Introduction
Hannah (2004) noted that sea levels have been increasing at the ‘average rate of 1.6 mm/yr over the last 100 year’.
Rahmstrorf (2012) argues that ‘losing just 1% of the present continental ice would raise sea level globally by about 75cm’.
Parker and Ollier (2019) stated that there wasn’t a coherent long term trend that can be observed from sea levels specifically in the Pacific and Indian ocean.
Duvat (2019) meta-analysis of 702 islands in the Indian and Pacific ocean found that only 11.4% of islands contracted while, 88.6% of islands remained consistent or enlarged over the past decade (Duvat, 2019).
Research Question
Old question - Is there a similarity between sea levels in islands within the Pacific Ocean?
After receiving the feedback, we adapted our original question as it lacked direction.
New question - Is there a relationship between tide gauge data and greenhouse gases in New Zealand over a decade?
Overview of the Data
In this project 8 datasets was used - all data obtained from the Wellington, New Zealand area. Most of variables are numeric and continuous.
The dataset including the New Zealand tide gauge data stems from the year 1944 to 2019 - including two variables the year and recorded tide gauge data.
The 7 greenhouses gases dataset show the site, year (1999 to 2019), month and value of the corresponding gas found in the ocean. The 7 greenhouses gases names in the dataset are the following:
Methane incl. C13/C12 in methane
Carbon Dioxide incl. C13/C12 in carbon dioxide and O18/016 in carbon dioxide
Carbon Monoxide
Nitrous Oxide
Sulfur Hexafluoride.
Methodology
We used R and R studio to conduct the analysis for the secondary research selected
K-means clustering to observe the trends in the data chosen
Classification - classifying the greenhouse gases
Linear Regression Model - to predict and compare the relationship between the tide gauge data and the greenhouse gases
Exploratory Analysis
EDA: Graphical Exploration
New Zealand Recorded Tide Gauge - Scatter plot that shows the annual tide gauge data - the plots are concentrated in the top portion of the plot with the line showing the slight positive trend present in this graph suggesting that sea levels have been slowly rising over the 1944 to 2019.
Conclusion
The project was limited by the fact that the datasets were sourced from multiple different datasets. This limited the range of variables that could be employed in the methods. To improve on this weakness, a bigger emphasis on finding multiple datasets from one source would have been helpful.
Additionally, the covariates were not scaled before the linear regression model, which could affect the accuracy of the model output, especially the coefficient output, which is sensitive to changes.
Furthermore, supplementary developed variable selection could have been explored to further assess the relationship between the coastal stations and to view if there are any explanatory outliers.
The weakness of the study is that the number of years for the greenhouse gases is limited and only measures a decade. Additionally, there are separate datasets for the greenhouse gases, which means that it takes more time to explore the relationship between the variables. On the other hand, the study has several strengths, such as having varied variables that allow researchers to pinpoint the relationship between greenhouse gases and the tide gauge data.
Final Sea Level Project Compressed.pdf