The project

In the growing field of landscape dynamics under climate change, the role of aeolian processes becomes prominent as precipitation decreases and desertification occurs. In this context, hot and cold deserts display a mosaic of specific depositional (e.g., sand and snow dunes) and erosional aeolian landforms (e.g., yardangs and sastrugi) that represent effectively (paleo)-environmental proxies, even in presence of vegetation (Figure 1). This project aims to reproduce these landforms by using physics-based cellular automaton (CA) models and simulation-based Computer Graphics techniques to combine emergence mechanisms and long-term dynamics with the realism of computer-generated 3D scenes.

Context

Aeolian landforms and especially dunes are major socio-economic issues and often a priority in land use planning. For example, the preservation of coastal dunes is essential in the context of rising sea levels. This problem directly concerns the Atlantic coastline in France and many coastal areas in Europe (e.g., North Sea, Baltic Sea) and is often accompanied by heavy expenses on the scale of the local communities. The stability of vegetated dunes is also essential to combat desertification, a phenomenon that is increasing in all scenarios of global warming, including in Western Europe. In addition, and regardless of the major scientific issues surrounding planetary climate and sediment transport dynamics, desert areas are still the subject of power struggles in Central Asia, Africa, the Arabian Peninsula and the poles. In these territories rich in raw materials, aeolian processes play a critical role and their environmental impact leads to important investments in fundamental research and infrastructure management at the scale of these territories. A surprising paradox in the study of aeolian landforms is that they are highly organized patterns on a large scale and of astonishing complexity on a small scale. A consequence is that they are often the first landforms observed from space, even at very low resolution, and at the same time difficult to decipher from the ground. This project, merging physically-based modeling, simulation and Computer Graphics techniques, will break this scale issue by offering a visual solution that satisfies physics and realism.

Artistic views and the incorporation of some degree of realism have always been used to illustrate scientific results. Beyond their pedagogical and communication interests, these visual aspects can also be generated based on scientific knowledge using virtual landscape simulations (Figure 2). This domain is experiencing rapid growth and it represents an opportunity to facilitate research, concept development and data acquisition in Earth Sciences, including studies on aeolian morphodynamics and desert mutations. This is particularly the case with regard to the increasing use of remote sensing data (satellite images, digital topographies, etc.). Although they provide new opportunities to analyze the morphology and distribution of landforms associated with sediment transport, the experience shows that these practices lead to a loss of cognitive cues on the length and timescales of the mechanisms that govern their evolution. The fluent manipulation of these data from one spatial or temporal scale to another is an advantage from a technical point of view, but also a source of subjective bias when it comes to the actual quantification of the examined morphodynamics. modeling and experimenting with these data in 3D with an increased level of realism should thus enable researchers, students and other audiences to better understand the nature and history of the landforms under consideration in order to develop an adequate mental representation of their dynamics. This approach is even more important for landscapes in inaccessible areas, due to geopolitical issues on Earth or on extraterrestrial planetary surfaces.

ANR reference

We acknowledge financial support from the French National Research Agency (ANR-23-CE56-0008/EOLE).

Partners

LIRIS laboratory

The LIRIS lab (UMR5205)
Origami team
People involved: Eric Guérin, Eric Galin, Adrien Peytavie, Hugo Schott

IPGP laboratory

The IPGP lab (UMR7154)
Geological Fluid Dynamics, Tectonics, Planetology and Space Sciences teams
People involved: Clément Narteau, Olivier Rozier, Sébastien Rodriguez, Laurie Barrier, Antoine Lucas, Jeanne Alkalla