__Coordination of WP3:__ C. Narteau (IPGP)

Despite recent progress, more research is still necessary to investigate the morphodynamics of mobile sedimentary structures in a quantitative way, especially at large length scales where different levels of complexity may arise. Hence, **WP3** aims to:

- develop new numerical methods to investigate the physics of sediment transport at a regional length scale in order to compare with climate global and meso-scale models developed at the LMD.
- test the influence of cohesion on grain interaction (LGL) and the impact of cohesion and topography on dune morphodynamics (IPGP, LPGN).

The planetary dunes database, built by the contributors of **WP1**, is of critical importance for **WP3** because numerical model parameters need to be set-up to the appropriate values of the relevant physical parameters in the different environments under consideration. In addition, theories and outputs of numerical simulations require comparisons with natural examples. Most importantly, models can also be used as predictive tools to feed the database, improve the optimization of the current observational practice and the development of new ones. These interactions will ensure permanent collaboration between the members of the different Work Packages.

An important objective of this project is also to couple climate and dune modeling. Following the collective and interdisciplinary philosophy of the EXO-DUNES project, numerical (IPGP) and climate modeling (LMD) will also be compared to laboratory experiments (**WP2**).

**Planetary climatic environment for the formation and evolution of dunes: Mars and Titan **

__lead__*: A. Spiga (LMD) *

__coll.__: B. Charnay, S. Lebonnois, F. Forget (LMD)

Over the duration of the project, LMD will provide numerical predictions from carefully-validated models of the wind and stability conditions close to the surface in the atmospheres of Mars and Titan. Those outputs from LMD climate models will be employed as inputs for dune formation experiments and models in collaboration with the WP2 and the tasks 3.2, 3.3 and 3.4. The consistency between dune observations, formation processes and predicted climate conditions will be discussed and possibly used to validate and improve the global and regional climate models, in close collaboration with the entire WP1.

As far as Mars is concerned, the LMD global-scale, regional-scale and turbulent-resolving models (Spiga and Forget, 2009; Spiga, 2011) can help to study how the interplay of wind variability at different spatial and temporal scales shape individual dunes and organized dune fields. Horizontal resolutions as fine as 1 km even allows for studying the interaction between wind, dune and topography (see WP2, tasks 3.2 and 3.3) to understand the morphology and location of dunes in specific places (e.g. basaltic dunes within craters, ice dunes of the North Polar Cap and gypsum dunes of the Circumpolar dune field, in collaboration with Tasks 1.M.1 and 1.M.2). In peculiar cases, it is thought that dunes may be relatively ancient (millions of years), providing potential clues for paleoclimate environments which can also be simulated through LMD models (Forget et al., 2006; Madeleine et al., 2009).

For Titan, global circulation modeling (Lebonnois et al., 2009) will be used similarly as inputs for dune morphology models. Observational and theoretical constraints provided through dune analysis will be particularly significant to help validate the climate models close to the surface, where many parameters are poorly known. For instance, the LMD team recently showed that Titan exhibits a very original boundary layer structure, distinct from other terrestrial planets, which can explain both winds at the Huygens landing site and dune spacing (Charnay and Lebonnois, 2012). Hence analyzing dune observations, through collaborations with the task 1.T, will impact the understanding of climate processes. LMD will also pursue the development of mesoscale models for Titan by building a version adapted from the Martian model.

**Real-Space Cellular Automaton dune models**

__lead__*: C. Narteau (IPGP) *

__coll.__*: F. Métivier, O. Devauchelle, O. Rozier, P. Lu (IPGP), Z. Dong (CAREERI) *

The REal-Space Cellular Automaton Laboratory (RESCAL) is a free software under GNU licence developed in IPGP to study complex geophysical systems (www.ipgp.fr/~rozier/rescal.html). In this package, the dune model is a hybrid approach that combines a cellular automaton of sediment transport with a lattice-gas cellular automaton for high Reynolds-flow simulation (Narteau et al., 2009; Zhang et al., 2010). In all natural environments where the dune instability can be observed, this model can be used by setting the length and time scales to the appropriate values of the relevant physical parameter. Thus, we can set up the model to reproduce all types of dune features on Earth but also on other planetary bodies (see www.ipgp.fr/~narteau/dunes.html).

In this task, we will primarily focus on:

*(1) Models of d*

__une fields using the outputs of climate modeling____:__Because they are constructed from an elementary length scale that incorporates all the diversity of the small scale processes, cellular automaton dune models are the most efficient numerical methods to investigate the dynamics of a significant portion of a dune field over long times. As a consequence, they are the best modeling alternative available to inject the output of climate global and meso-scale models which have coarse resolution if compared to the grain or even the dune scales. In practice, at each time step, we will have to reset both flow strength and orientation of the model to the variable conditions given by the climate models (see

**Task 3.1**). Then, to adapt to new flow orientation, the sedimentary layer will be located on the top of a rotating table. To modify flow strength, we will change the threshold shear stress value for motion inception.

