Modeling ice and snow on planets
The presence and behavior of ice and water are key to understanding the evolution of Mars’ geology and climate, the potential for life on Mars, and the future of exploration.
This enhanced-color image shows a 45-meter-diameter crater on the surface of Mars, appearing as a glistening sphere on a rusty red background. Glistening material is interpreted as water ice slowly subliming. Photo credit: NASA/JPL-Caltech/University of Arizona.
published in the Journal of Geophysical Research: Planets outlines a unified model for investigating evaporation rates on all planets and moons with atmospheres, with Earth, Mars, and Titan as test cases, using recent advances in the science of atmospheric boundary layers. The paper focuses on the near-surface part of the atmospheric boundary layer, where wind is influenced by friction over the planet’s surface.
The new model can help investigate the behavior of water and ice found on the surface of any rocky body with an atmosphere, ranging from Earth, to Pluto, to exoplanets.
The model results improve constraints on the stability of ice on Mars, and will help scientists understand whether ice can melt to form liquid water under present‐day conditions.
We spoke with authors , a postdoctoral researcher at the Jet Propulsion Laboratory (JPL), and Gary Clow, an INSTAAR research scientist who recently concluded a long career as a geophysicist with the U.S. Geological Survey.
Q: What does your model tell us about ice and evaporation on Mars?
Khuller: About a third of Mars has ice buried below the surface, probably deposited as dusty snowfall during a series of past Martian ice ages. This Martian ice is just like the snow and ice we have on Earth, and it usually contains a small amount of dust in it.
Over the last two decades, we have found hundreds of places where the buried ice becomes exposed at the surface of Mars due to meteor impacts and the overlying dry material slumping down on steep slopes.
Aditya Khuller, a postdoctoral researcher at the Jet Propulsion Laboratory (JPL)
Once the ice is exposed, it seems to disappear some time within a few Mars years (roughly twice the length of an Earth year). On Earth, ice usually melts when heated by the Sun. However, because the Martian air is extremely dry, it is commonly claimed that ice cannot melt on Mars under present-day conditions. The long-held view has been that the amount of energy lost from sublimation, when the ice turns into gas without melting, exceeds the incoming solar radiation at the surface.
But Gary had done some work in the 1990s that suggested that maybe the energy losses from ice sublimation on Mars were much lower than commonly believed, because the equations being used by other authors were simplistic. He didn’t complete this project then, because he began focusing more on ice on Earth. So when I began emailing him a few years ago, we decided to resurrect his old atmospheric turbulence work, and update it using advances made over the last three decades.
Before we applied our simulations to Mars, where we only have indirect ice measurements from one landed location, we wanted to make sure our simulations were accurate. So we compared our results with different types of measurements made on Earth, Mars, and Titan. We found that our simulations agreed up to 71% better with measurements compared to previously developed models used for Mars ice and water.
Our new simulations for Mars ice predict significantly higher or lower rates of evaporation when compared to previous estimates, depending on atmospheric conditions and how rough the surface is. We are using our new results to find the conditions under which ice could melt on Mars to produce near-surface liquid water and determine how the climate of Mars has been changing over time.
Q: So, in some places ice could melt on Mars, according to your model?
Khuller: We are currently in the process of investigating when and where ice could melt on Mars today. The answer would depend on how hot the ice and the land around it get relative to the air, how rough the surface is, and how windy it is.
Some of Gary’s other work from the 1980s suggests that dusty (about one percent) Martian snow could be melting on Mars today. But that previous work of his used very simple atmospheric calculations that we have now improved upon. Nevertheless, he found that dense, dusty snow could be melting below the surface because sunlight heats up the dusty snow below the surface due to a greenhouse effect. Just like on Earth, small amounts of dust mixed in with the snow lead to the snow becoming darker and warmer.
At the surface, the snow could still be sublimating. We are now building on his and other Martian ice work conducted since the 1980s to figure out if there are some places where ice could be melting on Mars.
Q: The model was tested against observational data from Mars. Would it work on any planet?
Fig. 6 from the paper, showing images of in situ observations of H2O sublimation on Mars at the Phoenix landing site and data/model comparisons of key factors like surface roughness, wind speed, and surface temperature. See paper for details.
Khuller: Yes, we have tried to make our model as flexible as possible so that it should work on any planet or moon with an atmosphere. We are hoping to collaborate with exoplanet experts, for example, to see how we can simulate the evolution of oceans, glaciers, and lakes on a wide range of planets beyond our solar system.
