Onto NASA’s Goddard Space Flight Center


“Interdisciplinarity.” That’s what drew PhD candidate Venkata Shashank Konduri to Northeastern.

Konduri is pursuing a PhD in Interdisciplinary Engineering, meaning he is taking coursework from different disciplines and performing valuable research on complicated, multi-subject problems. “During the course of my PhD, I took classes from various departments like civil engineering, marine science, policy, and electrical engineering, and others. It gives you a chance to interact with professors with different areas of expertise,” he says.

Konduri’s research “consists of elements from AI, machine learning, data science, computing, remote sensing, ecology, hydrology—all of these elements together. Most problems in the real world are interdisciplinary, right? You can’t just be an expert in one field, and solve a real-world problem,” Konduri explains.

The real-world problem Konduri wants to solve relates to vegetation distribution and its environmental drivers. “In order to study the distribution patterns, I use remote sensing data, such as satellite imagery, or in some cases, airborne imagery, collected over different spatial scales. So we could be talking about landscape-level or a county or a state, or even an entire country,” he says. “And, in addition to remote sensing data, I’ve also used methods in machine learning to try to understand how vegetation responds to different environmental conditions and how satellite imagery could be used to detect certain vegetation types on the ground.”

Konduri earned his bachelor’s degree in agricultural and food engineering at the Indian Institute of Technology. After joining Northeastern, he says “I took some graduate courses in the Civil Engineering department in which we focused on using remote sensing data to study vegetation properties. I also took a course in the Electrical Engineering department called machine learning and pattern recognition.” These classes and guidance from his advisor, Professor Auroop Ganguly, helped him complete his first major research project.

“So the USDA or the US Department of Agriculture creates a map of all the crop types that are grown in the U.S. at a resolution of about 30 meters…But the USDA makes this map available to the public only after a few months after harvest,” Konduri says. “The key finding of our study was that using satellite imagery, you can actually monitor crops in near real-time. And you can create a map with a sufficient level of accuracy as early as July or August of that growing season, which is three to four months before harvest and almost six to seven months before the USDA can give you a map,” Konduri explains.

Konduri’s interest and skill in this field earned him the opportunity to work at Oak Ridge National Laboratory where he has been since 2018. His research there examines the spatial distribution of plant communities in Alaska and their environmental drivers. Specifically, Konduri is trying to come up with ways to make high-resolution maps of vegetation using airborne hyperspectral imagery. “Vegetation found in the arctic tundra ecosystems is characterized by high diversity and heterogeneity. Usually, satellite data will be a fairly coarse resolution, meaning if you look at the satellite imagery, it will look pixelated to you. It’s like looking at a low-resolution image… So, we decided to use airborne imagery, which is higher resolution and is free from cloud cover issues… instead of coming from space, the imagery is coming from a plane that was flown over certain regions in Alaska. I use some machine learning tools to create a map of shrub types,” he says.

Through his research, Konduri wants to improve the performance and credibility of earth system models. “You’ll often hear climate scientists making projections about the future. ‘So in the year 2050 or 2100, they predict for example that sea levels will rise, or the temperatures would go up, precipitation patterns might change, extreme events could become more frequent, etc…They make those projections using earth system models,” he says.

An earth system model uses mathematical equations to capture a myriad of complex physical, chemical, and biological processes taking place on our planet. Vegetation plays a critical role, by regulating the exchange of water, carbon, and energy between different components of the earth system such as land, atmosphere and the oceans. “The maps that I’m creating right now would hopefully help contribute towards a better representation of vegetation in these models,” Konduri said. “The insights that I get from my research could also help improve the process-level representation of how vegetation is affected by environment, thus improving the credibility of these models.”

Konduri is looking forward to graduating this spring and moving on to his new role as a research associate at NASA’s Goddard Space Flight Center. Instead of studying the Alaskan wilderness, Konduri will survey savannah ecosystems using spaceborne Lidar imagery.


Related Faculty: Auroop R. Ganguly

Related Departments:Civil & Environmental Engineering