Atmospheric Science Book Review technology

Measuring Wind is for the Birds!

Article: Yoshinari Yonehara, Yusuke GotoKen YodaYutaka WatanukiLindsay C. YoungHenri WeimerskirchCharles-André Bostand Katsufumi Sato (2016) Flight paths of seabirds soaring over the ocean surface enable measurement of fine-scale wind speed and direction PNAS 2016 113 (32) 90399044; published ahead of print July 25, 2016, doi:10.1073/pnas.1523853113


Wind is crucial part of all weather systems, and therefore accurate wind measurements are necessary inputs for weather prediction models. Even is an age of satellite weather reports and widespread weather station buoys, we still have trouble measuring wind speeds over coastal oceans well. Satellites that measure wind over the ocean perform well over the open ocean, but inaccurate near the coast. This is due to the high variability of winds in coastal regions, changing over minutes or hours. Satellites usually only cover a specific location once or twice a day, so miss most or all of this variability. Buoys can be placed in coastal areas to get very frequent measurements, but it is difficult and expensive to cover a wide area since they only provide measurements at a specific point.

What if we could utilize denizens of the wind, birds that spend long periods aloft and cover wide areas to measure winds? Recently, researchers Yonehara and colleagues have been experimenting with such a technique.


Fig. 1
Fig. 1: Speed measurements from five minute sections of bird flight (A,B,D,E) were plotted against the birds heading (C,F). After fitting a sine wave (red line) to the measurements, the researchers found the midpoint of the fastest and slowest speeds to find the average wind speed during those 5 minutes. The heading of the maximum flight speed on the curve was taken to be the wind direction.

Inspired by other recent successful uses of animal borne measurement devices, the team equipped three species of birds with small, lightweight GPS trackers. They chose two species of albatross and a type of shearwater as the birds are large enough to carry the tags, live in coastal areas, and take long soaring flights to forage for food. The GPS tags recorded the birds’ positions once every second as long as the batteries lasted, and were removed by the researchers when the birds were recaptured. Using the time between position reports, the speed and direction of flight can be determined.

The real innovation occurs when the scientists assume that the birds fly at a constant speed. This is a fairly reasonable assumption as the species chosen soar and swoop, rather than flap consistently. This also means that they change directions fairly often, which is crucial to the method. By making these assumptions, the scientists can make the final assumption that any changes in the recorded speed of a bird are due to changes in wind speed. If the bird starts flying faster, there is some tail wind; flying directly into a headwind will produce the slowest speeds. Because the birds change direction often, a 5-minute section of flight data will usually show both reduced and enhanced speeds (Fig. 1). When flight speed is plotted against heading, a sine curve can be fitted to the GPS observations. The mid point between the maximum speed and the minimum speed should correspond to the actual wind speed. Wind direction can be obtained by examining the direction the bird was heading during maximum speed periods, where it had a perfect tail wind.


Fig. 2: Bird-measured wind speed is shown in color along the bird track, while arrows represent wind direction. The insets show the bird-measured wind in black arrows, with satellite-measured winds in light grey. The Red arrow is the bird measurement closest in time to the satellite measurement

To test this novel method of wind measurement, the team used bird tracks that were overlapped by satellites. This only accounted for a small portion of the total bird flights; 20 5-minute sections out of over 600 hours of data. However, it was enough to calculate that the correlation between the wind speed and direction together from both sources was very good (Fig 2.). When looking at wind direction alone, the bird tags matched well in strong winds, but not very well in low winds. This may be due to poor satellite performance for low winds near the coast. Looking at speed alone, the bird tags generally underestimated the wind speed compared to the satellites. However, they did follow the same trends as the satellite reported wind speed, so a conversion factor of some kind could probably be used to get the right wind speed.


The results of this study show that wind is another example of a natural feature that sensor equipped animals are well suited to monitor. The bird based measurements provided data in an area that satellites have difficulty measuring, and with much better spatial coverage than a bout or stationary platform could provide. By augmenting existing data sources with this data, weather prediction models could produce more accurate forecasts in coastal areas. Additionally, the GPS tags lend insights to the general behaviors of the birds themselves, and how they behave in different wind conditions.


What other animals do you think could provide helpful measurements to humans? How would you feel about you local weather station providing wind reports from birds?


Leave a Reply

Your email address will not be published.