Exploring the frontiers of data science, across space and time

Growing up in Wuhan, China, Xiao Huang dreamed of becoming an astronaut. “I watched a lot of ‘Star Trek’ and ‘Star Wars’ movies,” recalls Huang, an assistant professor in Emory’s Department of Environmental Sciences.

During high school, Huang discovered that he enjoyed learning coding and algorithms. His career goal gradually shifted from astronaut to computer engineer, the profession of both his parents.

But while Huang had an affinity for technology, he did not feel highly motivated as an undergraduate at Wuhan University. “I wasn’t a good student,” he says. “I escaped class a lot and played video games most of the time.”

One day an advertisement for an international cultural exchange program caught Huang’s eye. “The translation of the ad from Chinese,” he explains, “would be something like, ‘Explore the world and be brave.’”

Huang spontaneously applied and was accepted into the volunteer program. He spent two months teaching geography to primary school children in Tunisia.

“It was a super cool experience,” he says. “I met other volunteers from across the world. And I really enjoyed getting to know the Tunisian children, talking with them about the cultures of different countries.”

The heartfelt experience helped Huang find his purpose and bring his various interests together in a meaningful way. He began forging a path to a career as a geographer, exploring the frontiers of spatial-temporal data science.

After completing his bachelor’s degree in Wuhan, focused on interpreting satellite imagery, Huang went abroad for graduate school. He received a master’s degree from Georgia Tech’s School of City Planning and Architecture before going on to the University of South Carolina for a PhD in geography.

“I love geographical science and computer technology,” Huang says. “I want to use my knowledge of these fields to help humanity, especially socially disadvantaged communities.”

Huang joined the Emory faculty in 2023. His research interests literally encompass the world, combining satellite views of the planet from space with detailed, on-the-ground data.

He cites what is called “the first law of geography,” posited by Waldo Tobler in 1969: Everything is related to everything else, but near things are more related than distant things.

“When you form a community,” Huang explains, “that community shares something unique. Geography is about finding that uniqueness about different places, looking for patterns in data that reveal local variations.”

Satellite technology is transforming the field of geography. “It’s kind of like being an astronaut in that satellites give you a view of Earth from space,” Huang says.

In addition to large NASA satellites that circle the Earth every 16 days, space contains constellations of small satellites launched by commercial enterprises. It’s now possible to monitor almost the entire surface of the Earth in real time.

Huang can combine high-resolution views from space with data generated on the ground. Sources may include everything from Google Maps View to historical printed maps to information he collects from the field to publicly available datasets related to the environment, human behaviors and health.

Huang then taps advances in artificial intelligence to develop algorithms that efficiently analyze the resulting datasets and identify patterns.

“For example,” he explains, “an expert geographer might have to spend weeks poring over thousands of satellite images to map all the changes over time that led to deforestation of a region. AI can do this same task in seconds.”

The results can provide insights into how to design more equitable cities, improve management of natural resources, prevent or lessen the impact of natural and human-caused disasters, and improve public health policies.

Huang has collaborated on studies yielding useful data on U.S. trends regarding flood risks, settlement in hurricane prone areas, human mobility dynamics during COVID-19 and more. One of his current collaborative projects involves developing a Web-based application for Washington, D.C., to allow wheelchair users to navigate through available sidewalk networks, avoiding sections that may be currently impassable.

The Huang lab is also developing predictive models to help public health officials stay one step ahead of disease-carrying mosquitos.

In Ethiopia, Huang and colleagues are working on a new approach to control an invasive species of mosquito, Anopheles stephensi, which poses a growing threat of urban malaria in Africa. The high-tech, low-cost method combines satellite data with on-the-ground ecological and sociological data. The international research team, led by Gonzalo Vazquez-Prokopec, Emory professor of environmental sciences, aims to zero in on water sources allowing the mosquito larvae to survive the dry season.

Their research has pinpointed small water cisterns used at construction sites as a key habitat for the larvae, particularly water cisterns with algae growth. Satellite imagery gives the researchers a view from space at a resolution of one-third meter, allowing the researchers to map the unique spectral signatures of the cisterns throughout a city.

“We use multispectral satellite imagery to see how cloudy the water is and how much algae it contains,” Huang says. “Algal chlorophyll bounces back near-infrared light, while suspended silt soaks it up. That contrast gives us clear spectral fingerprints, allowing remote-sensing indices — such as turbidity models and the Normalized Difference Chlorophyll Index — to turn raw pixel data into solid water-quality numbers.”

The research team is gathering additional on-the-ground data related to environmental and sociological factors tied to the presence of the larvae. Huang will apply artificial intelligence techniques to develop an algorithm based on this data to rank the water sites from the highest-to-lowest probability for containing stephansi larvae.

The result will map the highest priority sites for treatment with a slow-release, biological larvicide that lasts for six months and is harmless to humans. Local public health workers will have access to this map via a mobile phone app to make their work as efficient and effective as possible.

The Huang lab is taking a global approach for another project aimed at controlling mosquitos. Funded by an $800,000 NASA grant, it draws from a NASA program known as GLOBE (Global Learning and Observations to Benefit the Environment). The GLOBE Program allows people in more than 125 countries to engage in hands-on data collection to increase scientific understanding of Earth.

Huang’s project is tapping the data focused on tracking species of mosquitos. People use the GLOBE Program’s GLOBE Observer app to upload their smart-phone photos of mosquito larvae, along with the dates and the GPS coordinates of where the photos were taken.

The Emory lab members will link available data on weather and the environment — such as temperature, rainfall, altitude, soil type and vegetation cover — to the mosquito data. They will then develop an algorithm to rank the risk for the presence of various mosquito species within different habitats, depending on an array of variables.

The collaborative project also involves scientists from the University of South Carolina who will model the potential for interactions of the mosquitos in the high-risk habitats with people, based on the location of human settlements and other factors.

A team from the University of Florida will build on the data to develop AI models for predicting emerging mosquito habitats, and the potential for outbreaks of mosquito-borne diseases, under different climate scenarios.

“Climate change is a global issue so we’re not just focused on a single city or country for this project,” Huang says. “Mosquito-borne diseases are already a huge problem in many parts of the world. The changing climate, combined with rapid urbanization, may drive the spread of these diseases, including to new areas. We want to give public health officials the tools to help prevent that.”

Story and design by Carol Clark