Earthquakes Tsunami Correlations

Introduction

In this project, developed over the course of the summer semester for the course ‘Programmiertes Entwerfen 2’ (Programmed Design 2), I focused on visualizing data from two datasets to determine possible correlations between earthquakes and tsunamis. The datasets are from Kaggle.com and contain comprehensive information on earthquakes and tsunamis worldwide. My goal was to find out whether tsunamis were triggered by earthquakes and to visually represent these events.

abgabe.pdf

In the following chapters there will be all kind of visualisations beeing displayed.

In this section, you can see the landing page with an overview of all data points from both datasets. Blue points represent tsunamis, and red points represent earthquakes. This view serves as a first impression for the user.

When the user opens the control panel, they have various options to manipulate the view, allowing them to reduce the total number of points and focus on a smaller group.

This is an example of a visual representation of all earthquake events contained in the dataset when tsunami events are disabled. In this view, the contours of the individual tectonic plates are clearly visible.

Visualization of the view when all earthquake data points are hidden. It is clear to see that the tsunami dataset is smaller and less extensive.

This view demonstrates the disabled buttons in the panel, indicating that the tsunami points are turned off and only the earthquakes with the highest magnitude are activated. This allows the user to see at a glance where significant earthquakes have occurred on Earth.

This visualization shows all earthquake data mapped on a 3D globe, taking into account magnitude and depth values. The model was created using Three.js.

The model can be rotated thanks to the THREE.OrbitControls function provided by Three.js. This allows the user to gain a completely new perspective on the data after previously only experiencing it in the 2D model.

The points are loaded again from the Earthquakes.data dataset and mapped onto the globe based on the longitude and latitude values. Additionally, a magnitude filter function is used again to display earthquakes of different strengths.

The zoom function allows users to interactively explore the depth of earthquakes by zooming into the globe and making the depth visible.

Conclusion

Now I come to my own results, which have caused me problems during this project, but also to the interesting findings that emerged in the process. The different formatting of the individual datasets made it very difficult for me at the beginning to verify the completeness and accuracy of the datasets, as it was very important to me that the visualization was based on actually logical data. One difficulty was that the datasets were of different sizes: there were 24,000 earthquake events but only 2,000 tsunami events, and I was concerned that I would not find any data that could be confirmed as related.

My initial hypothesis that tsunamis are correlated with earthquakes was confirmed for over 400 events, as both the geographical locations and the temporal data of the events could be unambiguously matched. There were partially incomplete intensity entries in the tsunami dataset, but these were identified and excluded from the comparison, ensuring that all data points visible in this visualization consist of complete elements from the datasets. A clear correlation between earthquakes and tsunamis from two different datasets was established. Additionally, it is worth noting that only significantly strong earthquakes generally triggered tsunamis. In this dataset, less than 3% of earthquakes with a magnitude less than 6 triggered tsunamis.