

So, despite a considerable decrease in per-capita CO2 emissions, total CO2 emissions remained relatively constant. This illustrates that reductions in per-capita CO2 emissions do not necessarily lead to reductions in total CO2 emissions at a national level.

plot(australia_co2$population / 1000000,
     australia_co2$co2,
     pch = 19,
     col = "steelblue",
     xlab = "Population (Millions)",
     ylab = "CO2 Emissions (Million Tonnes)",
     main = "Graph 3: Population vs Total CO2 Emissions")

model <- lm(co2 ~ population, data = australia_co2)

abline(model, col = "tomato", lwd = 2)

text(australia_co2$population / 1000000,
     australia_co2$co2,
     labels = australia_co2$year,
     pos = 3,
     cex = 0.6)

Graph 3 illustrates the relationship between total CO2 emissions and population. However, there is a very weak negative linear relationship between the 2 variables, with a Pearson correlation coefficient of -0.024, which is supported by the nearly horizontal regression line. This shows us that while Australia’s population continued to grow steadily, there were many fluctuations in total CO2 levels, demonstrating that other factors like changes in technology, energy sources, economic activity, or efficiency may have influenced total CO2 emissions, and population was not the main factor behind the change in total CO2 emissions.

There was a substantial decrease in total CO2 emissions from 2019 to 2020, despite there being an increase in the Australian population. This decrease coincided with the COVID-19 pandemic, which caused a decrease in demand for transport and other energy-demanding activities due to restrictions set by the Australian government. This is an observation that is supported by the Australian Government’s State of the Environment 2021 report.

Moving on, the correlation analysis only examines linear relationships, so more complex patterns may have been overlooked. The period during which the COVID-19 restrictions took place coincides with the period of the investigation, making it hard to distinguish between temporary reductions in CO2 emissions as a result of the imposed restrictions and reductions due to improvements in infrastructure. Lastly, since the dataset only contains fossil fuel and industrial CO2 emissions, it makes it more difficult to assess Australia’s overall climate, as other greenhouse gases are not included.

start <- australia_co2[1, ]
end <- australia_co2[nrow(australia_co2), ]

results <- data.frame(
  Variable = c("Population", "Total CO2", "CO2 per capita"),
  Value_2005 = c(start$population, start$co2,
                 start$co2_per_capita),
  Value_2023 = c(end$population, end$co2,
                 end$co2_per_capita)
)

results$Percentage_Change <- round(
  (results$Value_2023 / results$Value_2005 - 1) * 100,
  2
)

print(results)

        Variable   Value_2005   Value_2023 Percentage_Change
1     Population 20294490.000 26451125.000             30.34
2      Total CO2      386.176      382.972             -0.83
3 CO2 per capita       19.029       14.478            -23.92

cor(australia_co2$population, australia_co2$co2)

[1] -0.02384948

Recommendation

The Climate Change Authority should investigate both per-capita and total emissions together when finding ways to reduce emissions and improve the climate. They should also investigate household and other sector-specific emissions to find ways to increase the rate of total emissions decrease as their population continues to grow.

AI usage statement

OpenAI’s ChatGPT (GPT-6, https://chatgpt.com/) was used to assist with refining the research question and troubleshooting R code for integrity checks and visualisations. The analysis was conducted using the Our World in Data dataset, with results reviewed in RStudio. The grammarly google extension was also used to fix any grammar errors

Acknowledgements

I referred to the exemplar for the individual project uploaded by my tutors at this link for my formatting: https://canvas.sydney.edu.au/courses/74355/pages/project-briefs

References

Climate Change Authority. (n.d.). About the Authority. https://www.climatechangeauthority.gov.au/about-cca

Commonwealth of Australia. (2022). Climate Change Act 2022 (Cth). https://www.legislation.gov.au/C2022A00037/latest/text

Department of Climate Change, Energy, the Environment and Water. (2022). Greenhouse gases. In Australia state of the environment 2021. https://soe.dcceew.gov.au/climate/pressures/greenhouse-gases

Our World in Data. (n.d.). CO₂ and greenhouse gas emissions dataset [Data set]. GitHub. https://github.com/owid/co2-data

Ritchie, H. (2022, February 9). CO₂ emissions dataset: Our sources and methods. Our World in Data. https://ourworldindata.org/co2-dataset-sources

United Nations Framework Convention on Climate Change. (n.d.). The Paris Agreement. https://unfccc.int/process-and-meetings/the-paris-agreement

References cited using APA for the data source and any external sources. In-text citations used when appropriate.

Appendix

If you utilise hypothesis testing in your evidence section, document the HATPC process for your hypothesis test here.