Data art – Open data: Abbot Point (Bowen) 2013 hourly

Data art – Abbot Point (Bowen) 2013 hourly meteorological Queensland Government

Using this data set was perfect for the final outcome – it has all the values in the positive range so it was easy to work with. For this data art example I used the Date, time, wind direction and wind speed to make the animated art.


Every frame of the animation is mapped to the time series so in this example the time is based on hourly readings for a year which is 8761 frames which is almost 6 minutes long.  Interesting looking through datasets it’s not that common to find hourly readings unless it’s from a natural source.

In this data driven animation the wind direction controls the direction of the lines and also the hue, I’ve added a scale to the lines to reflect the wind speed but this is a subtle effect.


Interesting to see the patterns that emerge in presenting data in a non-spreadsheet or graph format.  You can see a pattern in the AM when the winds slow down and some really turbulent patches where the must have been a storm passing by.  I am very conscious of making this art rather than practical or logical – excel and data programs are far better at that – my intention is to use the streams of data to control the visual.

Dataset: air-quality-monitoring

Data art – 2011-2015—Thursday Island tide gauge

Cause why not? I like the idea that the data produced by this unassuming automatic tide gauge on Thursday Island can be used for art.  I had to extract the date and the value from the single column of data (example 120520110710  2.558) I used to use open refine for this but it is easy enough in excel these days to pull out the sections.


I could only use a section of days as it gets into the 1000’s of frames and weighs everything down. I kept this very simple the tide is represented in 3 ways – the circle the water height and the volume of the ocean – I figured if the tide was high the waves volume should drop and if low the volume increase.

The ebbing is perfect for data art – the natural variance you get in the data is something that would be hard to animate and have it look realistic.  The fluctuations from the standard give it some variance that make it interesting.. Higher than usual tides and patterns between the days.

Data: Thursday Island Tide gauge

Data art – SoE2015 Southern Oscillation Index

Another time series data driven animation. This time using the historic records of the SOI (Southern Oscillation Index) since 1876 as the input.  Because the SOI is a dance between positive and negative it was challenging to wrangle into some form of data art.  Originally I explored what visually represents the aspects of positive and negative – batteries, yin yang, hot/cold but when you apply it to a single object (rather than something that is time based) the data just produces oscillations without a time perspective.


This dataset only has 4 columns of data so it was easy to work with YEAR, MONTH, SOI, SIX MONTH MEAN.  I had to scale the values of the SOI as the max ranges seem to be around 50 which doesn’t highlight the fluctuations that well.

Due to the nature of the SOI being negative it also was perfect to control the volume of the wind as it goes negative and positive when the SOI is positive it turns the volume up.  It’s interesting to look at a dataset and see the values as potential for art rather than static values.  Inputs that require array values are perfect [RGB, audio, position…] you can plot the data in unique ways that might highlight a pattern that isn’t obvious.

Data: Historic SOI