From predicting wildfires to mapping urban heat islands, the fusion of AI and geospatial data is reshaping how we understand and respond to climate change. Behind these innovations are people like Ryan Kmetz, who are redefining what’s possible when technology meets environmental science.
When Ryan Kmetz reflects on his career, the central thread is his commitment to utilizing data-driven approaches to mitigate environmental impact. He began in biology and medical school, but soon realized that environmental science—and ultimately, climate solutions—were his true focus. This decision launched his journey through graduate studies and roles spanning consulting, sustainability leadership, and data science, each unified by a purpose to drive climate action.
A core driver in Ryan’s story is a restless curiosity—not just to report problems, but to use technology to solve them. While consulting in regulatory compliance, he realized supporting minimum standards was not enough; he wanted to transform how climate challenges are addressed. Transitioning into the intersections of sustainability, geospatial science, and AI, Ryan became focused on combining these fields to create effective climate solutions.
From Biology to AI: A Shift Toward Systems Thinking
Like many climate professionals, Ryan didn’t start in technology. But he always had an interest in computers, and when he stepped into sustainability leadership roles at universities, he quickly realized that enthusiasm and storytelling alone weren’t enough to convince decision-makers to change course. “What I came up against a lot,” he explained, “was that you actually needed the data and analytics to make the case for administration to do X, Y, Z.”
That need for data drew him deeper into Geographic Information Systems (GIS) and remote sensing, tools that combine maps, satellite data, and analytics to uncover patterns invisible to the naked eye. His early projects in urban forestry used AI-powered models to detect tree canopies and build digital maps of urban forests. When paired with LIDAR (laser-based remote sensing), these models could produce 3D views of tree crowns, canopy coverage, and even ecosystem services like cooling or carbon storage.
Ryan’s first step into applying AI for the environment came through GIS. He began by running models that could segment tree canopies and reveal the structure of urban forests, a process that helped him better understand how ecosystems function in cities. That experience sparked a bigger set of questions: If AI can help us map forests, could it also help us anticipate flooding, heat waves, or air pollution events? Could it help us design solutions to prevent the worst outcomes?
Building Tools That Anyone Can Use
Ryan believes that the most powerful climate solutions are not just technical, but accessible. He often asks himself: Can I build something that my mom could use, but also the engineer at the Department of Public Works could use too?
One of his favorite projects, which he calls the GIS Data Explorer, embodies that philosophy. The tool lets users search for topics like flooding or wildfire, automatically pulling relevant datasets from Esri’s Living Atlas, a massive library of geospatial data. It summarizes the data, explains how it might be used, and even suggests related datasets that could enhance analysis.
Instead of expecting people to have programming expertise, Ryan designed it so that anyone, from a city planner to a high school student, could begin exploring geospatial climate data. He took this approach in the development process of this product:
- What is the problem?
- How can we solve it?
- How can we build an effective tool? And that tool should work across different types of stakeholders.
AI Meets Climate: Beyond the Hype
When people hear “AI for climate,” they often think of ChatGPT-style language models. But as Ryan pointed out, those are just one piece of the puzzle. Large language models (LLMs) may make information easier to access, “basically Control-F on steroids,” but the real breakthroughs lie in machine learning applied directly to geospatial and environmental data.
Take air quality forecasting. Ryan sees enormous potential for AI to improve how we predict pollution events, especially as new sources of emissions emerge. One example: data centers. With the boom in AI computing, massive data centers are popping up across the U.S., often powered by natural gas backup generators. “That natural gas combustion has air quality ramifications,” Ryan explained. If those generators run during heat waves or winter inversions, they can worsen pollution during already dangerous conditions.
But machine learning models could help anticipate these scenarios. By forecasting poor air quality days and tying that to data center operations, it might be possible to stagger power usage, switch to battery storage, or avoid running generators at the worst times. It’s a practical example of how AI could directly reduce environmental harm in real time.
The Power of Seeing the Invisible
Ryan lights up when talking about remote sensing. “I’m a total geek,” he admitted. “I love hyperspectral data.” Unlike ordinary satellite imagery (which just captures red, green, and blue), hyperspectral sensors detect dozens or even hundreds of wavelengths—including those invisible to human eyes. This data reveals subtle details about vegetation, water, soil, and buildings.
