How Do You Measure Nature?

All the trees that line our streets and grow in our parks and yards clean our air, protect us from the hot summer sun and lower energy costs by shading our homes… Our urban forest makes our city livable, and we must take care of the nature that cares for us. New York City Mayor Zohran Mamdani 

By now, most Americans know what a heat dome is because many have lived through one. As summers grow hotter and cities search for ways to reduce heat, improve air quality and support public health, nature is having its moment. Trees, parks, gardens and other greenspaces are increasingly being looked to for solutions. 

In New York City, “Vital Parks for All” is investing billions of dollars in neighborhoods to expand access to quality greenspaces through park acquisitions, infrastructure upgrades, new facilities, improved park maintenance and tree planting in its most heat-vulnerable communities.  And the city’s 2026 Urban Forest Plan  hopes to increase its tree canopy to 30 percent, which would put it in league with Singapore, one the world’s greenest cities. 

 But beneath the greenspace enthusiasm lies a more complicated question: Is improving health simply a matter of planting more trees? 

That question brought researchers from across the United States and Europe to the University of California, Davis for a workshop hosted by the Center for Occupational and Environmental Health (COEH). Led by COEH Director Peter James, Sc.D., the gathering assembled experts in environmental epidemiology, health geography, GIS, urban planning, climate, neuroscience, technology and medicine to begin designing a nationwide repository of nature-related data. 

Yet before participants could discuss databases, maps or technology, they found themselves confronting a more fundamental challenge: What exactly is nature? 

For decades, researchers have relied on measures such as tree canopy coverage or satellite-based vegetation indices to estimate exposure to nature. But workshop participants argued that these metrics often capture only a fraction of the story.  

“Green spaces, which are living organisms, grow over time. They die, they change throughout the seasons, they rustle on the winds,” said Ray Yeager, Ph.D., health geographer and assistant professor at the University of Louisville. “Unlike other things, we are co-evolved to be in these spaces.” 

Comparing nature to art, Yeager said we could measure how many times a person views art, how pleasing it was, how far a person lived from an art museum and the quality of the art. But how would we measure art that made us feel something unexpected? And how do we distinguish between good and bad art, and who decides if it’s one or the other?  

Workshop participants noted that nature can take many forms. It might be a forest, park, schoolyard, garden, cemetery, agricultural land or blue space. And whether some landscapes are managed by people and others mostly untouched, both may offer health benefits. The challenge is determining which characteristics of these spaces matter most. 

Data Innovations 

Turning from what the nature repository could house, the workshop focused next on the nuts and bolts of building one. With so much big data out there, constructing a tool that checks all the boxes is a challenge since it needs to be everything all at once: easy to use, fast, accurate, consistent, secure and visually appealing. 

Scarlett Gomez, Ph.D., a UCSF professor of epidemiology and biostatistics and director of the Greater Bay Area Cancer Registry, opened with a question she said she used to field constantly, “Why can't we have cancer data below the county level?”  

In California, county-level data can obscure important local differences, especially in large counties like Los Angeles. But breaking data down to neighborhoods creates a new problem because many areas have too few cancer cases to generate statistically reliable rates. To overcome this, Gomez and colleagues combined census tracts into larger geographic zones with enough people and cases to produce stable estimates. 

“This was a huge advance,” said Gomez. “Multiple states have since taken up the platform and put their own cancer registry data into it.” 

That kind of innovative thinking laid the groundwork for the UCSF Health Atlas, a mapping tool Gomez helped develop that's considered the gold standard in public health mapping. The tool launched in California in 2020 and now hosts more than 200 variables from cancer and heart disease rates, to demographics, income and pollution exposure. The data comes from sources like the Census Bureau, CDC and EPA, among others. Users can layer any two variables to see how they intersect, filter down to a single state, and export the underlying data for their own analysis. 

At UC Davis, GIS Analyst Peter Flores is leading development efforts for COEH’s nature repository using Databricks, a server within the university's secure data environment. The long-term vision includes multiple access points for different audiences, from public-facing dashboards to advanced analytical tools for researchers. 

Databricks is a cloud-based data warehouse that controls who and how people access data, which helps keep it HIPAA-compliant. It can analyze huge data sets quickly and use Notebooks, Python, R, SQL and Scala for data engineering analytics.   

One of the big advantages of Databricks is that since it’s HIPPA-compliant, health data can live in it and be linked through a single server. And since it’s part of UC Davis Health, the health system’s electronic medical records can potentially link to it, too.  

Databricks uses medallion architecture to organize data into three stages—raw (bronze), cleaned (silver) and business-ready (gold)—so data gets more refined and trustworthy at each step. Although it houses data in all its forms, the final product will be off-the-shelf, ready-to-use, gold-level data.  

There also will be multiple access points for different types of UC Davis users, including Databricks SQL Dashboards for non-technical users and Notebooks for data scientists who need to query and do advanced spatial analysis and machine learning.   

External users will access the ESRI Enterprise Portal. “It's a lot of out-of-the-box tools, similar to what UCSF is doing with the Health Atlas,” said Flores. Internal and external users will be able to layer data sets and colors to visualize how they intersect. 

Scientist Anne Neale ended the session with a tour of the EPA’s EnviroAtlas which she helped to design. EnviroAtlas helps users understand the benefits people get from nature—like clean air, clean water and flood control—and how those benefits relate to human health.  

The tool incorporates over 400 data layers nationally, many summarized by 12-digit hydrologic unit codes, or sub-watershed basins, providing roughly 90,000 similarly sized spatial units across the country. Similar to Gomez’s cancer data, it’s another example of a small, consistent geographic unit making environmental and health data meaningful. 

Through this nature repository, our goal is to capture the many dimensions of nature so we can better understand its ‘active ingredients’—what aspects of nature matter most for health, for whom, and under what circumstances.  Peter James, Sc.D. 

 

What Comes Next 

Just 30 miles from UC Davis is America’s greenest city, according to the Treepedia Index. Sacramento, known as the “City of Trees” since its Gold Rush days, has tree coverage averaging an enviable 23.6 percent or almost twice as much as New York City’s.