Clow: I was thinking of the model strictly in terms of Mars at that time we started. But working with Adi, I said, why stop at Mars? Why not Titan? And why stop there? Why not exoplanets? If we make the model general enough, we could even apply it to investigate the dunes on Pluto. We could calculate all the stresses that occur on the surface to get particles moving to make the observed dunes. That would be cool!
So, the way this model ended up being designed, you can use it for all kind of applications.
When I see that little helicopter flying around Mars, I’m thinking, that little guy’s in the boundary layer. We can calculate the shear stress that that thing’s flying in. That’s an application that would be of interest to people at JPL designing helicopters. Now they’re designing one for Titan, to look at dunes there among other things.
Dunes, aeolian transport, erosion rates. Anything that involves wind, the model can be applied to. You can calculate how much stress is on the surface that’s trying to get a little grain saltating.
Another thing of interest is being able to calculate what the temperature is at the surface of a planet and how that might change over time. Our model then can be incorporated into larger models to calculate the energy balance at the surface, which controls the temperature on Mars, Titan, Pluto, or Earth in the distant past when the atmosphere was different. The model could become a component of a larger model, to help you understand some other problem. It’s integral to understanding the environment down in the atmospheric surface layer that interacts directly with a planet's surface. So, the model can do much more than just calculate sublimation and evaporation.
We actually wrote the paper using all kinds of examples, but the reviewers didn’t like it. They wanted us to focus on one planet and one issue. But that opens the door to writing more papers where we investigate these other examples in more detail.
Q: What inputs and outputs does the model have? In other words, what do you need to know going in? What do you find out from running the model?
Gary Clow, an INSTAAR research scientist
Khuller: Once a site has been chosen in terms of planetary body, atmospheric surface pressure, and surface roughness, the model is driven by the temperature and relative humidity at two heights and the wind speed at a single height within the atmospheric surface layer. The overall planetary boundary layer height must also be specified. These quantities can come either from meteorological measurements or synthetic data from a climate model.
By running the model, we can calculate how the wind, temperature, and humidity vary with height. Additionally, we can calculate the energy and amount of volatiles lost or gained from the surface to the atmosphere. We can use this information to predict whether ice can melt, how the climate is changing, how future Martian missions will be affected by the weather, and more.
Clow: The atmospheric boundary layer is that layer between the surface of the planet and the "free" atmosphere that is running above it, which isn’t affected very much by the surface. What affects the surface temperature and the amount of evaporation and the wind shear at a planet's surface all depends on the wind speed, temperature, and humidity gradients from the surface up through the boundary layer. It’s these gradients that determine what the sensible heat flux is from the ground to the atmosphere, and the rate of evaporation or sublimation rate from the surface to the atmosphere. The wind speed in conjunction with buoyancy creates turbulence that transfers heat and volatiles (water vapor on Earth or Mars or methane vapor in the case of Titan) up from the surface into the atmosphere.
Q: The model has roots in some advances in earth science. Can you tell us about that?
Clow: One advance we made is that the equations built into the model take into account the buoyancy due to both the temperature and humidity gradients. On Mars, the primary volatile—H2O—is much lighter than the ambient gas, which is CO2. So if you have a parcel of air that’s saturated with water at the surface of Mars, if the air above it is drier, it wants to go up. Even if the surface layer is isothermal, the parcel will still want to buoyantly rise. This effect is much stronger on Mars than on Earth because of their different atmospheres. This additional buoyancy affects all the atmospheric stability equations that determine the wind, temperature, and humidity gradients. We explicitly built this buoyancy effect into the model.
This has been done for Earth, too, but it’s been sort of hard coded in, where they just put in Earth’s atmospheric molecules. In every atmospheric science book I’ve ever looked at, the molecular effect on buoyancy is listed as a constant. They don’t list the equation that was used to calculate the constant. But you can’t just take constants from Earth and apply them to other planets. So we built in the actual equations into our model instead.
Another advance we made was to include the effect of gustiness at the surface due to large-scale convective structures that occur during really unstable atmospheric conditions. These structures are commonly observed in Earth’s deserts on hot afternoons, and they may be common on Mars too under certain conditions. The gustiness not only affects the wind profile, but the temperature and humidity profiles as well.
We haven’t done anything really new other than to pull all these pieces together and integrate them in once place. I don’t know that anyone has done that before—certainly no one’s done it for other planets.