Using artificial intelligence and weather radar to predict bird migration

Artificial Intelligence (AI) is being used more and more recently, and is now being used to predict bird migration using old weather radar images.

Here’s how it works.

What you need to know

  • Most birds migrate at night
  • More than 4 billion birds migrate across North America each year
  • Weather radar can track birds

A mass migration occurs twice a year, in spring and fall. It is estimated that more than four billion birds fly over North America.

In one night, during the peak migration season, experts say more than 500 million people could move. The journey is difficult, as the winged warriors travel hundreds to thousands of miles and must contend with harsh weather, food shortages, and light pollution.

Bright, constant lights can confuse birds and cause them to change their normal migration route, sometimes resulting in death or injury. That’s part of the reason the Federal Aviation Administration (FAA) in 2016 required buildings and towers to have flashing lights.

(Photo by Derek Hotchkins)

This attraction to light is also one of the reasons why Central Park in New York City is one of the best birding spots on the East Coast. A wide range of migratory birds are attracted to the bright lights and use the park’s expansive forests as a resting place.

Another example of the effect of light on bird migration is the annual tribute to lights on September 11. Every September, the migration season, a massive array of floodlights is erected near the World Trade Center site.

On a clear night, the lights can be seen from more than 50 miles away. Ecologists soon noticed many dead or injured birds appearing near the monument. The birds were attracted by the lights and they flew around to the point of exhaustion.

There are now volunteers from the NYC Audubon Society stationed at the floodlights each year. They work in two-hour shifts and count the birds. If the number of birds exceeds 1,000, the lights are turned off for 20 minutes to allow the birds to escape the pull of the display.

(AP Photo/David Ack)

In the past, ornithologists, a scientist who studies birds, tracked bird migration through banding. They would place colorful plastic or metal bands on the birds’ legs and then scientists in other parts of the world would observe the birds and share the species they saw with the bands.

As you can imagine, with the huge number of migratory birds, banding only captures a small percentage of the group’s movements each spring and fall. In addition, they could not accurately predict which nights would see more migratory activity, so they could alert local communities to reduce the use of outdoor lights.

However, in 2018, the science of bird migration took a huge leap forward. BirdCast was born.


This is a project developed and managed by Colorado State University, the Cornell Laboratory of Ornithology, and the University of Massachusetts.

Using artificial intelligence, weather, and historical weather radar data, BirdCast can predict bird migration density, up to three days ahead of time.

Since the 1940s, weather radar has been able to detect things like insects, birds, rain, and snow. As weather radar technology improved, the data was saved in digital form.

Perching birds create a distinctive radar signature when they fly. It looks like a circle on a radar screen.

Animation of the fundamental reversal, starting at 6:10 AM and ending at 7:17 AM on 8/2/2010 (NOAA)

Machines have been taught how to filter out this chaos when tracking storms. This same filter is now used to comb through historical radar records to create decades of migration records.

The computers were taught what birds looked like on radar compared to rainfall. Computer programs were then tasked with reviewing decades of data, removing rain, and uncovering bird migration data. The results showed that the number of birds flying at night depends mainly on wind, temperature, atmospheric pressure and rainfall.

Now that ornithologists know the main factors that influenced migration, they can now use this information to predict future migration.

The next step that experts would like to address is knowing what type of migratory birds they are. The Cornell Lab of Ornithology has created software that can listen to a bird’s call and determine its species.

They hope to incorporate this technology into BirdCast to provide a more complete picture of one of the world’s largest migrations.

Our team of meteorologists dig deep into the science of weather and analyze timely weather data and information. To view more weather and climate stories, check out our weather blogs section.

(Tags for translation)New York State

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *