Our Mission

At WING, we strive to advance the fields of soundscape ecology and remote sensing by harnessing innovative research methods. Our mission is to understand the intricate relationships between bird vocalizations and environmental changes, ultimately to forecast natural disasters in the Chesapeake Bay region.

Photo: Chuck Homler / Focus On Wildlife

How We Push The Boundaries

WING plans to develop a new methodology that opens the gate for understanding and preparing for climate change and its effects in the Chesapeake bay.

Most earlier research on bird sounds looks at what happens after a disruption, or at slow changes in habitats over many years. We want to catch the small changes in a landscape's sounds before a weather event arrives.

To do this, we will place microphones in the field that record continuously, giving us a running log of what the landscape sounds like. We will then match these recordings directly to satellite images. Matching isn't easy - satellite pictures contain a lot of visual "clutter" that makes it hard to tell which part of the picture goes with which recording.

So, we will use a technique called Object-Based Image Analysis, which groups the pixels in a satellite image into meaningful areas like marshes, forests, and shoreline. With the map divided into clear habitats, our computer models can connect each recording to the right piece of land.

Our Impact

Climate change is usually measured with things like sea level and temperature, and the living world is often left out. WING treats birds as living sensors, adding a new way to understand and anticipate environmental change. This matters most for rural and coastal communities in Maryland, which often have fewer weather stations and are more at risk from storms, flooding, and drought.

Our approach uses low-cost equipment and free tools, so it could make early warning available to more people. The results could help government agencies like NOAA and conservation groups like the Chesapeake Bay Foundation prepare for and respond to disasters. Our work also tests a new way to combine very different types of information, which could make environmental forecasting more reliable everywhere.

250+

A review of more than 250 studies found that areas with a greater variety of species tend to have healthier ecosystems. This is why bird communities are a good way to gauge the health of an environment.

2 Billion

The number of bird sightings and observations on eBird, a public database where birdwatchers around the world log what they see. This gives our project a huge amount of data to learn from.

85 Million

The number of bird banding records the U.S. Geological Survey (USGS) holds, which come from tagging and tracking birds over time. These records show where birds travel and how their populations change.

Our Subgroups

Our question touches biology, sound, satellites, and computer science, and no one person can master all of them. So, we divided into four teams, each going deep in one area while relying on the others. The wildlife team tells us which birds and behaviors matter, the sound team turns those behaviors into recordings we can measure, and the remote sensing team shows what actually happened on the ground. The data team then ties it all together and builds the prediction tool.

This is what we call a "multimodal" approach, which simply means using more than one kind of evidence. We do this because every method has weak spots: microphones can miss quiet birds or be drowned out by noise, and satellites can be blocked by clouds. But when a change in birdsong lines up with a change seen from space, we can trust the signal far more, and false alarms become less likely.

Weather Events & Wildlife

Storms, heat waves, and droughts change how birds live, from when they migrate to how they find food and even how they sing. We study Chesapeake Bay birds, from species found almost anywhere to those that depend on salt marshes, since each type may react to the same storm differently.

Remote Sensing

We use free satellite data, including radar that can see through clouds and in the dark, to map plant health, water, and flooding around the Chesapeake Bay. These maps give us an independent record of what happened on the ground, so we can check whether bird sounds were telling us something real.

Ecoacoustics & Soundscape Ecology

We place small microphones in the field to record the sounds of a whole landscape, then use free computer tools to identify which birds are singing and how often. Landscapes get quieter after severe storms, so we want to know if they also change before one, which could make birdsong an early warning sign.

Environmental Data Science

This team cleans and organizes our large, messy recordings and satellite images, accounting for yearly patterns like seasonal migration so normal changes aren't mistaken for warning signs. They then train computer models, programs that learn patterns from examples, to test whether bird behavior can predict environmental disruptions.

Guiding Questions

To keep WING focused on our research goals, we use these questions to guide every step we take.

Can birds truly warn us before a natural disaster hits?
Can we verify what the birds are telling us using remote sensing?
Can we create a predictive model that detects future natural disasters using bid and remote sensing data?
How does this benefit the Chesapeake bay and the communities that rely on it?