How the Science Equivalent of Impressionist Paintings Can Make You Feel Data


A group of artists shook the world in the 1860s by painting what they saw, thought and felt. They became known as the Impressionists and they weren’t interested in recreating perfect visual appearances like hundreds of artists before them.

Instead, painters like Claude Monet struggled to find a new way to represent the world in order to keep it alive and real. They did this by creating an “impression” of how a person, landscape, or object appeared to them at a certain point in time. In doing so, they captured every aspect of their changing societies and transformed the very nature of how people think about and engage with art.

Our world today is shaped by equally intangible elements, such as data. Like the Impressionists, scientists need to visualize these things in a way that can help people see the world (and how it’s changing) anew.

In 2020, the average person created at least 1.7 megabytes of data per second browsing online banking systems, emails, medical records and social media. To try to represent data, scientists usually use graphs or tables. With much of society now suffering from what has been described as data fatigue, traditional methods of representing all the facts and figures in circulation are unlikely to be effective.

For example, the smart meters that were introduced to households in the UK were supposed to motivate people to save energy through a better understanding of where it was wasted. But research suggests that many people find data visualizations confusing and difficult to relate to daily household activities.

Just when people need to engage in the effort to avoid the worst consequences of climate change, data fatigue distracts them. As the 19th century impressionist movement did for art, 21st century science needs a new way to represent data.

An impression of island life

Data impressionism is meant to imbue the data with a vividness that enhances understanding and perhaps even influences the behavior of those viewing it.

The idea is to make the data more perceptible and therefore easier to interpret. A data print should only represent accurate data, but unlike traditional graphs and figures, it is designed to get people thinking about how the information makes them feel.

A data impression that my colleagues and I developed echoes the work of Impressionist painters who used changing light and color to depict an impression of a scene, such as Claude Monet in his 1872 painting Impression, Sunrise.

Flat Holm Island is a nature reserve located a few miles off the coast of Cardiff, Wales, in the Bristol Channel. Flat Holm contains rare plants, such as sea lavender and wild leek, as well as a colony of sea gulls. Its protection depends on its ability to remain a place of interest in the public consciousness. Much data has been collected on the biodiversity of Flat Holm over the years, and a local weather station monitors sunshine, wind and rainfall.

A temporary exhibition at the Techniquest science museum in Cardiff depicts some of these data streams using colored LED lighting, moving parts and reflective surfaces. An online application has been developed to support the exhibit and is now used by the island keeper to count and report the number of seagulls, butterflies, shelducks and other species. An impression of this data is then revealed to the museum public through an interactive map.

A simple press of the seagull button on the screen unleashes a pattern of color and movement. If many seagulls have been counted on the island, the changing colored LED lights are vibrant and fast. Though few seagulls were reported, the flickering LED lights are slow and quiet.

The exhibit also captures a real-time impression of the weather on the island. Data extracted from the weather station transforms prism-shaped panels on a mechanical display to give an impression of sunny, cloudy or rainy weather.

The exhibit allows visitors to immerse themselves in a visual display of life on the island. Like Impressionist paintings before, it uses aesthetic elements and principles to make the data more real. If scientists want to successfully engage people with complex data, they must generate experiences that allow them to connect and relate to it.

This article originally appeared in The Conversation.

Fiona Carroll is a Lecturer in Human-Computer Interaction at Cardiff Metropolitan University, Aidan Taylor
is Senior Lecturer in Computer Embedded Design at Cardiff Metropolitan University and Jon Pigott is Senior Lecturer in Art and Design at Cardiff Metropolitan University


Comments are closed.