Confronting Grand Challenges in environmental fluid mechanics

Environmental fluid mechanics underlies a wealth of natural, industrial, and, by extension, societal challenges. In the coming decades, as we strive towards a more sustainable planet, there are a wide range of Grand Challenge problems that need to be tackled, ranging from fundamental advances in understanding and modeling of stratified turbulence and consequent mixing, to applied studies of pollution transport in the ocean, atmosphere, and urban environments. A workshop was organized in the Les Houches School of Physics in France in January 2019 with the objective of gathering leading figures in the field to produce a road map for the scientific community. Five subject areas were addressed: multiphase flow, stratified flow, ocean transport, atmospheric and urban transport, and weather and climate prediction. This article summarizes the discussions and outcomes of the meeting, with the intent of providing a resource for the community going forward.

Figure 1

The 2030 Agenda for Sustainable Development, adopted by all United Nations Member States in 2015, provides a shared blueprint for peace and prosperity for people and the planet, now and into the future. At its heart are the 17 Sustainable Development Goals (SDGs), which are an urgent call for action by all countries—developed and developing—in a global partnership. They recognize that ending poverty and other deprivations must go hand in hand with strategies that improve health and education, reduce inequality, and spur economic growth—all while tackling climate change and working to preserve our oceans and forests [1]. Of these, environmental fluid mechanics plays a substantial role in achieving SDGs no. 6, 7, 9, 11, 13, and 14.

Figure 2

Multiphase flow processes in the environment at different scales: above us in the atmosphere, around us on land, and below us in bodies of water.

Figure 3

(a) Granular segregation leads to nonrepeatable laboratory experiments. Release of 1 kg of a $50\%–50\%$ granular mixture of white (0.425–0.600 mm) and red (1.0–1.3 mm) glass ballotini from a 2-m-long chute inclined at ${25}^{\circ }$, showing different run-out lengths and levee structures. Experiments by Ms. Elze Porte. (b) Granular forces are transmitted in a nonhomogeneous force-chain structure. Snapshot of a photoelastic 20-particle-deep avalanche experiment down a two-dimensional channel with rough base. Experiments by Dr. Amalia Thomas.

Figure 4

Particle configurations at different times during the settling process of cohesive sediment. The color reflects the vertical particle velocity. Adapted with permission from Ref. [18].

Figure 5

Contours of the spanwise vorticity for simulations with the same initial Reynolds number and bulk Richardson number but (a–c) prone to primary KHI where the initial shear layer depth and density interface depth are equal and (d–f) prone to primary HWI where the density interface is significantly “sharper” than the shear layer depth; at $t=t_{2d}$ (scaled with advective time units $d_0/U_0$), when spanwise-averaged TKE is maximum; $t=t_{3d}$ when three-dimensional perturbations are maximum; $t=t_{2d}+100$. Note “overturning” by the primary KHI and “scouring” by the HWI. Used with permission from Ref. [72], copyright CUP, all rights reserved.

Figure 6

Potential oil spills detected by satellite ESA Synthetic Aperture Radar (SAR) in the period 1999–2004 in the Mediterranean Sea [98].

Figure 7

Diffused concentration $c({\mathbf{x}}_1,t_1)$ at time $t_1=t_0+90$ days, with the advected positions of material Agulhas ring boundaries (identified as diffusion barriers) overlaid in red. Adapted from Ref. [111].

Figure 8

Beached oil distribution from ensemble simulations for different release points around the Caribbean Islands, simulating accidental oil spill transport produced by the flow field conditions in 2013. Experiments by Dr. Antonio Augusto Sepp-Neves.

Figure 9

Integration of field measurements, laboratory measurements, and high- and low-fidelity computational models to provide predictions with confidence intervals.

Figure 10

Simulation results for JU2003: iso-contours of Q-criteria colored by velocity magnitude obtained from LES [147] (left), pollutant source and sensor locations (center), and RANS predictions of pollutant concentrations with 95% confidence intervals compared to field measurements [155].

Figure 11

Future prediction system advances will arise from improved resolution delivering a more accurate representation of physical processes, more complex models delivering a better representation of the entire Earth-system that is highly relevant for longer-range prediction, and larger ensembles delivering more reliable forecasts from better uncertainty statistics. Increasing both ensembles and model complexity has at least a linear impact, while better resolution has at least a cubic impact on the computational cost of simulations.