Earthquakes visualization

How our visualization works?

The methodology used for this visualization is based on eye movement plots.Where eye movement traces are transformed from a sequence of points into a sequence of lines preserving information of duration. Such mapping linearizes the corresponding visualization and allows a stacking of the region event plots.
The biggest challenge linearizing earthquake events is comparable visual expressions for each of the locations. For different event to be placed in the timeline, each location should have a unique linear value. This allows visualization tool to represent events that appeared at the same timeslot. However, there are hundreds or even thousands of earthquakes each day, therefore, density of linearized events is expressed by color gradients from red, which is usual density, to yellow - a significant density.

Location stamps

As a first step moving toward earthquake visualization, location points are transformed from 3D to 2D. This is done using stacked line-based, splatted diagram technology which is mostly used for eye movement tracking. Here line is drawn considering location coordinates and duration in time. By applying usual transformation, result do not satisfy our case because of spherical 3D data alignment. The distance between points is different in every latitude, therefore, northern and southern values should be much more alike. Furthermore, it is a circle in shape which means that sides of the grid should become as one. As we see in the image bellow, none of the requirements are met as lines are in diagonal symetry when using transformation used for eye movement tracking.

Location lines in the grid are moving from the bottom to the top and size is changing respectively to the North - South geographical position. This representation is necessary for the poles to distinguish. As data points at poles are much closer to each other than at the Equator, they are made to be more like to each another.
Meanwhile slope of the line depends on East - West geographical position. There are 90 degrees lines at the Prime meridian and angle slopes to the right toward East side, to the left - toward West side. This is made for the sides of the map to be same due to the spherical shape of the Earth.

The transformation used for each line l (sequence of two points) from each earthquake e (lognitude, latitude and time)
$e_{i} := (long_{i}, lat_{i}, t_{i}) , l_{i} := [(x_{i,1}, y_{1,i})(x_{2,i},y_{2,i})]$
is the following:

\alpha_{i} = \frac{long_{i}+180}{2}

cx_{i} = t_{i}

cy_{i} = \frac{lat_{i} + 90}{180}

\delta_{i} = \delta * cy + C

dx_{i} = \delta_{i} \cdot cos(\frac{\alpha_i}{180}\pi)

dy_{i} = \delta_{i} \cdot sin(\frac{\alpha_i}{180}\pi)

which leads us to the x's and y's needed for the line:

x_{i, 1} = cx - dx_{i}

y_{i, 1} = ct - dy_{i}

x_{i, 2} = cx + dx_{i}

y_{i, 2} = ct + dy_{i}

For the tool to work, grid of the events is pressed into one line so the frequencies and patterns would be visible in global perspective and not regional.
The image bellow represents 40 different earthquake data lines while using stacked line-based, splatted diagram technology. It is clearly seen that L2 is a lot more seismically active zone. The lines with the same slope in L2 represent events in Turkey area. Meanwhile, L4 shows earthquakes further from each other which have happened in Indonesia, Chile and Myanmar. The visualization tool shows intensity of earthquake attacks in a particular area and find patterns of event sequence. With a large amount of data, lines are splatted to clear visual clutter and make the diagram understandable.

However, for the exploration and research reasons, there are filters and histograms provided. By filtering location, magnitude or timeslot, the visualization like the one above may appear. Otherwise, all the data lines will be splatted together and only densities visualized in a understandable way.