Post by Jenny Seifert
Using a tool similar to a computer game, Melissa Motew is peering into the future.
Motew is a modeler. She uses computers and mathematics to simulate ecosystems and make sense of nature.
Her task is to shed light on what the Madison area’s environment could be like by the year 2070 and what this might mean for human well-being—how much food could we grow, how well could the land withstand floods and will we have clean lakes yet?
“We want to track what’s happening through time, so we can understand all of the changes,” says Motew.
A PhD student in the Environment and Resources program at UW-Madison’s Nelson Institute, Motew is also part of the Water Sustainability and Climate (WSC) project’s modeling team. They are simulating the future to understand possible challenges and opportunities for sustaining freshwater given the host of long-term changes affecting it, especially climate change.
The Yahara Watershed, the land area that drains into the Yahara River and lakes and includes Madison, is the project’s research specimen. The model Motew works with is like the watershed’s avatar.
Scientific models, in general, are conceptual representations of the natural world based on the scientific understanding of how nature works. Scientists create models by translating that understanding into a language of equations and computer code. After making its calculations, a model then outputs stories about what is happening or could happen in nature—details that are otherwise difficult to observe in real life.
The model Motew uses simulates key natural processes, like photosynthesis and phosphorus runoff, that take place in the Yahara Watershed’s assorted ecosystems—its farmland, urban land, wetlands, forestlands and grasslands—and shows how different changes in land use and climate could affect them.
The results of this “SimEcosystem” (not its actual name) will be the final piece of Yahara 2070, the set of scenarios developed by the WSC project to understand an array of possible futures for the watershed and the implications for achieving a desirable future.
Step one: Start with stories
When the WSC project team describes Yahara 2070, we often use a quote by poet Muriel Rukeyser to help: “The universe is made of stories, not of atoms.”
Yet, we assert, the universe is really made of both stories and atoms. But when it comes to how humans make sense of the world, the stories often come first—mythology precedes ecology.
And so began Yahara 2070—as four stories that depict four different ways the watershed could change over the next two generations, each laying out a different set of choices people could make in addressing freshwater challenges and how land management could play out as a result.
For example, say we collectively decide that eating only local, grass-fed meat is the way to reduce agriculture’s impact on the lakes. Or what if we abandon meat entirely for technologically advanced fake meat made from plant tissues or animal stem cells? Such decisions would have significant implications for our land and water.
Each scenario was also given a different climate, reflecting the scope of regional climate change trajectories. Humans respond to climate change differently in each scenario, too, ranging from the do-nothing approach to a geoengineering fail.
With such distinctive and provocative pictures of the future, the scenarios allow people to compare the alternatives and discuss what is desirable. They also challenge the models to flex their analytical muscles.
The models essentially link the social realm to the biophysical realm. They take the land-use decisions and climate changes the stories present, simulate the natural processes involved, and churn out “the atoms,” the details about what these changes would mean for a number of the natural benefits, or ecosystem services, future generations will need to live happy and healthy lives.
“The models let you run experiments you can’t run in real life,” says Motew.
Step 2: Simulate the system
Simulating ecosystems was not what Motew first set out to do in life. Music was her initial path, until she realized her love of science trumped her love of the trumpet, at least in terms of what she wanted to pursue in college. So she changed course to explore physics.
Post-graduation she landed a job at MIT Lincoln Laboratory, where she got her start doing systems analysis for the defense industry and became proficient in using computers to solve problems.
Ultimately, her interest lay not in problems related to national security, but in those related to nature. Ecosystem modeling seemed like a better fit.
“Ecosystems are incredibly complex and beautiful systems to study, and immensely satisfying to learn about,” says Motew.
She came to UW-Madison for graduate school to work with Chris Kucharik, a professor in the Nelson Institute and Department of Agronomy. Her master’s project entailed working with an ecosystem model nicknamed AgroIBIS, preparing her for her Ph.D. work on the WSC project, which Kucharik leads.
