Landscape Prediction: An Earthology of Moving Landforms


Recent space imaging developments have given rise to a spread of commercial services based on the temporal dimensions of satellite imagery. Marketed under umbrella terms such as environmental intelligence, real-time Earth observation or orbital insight, these imaging projects deliver the surface of the planet as an image flow encoded into video streams, where change and variation become a commodified resource on the one hand, as well as a visual spectacle on the other.

Postprocessed by computer vision and machine learning algorithms, these services extend the geospatial logic of GIS systems to the surface of the image. Paths become trackable, objects classificable and movements predictable. Seen from the satellite networks, the surfaces of the planet are imaged as visual feeders for data-extractive algorithms. As aerial images become data, then, the Earth is operationalised as a legible screen, where the predicted predates the perceived.

This workshop proposes to examine the use of video prediction techniques based on machine learning within this imaging context of the transformations of landscape. It seeks to explore the cinematic character of some of the active landforms of the planet, such as river thalwegs, drifting glaciers or crawling dunes. This way, the workshop will present the capacity to generate video-predicted landscapes as a platform to speculate with this particular entanglement between visual media and the surfaces of the planet, beyond the extractive and finantial contexts that have given rise to it.

The workshop is part of an ongoing research on the image character and temporality of the planetary surfaces developed together with Jussi Parikka and the Archaeologies of Media and Technology group.

Link to the workshop at Transmediale 2017: Transmediale 2017 Link to the workshop at Linz AMRO 2018: Art Meet Radical Openness AMRO 18

Selected texts:

2021 Jussi Parikka, From Planetary Depth to Surface Measure, or How to Read the Future from an Image Deep Mediations. Thinking Space in Cinema and Digital Cultures (eds. Karen Redrobe and Jeff Scheible), U. of Minnesota Press
2021 Bernabé Sauvage, Data is Beautiful. From data visualisation to data poetry Image and Imagery
2020 Abelardo Gil-Fournier, Jussi Parikka, “Visual Hallucination of Probable Events”. On Environments of Images, Data, and Machine Learning Big Data. A New Medium?, Routledge (London)
2019 Abelardo Gil-Fournier, Jussi Parikka, Visuelle Halluzination von möglischen Ereignissen oder Über Umweltbilder und Maschinelles Lernen Archiv für Mediengeschichte, 18
2018 Bethany Nowviskie, Reconstitute the world