Advancing geographic research through AI and remote sensing

Le Wang wants people to understand that AI is powerful—but it isn’t in charge.

In his work, Wang, a professor in the Department of Geography, uses artificial intelligence alongside remote sensing data to answer questions that are too large, too complex or too time-consuming for humans to tackle alone. In recognition of his research and accomplishments, Wang received the 2026 Outstanding Contribution Award in Remote Sensing from the American Association of Geographers (AAG). Established in 1992, the award honors AAG members who have made exceptional contributions to remote sensing through research, teaching and/or outreach. This marks the first time a UB professor has received the honor.

“I feel so honored to be recognized by this very important community,” Wang said. “Many of the names on the list of previous recipients are really prestigious researchers and educators in remote sensing, so it means a lot to me.”

The award highlights Wang’s broader effort to harness emerging technologies to better observe and understand our planet. The goal, he says, is to use new tools to generate knowledge that helps people make better decisions in the real world.

“AI is useful if you know how to use AI,” Wang said. “But AI cannot be the boss. You have to be the boss.”

The philosophy of  humans guiding the questions, interpreting the results and deciding what to do next, runs throughout Wang’s research, teaching, and curriculum he is helping develop around AI and remote sensing.

“Although I’m a professor and teach in geography, interdisciplinary work is really the nature of remote sensing. Not only can the images provided by remote sensing work for geographers, but also for environmental science, geology, sustainability, economics, international trade, and different disciplines, different domains, and a lot of real-world applications can be developed from image analysis. At the end of the day, the idea of these images is to show us what the Earth looks like on a daily basis,” Wang says.

Wang used this approach on a project he began in Texas. Before joining UB in 2007, Wang finished his Ph.D. at the University of California at Berkeley and started his faculty career in Texas where he became familiar with tamarisk—an invasive plant that grows along the Rio Grande and crowds out native trees like cottonwood and willow. Wang says that tamarisk is an ecological and water problem. In regions where freshwater is already scarce, its roots allow it to pull more groundwater than many native species.

Wang’s  NSF-funded research used AI and remote sensing to measure how quickly tamarisk was spreading along the river, and to evaluate whether control strategies were actually working.

Efforts to remove tamarisk included burning and cutting. Unfortunately, both approaches proved ineffective. A promising alternative came through biological control – using a small beetle developed by a USDA researcher that causes tamarisk to defoliate. But success in a lab doesn’t always translate to success in the field, and managers faced questions about where the beetles are going, and if they are making a difference.

That’s where satellite data, along with AI’s ability to process it at scale, became paramount. Wang’s team used imagery and AI methods to detect defoliation patterns and assess whether biological control was reducing tamarisk over time.

“I’m proud because we produced knowledge that didn’t exist before,” he said.

Wang shared his findings and highlighted where efforts were working and where additional focus was needed.

Wang’s research also extends to Florida, where he studies mangroves, which serve as coastal forests that function as natural barriers during storms. Wang says that mangroves are uniquely suited for their environment because they are the only woody vegetation that can grow in both salt water and freshwater, and they occupy many coastal zones.

As more people move to coastal areas, Wang notes, the risks from hurricanes and storm surge grow. Mangroves can help buffer those impacts by reducing storm intensity as storms move through dense root systems and forested shorelines.

To protect communities, though, researchers and agencies first need to understand the condition of the mangroves themselves—whether they’re healthy, stressed, dying, or being removed for development. Wang’s team uses AI and remote sensing to monitor mangrove health, anticipate future changes, and identify which coastal areas are sustainable or fragile.

The goal is to provide information that helps decision-makers maintain and restore coastal ecosystems that protect human populations.

After moving to Buffalo, Wang began asking a new question: What challenges can AI and imagery help address locally?

One of the most visible issues, he says, is abandoned housing. When homes sit empty and are usually marked by broken windows, boarded doors and overgrown lawns. The bigger challenge for city officials is that abandonment doesn’t always show up immediately in records. People may leave suddenly because of job loss or other hardships, and the city may not know in real time. Wang and his students are exploring how AI can help. Using Google Street View imagery, Wang’s team developed methods to automatically identify abandoned houses. The result was a map showing more than 2,000 suspected abandoned properties, which is information that could support better planning, faster response, and more targeted interventions.

Once those locations are mapped, Wang says, the analysis goes further to find where properties are clustering, what factors might be driving patterns, and what policies could help prevent the problem from growing.

In a graduate-level AI remote sensing course he developed at UB in 2025, Wang designed a class project around a seemingly simple question: How many trees are on UB’s North Campus?

University Facilities had field survey mapping data from 2021 and were curious about the status in 2025. Students in his GEO 586 (graduate-level course offered in a stacked format alongside its undergraduate counterpart, GEO 486) used current imagery and an off-the-shelf AI tool from Esri, which is made available to students, to update the estimate without writing code. They then presented their findings directly to campus stakeholders and discovered there are approximately 12,000 trees on UB’s North campus.

For Wang, those interactions are the same real-world experiences he is preparing students for.

Wang believes this approach is key to preparing students for a changing job market.

“Students are receiving training through meaningful projects like this, and it helps students develop a stronger skill set,” Wang said.

Remote sensing data has become widely accessible, Wang says, and AI tools are lowering technical barriers. The competitive edge, he says, comes from knowing how to apply and customize AI for specific domains, such as environmental monitoring, urban planning, resource management, and other issues.

Students trained in this space, Wang says, will be well positioned for roles in government and industry, where demand is growing for professionals who can transform large datasets into actionable decisions. AI can accelerate analysis, but it still needs people to decide what matters, what’s needed, and what to do with the results.

“We still teach critical thinking – how to formulate good questions and how to address good questions,” Wang said. “AI is useful, but you have to be the boss.”