By combining different indices—NDVI (vegetation health), NDWI (water index), NDBI (built-up areas)—Ryan and his collaborators can detect patterns of flooding, even beyond official FEMA flood zones. “If I can show you 10 years of data suggesting that a building has flooded 18 times, that’s critical information,” he explained. For communities planning development or disaster preparedness, that kind of insight can be life-saving.
In another project, he and his team analyzed U.S. Geological Survey stream gauges to develop machine learning algorithms that could forecast river flooding ahead of time. By marrying streamflow data with satellite imagery, they aim to provide early warnings to prevent tragedies like those seen in Texas after sudden floods.
Climate Solutions in Schools and Communities
Not all of Ryan’s work is technical. During his time at the Maryland Energy Administration, he managed over $24 million in projects to support the state’s goal of net-zero schools. Districts could apply for funding to either build new net-zero schools or retrofit existing ones, often with solar installations. While the work was more administrative, Ryan sees it as part of the larger puzzle: aligning infrastructure, policy, and community needs to achieve climate goals.
Earlier, in his university sustainability roles, Ryan focused on bridging the “town-gown” divide—engaging both campus and community in shared climate solutions. He worked on canopy solar projects, energy retrofits, and air quality monitoring, always emphasizing the importance of storytelling. “It’s not just about saving money,” he told facilities staff. “Switching to LEDs also means no more hazardous waste from mercury in fluorescents.”
For Ryan, data is never just numbers. It’s a way to tell stories that resonate with students, parents, policymakers, and neighbors alike.
The Policy Gap
Of course, even the best tools run into the limits of policy. Ryan has seen firsthand how climate policies often get watered down or abandoned when budgets tighten. “In the climate world, a lot of things are victims of convenience,” he observed. Maryland’s 2022 Climate Solutions Now Act was highly prescriptive, requiring state entities to cut emissions on a strict schedule. But with recent budget deficits, many of those mandates are already being relaxed.
It’s a reminder that data and technology alone won’t solve climate change. Policy, funding, and political will must align, and too often, they don’t.
Skills for the Next Generation
When asked what skills young people should focus on to thrive in the intersection of climate, data, and technology, Ryan emphasized the importance of data storytelling. Technical expertise—like being able to write Python scripts flawlessly—can only take someone so far. What truly makes an impact, he explained, is the ability to translate raw data into a compelling narrative that resonates with people, from policymakers to community members. Being able to show not only what the numbers mean, but why they matter, is the skill that can drive real change.
Ryan also stressed the value of geospatial skills. While artificial intelligence is quickly advancing in many sectors, its application in geospatial analysis is still relatively untapped. This makes it an area full of potential for growth and innovation. For young professionals who are interested in climate and technology, building competency in geospatial analysis could open doors to emerging opportunities where the demand will soon outpace the supply of skilled practitioners.
Equally important, however, is cultivating resilience. Ryan noted that entering this field—especially as someone from an underrepresented background—will come with unique challenges. The reality is that systemic barriers still exist, and they can make the path forward more difficult. But his message was clear: persistence and determination are crucial. The work of driving climate solutions is too important to step back in the face of obstacles. For young people, especially women and people of color, the call is to keep pushing forward, finding strength in both community and conviction.
Looking Ahead
Ryan doesn’t sugarcoat the challenges. He admits that climate progress in the U.S. lags behind Europe and that disasters may need to get worse before policy improves. Yet he also sees hope in the rising influence of millennials and Gen Z, in the growing sophistication of AI tools, and in the communities of practice, from PurpleAir sensor networks to grassroots climate groups, that are making environmental data more accessible.
At its heart, Ryan’s work uses AI and geospatial data to turn environmental challenges into actionable solutions. By making the invisible visible, he empowers communities to understand risks, drive adaptation, and influence policy—reinforcing the idea that climate solutions are as much about people as they are about technology or politics.
If you’d like to learn more about Ryan’s work or connect with him, you can find him on LinkedIn.