“AgroIBIS is a big, powerful model that simulates a lot of what happens in an ecosystem. That means there is a lot of science that you can do,” says Motew.
The model simulates the terrestrial landscape: the suite of biophysical processes for the vegetation and soil, the effects of the weather and climate, the movement of energy and nutrients, and the decisions people could make about what to do on and with the landscape—where and how to farm, where and how to build, etc.
But, when the project started, AgroIBIS was missing one thing: phosphorus, an essential element for understanding the future of the Yahara Watershed.
When it comes to freshwater quality in the Yahara, phosphorus is “public enemy number one,” as former Dane County Executive Kathleen Falk once put it. There is simply too much of it in our soils and waterways, creating a health hazard for both lake ecosystems and people.
Incorporating phosphorus into the model was necessary to understand what the various land-use and climate changes presented in the scenarios could mean for future water quality in the watershed. And so it was Motew’s first order of business.
Putting the “P” in AgroIBIS meant combing through the literature and piecing together code and equations that would allow the model to imitate how the nutrient moves through the system. Despite initially knowing very little about phosphorus, Motew successfully designed a representative module and wove it into the model’s thousands of lines of code.
Motew also had to test the model to make sure it acts like a real ecosystem, which entails comparing the numbers it generates, such as how much corn grows, against real-life field data.
“Field observations and models are complementary,” says Motew, explaining the iterative process that takes place between measurements and models. This enables scientists to home in on how exactly a system works, what measurements to take and what questions to ask.
AgroIBIS communicates with two other models to complete the simulation of the watershed. Next in line is a model nicknamed THMB, which imitates how water and nutrients flow throughout the watershed, based on what AgroIBIS tells it about happenings on the landscape. Importantly, it models the delivery of phosphorus from land to lake.
For example, say AgroIBIS simulates a cornfield in a rainstorm. It will estimate how much phosphorus runoff occurs, passing that amount off to THMB, which then determines how much phosphorus will end up in the lakes.
TMHB then sends that amount to the final model, the Yahara Water Quality Model, which calculates what all of the changes that happened in this “SimEcosystem” would mean for lake water quality.
“The model marches through time. It’s crunching numbers, calculating quantities. Our job at the end is to look at those quantities, analyze them and try to learn something from them,” says Motew.
Step three: Ask more questions
Motew explains the models are designed to be “forward capable,” to be able to mimic situations never before encountered. This sets them apart from many other models, which rely heavily on information about the past.
“Extrapolating the past into the future doesn’t really work, especially when so much change is taking place,” says Motew.
But the WSC models don’t forget the past entirely. Their scaffolding is based on how natural systems work at a fundamental level—knowledge gained from accumulated field observations—again demonstrating that models and field data go hand-in-hand.
As the model marches forward hour by hour, day after day until the year 2070, it tells detailed stories about not just how ecosystem services could change, but also why. Was it because of those droughts, or all that rain, or is all that pasture doing something we never expected it would?
While the models can’t predict the future, they can show us what is possible and how to prioritize decisions moving forward. Unlike many other models, they aren’t focused on the impacts of a specific land management practice, but on the bigger picture: what humans need over the long term.
“[This information] helps us understand what could happen and gets us thinking about what we should do to steer ourselves in one direction versus another. It can tell us what aspects of the future we really need to pay attention to and maybe those aspects that don’t deserve as much attention,” explains Motew.
One thing that excites Motew about the WSC models is the seemingly limitless questions they can help answer. How could climate change affect crop yields or carbon sequestration in the watershed? How could any number of changes to the landscape mosaic impact water quality?
“One thing I wonder is, how much control do we have in this watershed in terms of the actions we take. How much power do we have in the face of changing climate?” poses Motew.
And while Motew may not fit most people’s image of an ecologist, out in the field collecting measurements, she asserts her method of making sense of the world is not all that different.
“I typically don’t do fieldwork, but I try to make sense of fieldwork and develop tools to better understand ecosystems,” she says.
The fruit of her and the WSC modeling team’s labor will be revealed in early 2016.