Courtesy of The New York Times

Deforestation in the Amazon has dramatically increased in recent years, threatening the ecosystem and local communities.

October 31, 2020

Cornell Researchers Collaborate to Investigate Environmental Consequences of Hydropower Dams

Print More

The Amazon River Basin is under threat, largely at the hands of humans. 

To help change that, Prof. Carla Gomes, computer science, Prof. Alexander Flecker, ecology and evolutionary biology, and Rafael Almeida, postdoctoral researcher with the Atkinson Center for a Sustainable Future, designed an interdisciplinary research project to inform policymakers of the environmental consequences of human actions. 

According to Almeida, climate change, extreme weather events, deforestation, fires and mining — in addition to fisheries, Flecker added — are jeopardizing the integrity of the Amazon.

But the Amazon faces yet another seemingly counterintuitive environmental threat: hydropower dams.

Hydropower is a method of harnessing the energy from moving water and transforming it into electricity. Coveted for its low costs and use of water as a renewable source, hydropower has steadily grown over the past century, especially in highly populated areas such as Brazil, Almeida said. 

“In Brazil, the potential [for hydropower] in river basins outside of the Amazon has been nearly overexploited, making the Amazon the new frontier for hydropower development,” Almeida said. “If energy planners and policymakers want to continue with expanding hydropower in the country, then that is one important driver for [the] construction of dams in the Amazon.”

Almeida added that the energy harnessed from river basins in other, smaller Amazonian countries, like Peru and Bolivia, can also be exported to supply Brazil. 

But what makes hydropower dams dangerous for the health of the Amazon?

One reason is that hydropower dams disrupt the natural flow of water essential for river ecosystems. Almeida said rivers that normally transport nutrient-rich sediments are no longer able to fertilize the soil of downstream wetlands, and animals that rely on a seasonal pattern of water flow may be harmed, taking a toll on biodiversity.

Although the adverse ecological effects of hydropower dams are extensive, Gomes, Almeida and Flecker have focused their research efforts on the dams’ production of greenhouse gases. 

Depending on a dam’s design and location in the river basin, blocking the flow of water can lead to large inundated reservoirs of water that, according to Almeida, are conducive to the production of methane gas — a major culprit of climate change among the greenhouse gases. 

Microorganisms that live in the reservoirs can decompose vegetation and other organic material that have accumulated, producing the climate change-fueling methane gas in the process, Flecker explained. 

“This is something that we’ve been addressing in our research,” Almeida said. “How can you maximize hydropower production, [while] minimizing greenhouse gas emissions?”

To tackle this problem, Almeida and Flecker bridged the interdisciplinary divide and collaborated with Gomes’ team of computer scientists to sift through hundreds of possible dam projects and evaluate them for their environmental impact and electricity production. 

“When we deal with real problems in the academic world, you have your niches. I do X, and somebody else does Y, and nobody talks to each other,” Gomes said. “This project is awesome because you see, here I am a computer scientist talking to my colleagues in ecology and how we are bringing together completely different worlds.”

However, there are challenges associated with taking on a project consisting of over 35 people. Members of the team must collaborate across their vast cultural differences, in addition to their varying areas of expertise ranging from policy to biology to computer science.  

“Getting us on the same page — there was a lot of investment put into that,” Flecker said. “We had to really work through a whole lot. It is not just that we always agree on everything.”

But for Gomes, the rewards of the collaboration far outweigh any challenges the team had to overcome. 

“As a computer scientist I really feel much better working on this challenging, but very meaningful problem. So I think it’s well worth it,” Gomes said. “We expand our horizons dramatically by interacting with all these disciplines — it’s really, really exciting.”

Flecker explained that the research team inputted information from pre-existing dams and estimations for proposed dams into computer science modeling exercises that then weighed each dam against a set of environmental criteria — greenhouse gas emissions, changes to river flow, sediment trapping, fish biodiversity and connectivity.

“Connectivity is a really big one,” Flecker said. “You can think about it in one way as how fragmented does the river basin become when you put in a set of barriers?”

Disrupting the connectivity of the Amazon River basin can have profound effects on organisms, like fish and freshwater dolphins, whose life cycles depend on migration through different parts of the river, Flecker added. 

After narrowing down the set of environmental criteria, Flecker said the team had to attach a quantitative value to allow for computational analysis of each existing and proposed dam project. According to Flecker, quantification was a major constraint on the scope of the research project, as this forced the team to exclude unquantifiable factors implicated in the dam projects, such as social disruptions and disease transmission.

Gomes and her team’s method of computational analysis was to take the environmental criteria, optimize them for factors such as electricity generation and plot them utilizing visualization tools such as parallel coordinate plots. 

A graph that models the relationships between variables of different sizes and measurements, parallel coordinate plots are essentially a scientific way of ordering a vast array of different dam planning options. This allows for researchers to manipulate multiple environmental criteria and examine their potential outcomes. 

Gomes explained that artificial intelligence was also used to identify combinations of dams that would produce energy with the least amount of environmental disruption — a feat of computation far too vast for a human to accomplish.

The ultimate goal of this project is to provide policymakers with a tool that can aid and inform the decision-making process regarding where to build hydropower dams in the Amazon basin. 

Gomes said that since their research does not include other factors essential for policy-making decisions, like the socioeconomic effects of hydropower dam projects, the goal of the project was to empower South American leaders with the tools to understand the impacts of their policies. 

“We are not making decisions — we just want to have a scientific way of understanding the impact of the dams,” Gomes said. “We are providing the science, the methodology to understand the impacts. But we are fully aware that there are many, many [other] issues that are underlying the decisions.”