*(2) The influence of cohesion and topography on the morphodynamics of dune fields*

__:__First, to take into account the role of cohesion in the model, we will introduce a new field to the structure that characterizes the sedimentary state. It will be thus possible to determine the local effect of cohesion according to the convolution between this field-value and continuous functions derived from precise physical mechanisms. Second, the real-space cellular automaton dune model will directly take into account topographic obstacles and various types of boundary conditions. We will be able to study the interaction between the shape of linear dunes and topographic terrain features. This analysis will be done upstream of laboratory experiments to test if the topography may significantly alter the geometry of dune fields (

**WP2**). In relation with the problem of dune orientation on Mars and Titan (

**Tasks 1.M.1, 1M.2**and

**1.T**), our objective will be to observe how cohesion and topography may affect the dune morphology.

*(3) Sub-surface processes and sedimentary structures:* Subsurface processes may be critical for dune field behaviors, particularly if they stabilize the sedimentary layer at depth or produce chemical reactions that change the composition of the sediment. These subsurface processes may be driven by liquid percolation or where freezing and melting occur (such as for some dunes on Earth, Mars and Titan, see

**WP1**). Because the real-space cellular automaton dune model is 3D, it will be possible to introduce subsurface flow and to change the properties of the granular medium as a function of depth. In addition, we can pay more attention to the analysis of compositional heterogeneities (size, shape) in order to reproduce segregation and stratification patterns in avalanche dynamics (see

**Task 3.4**).

**Modeling of planetary dunes composed of cohesive and volatile materials **

__lead__*: S. Carpy (LPGN) *

__coll.__: C. Herny, O. Bourgeois (LPGN)

We currently have very little knowledge on the formation of dunes composed of exotic materials such as ices, sulfates and hydrocarbons. Ice, sulfate and hydrocarbon dunes have been discovered very recently on Earth (Frezzotti et al. 2002), Mars (Massé et al. 2010, Smith and Holt 2010) and Titan (Lorenz et al. 2006). Our goal is to understand how the shape, the internal structure and the temporal evolution of these exotic dunes are controlled by specific processes due to the physical properties of their constitutive materials, such as grain cohesion, grain volatilization by sublimation and grain formation by solid condensation. We will explore the role of these specific processes with numerical models. First, we will develop a numerical code to simulate wind flow and to map pressure and temperature conditions above a stepped topographic surface. Then, we will introduce in the code erosion and sedimentation laws, depending on pressure and temperature conditions and specifically designed for cohesive and volatile materials, which will be used to modify the shape of the topographic surface in response to wind flow. These numerical models will be constrained by comparing their results with observational data acquired on natural examples of ice and hydrocarbon dunes on Earth (**Task 1.E.2**), Mars (**Task 1.M.2**) and Titan (**Task 1.T**).

**Numerical investigation of out-of-equilibrum transport of grains **

__lead__*: P. Allemand (LGL) *

__coll. __: V. Langlois, A. Quiquerez (LGL)

We have developed a numerical model to simulate the motion of sedimentary particles when sheared by a fluid flow. The idea is to combine a Discrete Element Method to model the grains with a simplified description of the fluid flow. Sand grains are modeled as spherical particles that can interact through contact forces (‘molecular dynamics’) while the wind or water flow is averaged over horizontal layers that exchange momentum with the grains. Therefore we take into account both the effect of the flow on particles and the retroactive effect of particles on the flow. This allows us to account for the differential inertia between grains and fluid. We will first apply this model to simple configurations (one size of grains and constant fluid discharge) in order to describe more precisely the dynamics of the sand bed (in particular, what are the depth and velocity profile of the layer of moving grains). We will then investigate the behavior of the sand bed when the fluid flow is either accelerating or decelerating: we will study in particular how much time is needed for the flux of grains to reach a new equilibrium value, the dynamics of exchanges of grains between the static and moving layers, the dependence of mean velocity and depth of the moving layer on the fluid flow and grain properties. The aim of this approach at the scale of the grains is to provide a better description of the flux of grains out of equilibrium that can then be included in continuum (see **Task 3.4**) and discrete models (see **Task 3.2**) of the dynamics of dune fields. In relation with the **Tasks 3.2** and **3.4**, an interesting objective is to incorporate the role of cohesion at the grain scale.

Maj : 16/07/2014 (11)