Columbia Water Center welcomes Pierre Gentine, Assistant Professor of Applied Mathematics at Columbia University’s School of Engineering and Applied Science, as an affiliate researcher.
Pierre’s groundbreaking research on the way soil moisture interacts with the atmosphere has implications for many of CWCs initiatives—from developing more efficient irrigation systems, to water resource management, to understanding floods.
In this interview, Pierre explains why a better understanding of soil moisture/atmospheric interactions is needed both for a more accurate understanding of human caused climate change as well as water management at different scales.
Can you explain a little bit about your work and how it relates to Columbia Water Center?
I am researching hydrometeorology. It’s basically the link between hydrology, surface water and the atmosphere. We call it hydrometeorology because our scale is a little smaller than what climatologists typically use. We look at smaller timescales—typically over a day—and smaller spatial scales, an area of a few kilometers or a few miles, not too big.
The idea is to try to understand the daily cycles of energy and moisture at the land surface. If we understand how soil moisture evolves through the course of the day, we can extend that to a couple of days, and then to a season. Then we can relate the evaporation we observe over the landscape to the evolution of soil moisture.
This understanding has implications for water resource management, for floods and drought forecasting, also for irrigation management – it has all sorts of implications, especially for agriculture. It’s also interesting from a climate change perspective as well, in that it could help predict, for example, heat waves, floods and droughts– things that we don’t understand so well yet.
How would this research affect our understanding a heat wave or a drought?
When you look at the moisture in the soil and in the atmosphere, together they form a coupled system. In this system you can have feedbacks that could create a vicious circle.
So say, for instance, you start with a pretty dry season, in the summertime or spring, like what’s happening now in Europe. The problem is that as the dries out, there’s less soil evaporation; at the same time you are providing more heat to the atmosphere. You then have less precipitation, because you have less moisture coming from the soil. So in a sense it’s like a vicious circle that increases the heat-wave.
That’s a typical example. And its exactly the same for floods, it’s just the other way around. If you have too much moisture in the soil, then you have more evaporation, so more precipitation in the atmosphere, and more moisture falls back–a vicious circle.
Most people assume that moisture comes from the ocean or big bodies of water. How big an impact is this evaporation from the land surface?
Historically, people were not much interested in the feedback between the land and the atmosphere, because as you said, lots of moisture is provided by the ocean.
But in fact you can have some pretty subtle effects from the feedback between the land and atmosphere. For example, when the land warms you can get a kind of breeze effect that brings in moisture from remote sources. So you can have feedbacks that are subtler than what we previously understood.
So what you’re saying is that the interaction between soil moisture and the atmosphere even affects how moisture is drawn in?
Exactly, it can be non-local.
Typically dry land will warm up the surface. But if I have an ocean nearby that is very cool and moist, the warming can generate a breeze effect that brings soil moisture over the land. That’s what you typically have in the summertime over the Eastern U.S.
You don’t see any clouds over the ocean. In fact, you can see the border of the continent through the outline of the clouds. Right on the border, you can see the clouds forming.
The reason is that you provide moisture from the ocean with this breeze effect during the day, and then through the day you’ll see clouds forming, and later in the day you’ll see precipitation forming over the land, but not over the ocean.
This is something you can see from overhead, and you can see it evolving over the course of the day. So in the morning you won’t have any clouds, mid-morning more clouds, late afternoon even more clouds, and then you have precipitation or what we call convection.
Do you also look at things like forest cover, and deforestation effects?
Yes, exactly. So for example, the Amazon rain forest is sometimes called the “Green Ocean” because it has so many trees, and trees provide moisture locally. If you deforest a part of it, that part becomes like land, and you see some pretty funny pictures of clouds forming only over the deforested patches.
Depending on the scale of deforestation, there might be a point at which the forest can actually regenerate the savannah, and provide enough moisture so that the savannah could grow to become a forest again. But if the site of deforestation is too big, its possible that the forest won’t be able to provide enough moisture–so at some point the deforestation cannot be reversed.
When people talk about climate change, they usually mean “global climate change.” It seems like what you’re talking about is the interaction of the local with global system. Can you talk a little more about that?
If we think about anthropologic changes other than global climate change and you look at what’s happening in India it’s an interesting picture. People there have been using irrigation all over the place. In fact they’ve strongly modified the moisture and heat sources over the continent. It’s at such a big scale it’s almost like geo-engineering. So if you play at those scales then you are impacting the climate locally. These things are still poorly understood. What is the impact of irrigation on such a huge scale, a continental scale, on the pattern of the Monsoon?
Have you discovered anything in your research on this point?
It’s not finished yet, but I’m working on a conceptual model to understand the daily cycle of clouds and precipitation over land. If it works, it would have an impact in the sense that we will have a physically-based model to understand why we observe the daily cycle of precipitation over continents, during the summer, and why we observe precipitation at particular times of day.
Right now cloud cover is a major problem for climate change models–if you look for example at the IPCC reports, you can see that low-level clouds really are the major difference among models. It is really one of the major, major issues; the forecasts go all over the place. Basically under climate change nobody knows what will happen to those clouds. The reason is that we don’t have a physical understanding of how they form, and how we can relate those to the surface.
One thing that is subtler, is that clouds modify the planetary albedo, so they also modify the heat transport over long distances. So if you don’t get the right cloud cover and cloud models, then your understanding of heat transport is not correct. People need to understand that we are still facing issues in understanding what will happen.
Could you explain more how your research might impact agriculture?
My model should better explain why we get droughts and floods; and even more importantly, how droughts and floods are triggered and how long they’ll last. We hope to get a physical understanding of how long they last. That’s the main part.
If you consider water resource management, if we understand evaporation better over a basin, then we can be much more confident of our capacity to deal with water resource management because we have a better understanding of the underlying soil moisture state.
If you can have better predictions you can say “okay, we just need to irrigate within a week”–you can have some kind of range. Right now the problem is that our capacity to predict soil moisture is really poor.
In fact if we had something that was more consistent and physically-based, maybe we could apply that over large areas, using remote sensing, for instance, so we wouldn’t need any local measurements, something especially useful for developing countries, because [on the ground measurement] costs a lot. So you could use just remote sensing to predict your water or your soil moisture state and therefore your water resources; and you could have better management of your resources just based on pictures from satellites. That’s the ultimate goal.