Measuring an animal’s speed depends on the animal’s size, environment, and whether you need a quick estimate or precise data. The methods range from simple stopwatch-and-distance calculations to GPS collars, high-speed cameras, underwater acoustic sensors, and wind tunnels. Each approach comes with trade-offs in accuracy, cost, and practicality.
The Basic Principle Behind Every Method
Every speed measurement boils down to the same formula: distance divided by time. What changes is how you capture those two values. For a cheetah sprinting across open ground, you might time it over a known distance. For a fish in the deep ocean, you might track acoustic pings between underwater receivers. For a rat on a treadmill, you might calculate its position frame by frame from video. The technology gets more sophisticated as the animal gets smaller, faster, or harder to observe, but the core math stays the same.
GPS Collars and Tracking Devices
GPS is the go-to method for measuring speed in wild mammals, birds, and other larger animals. A small GPS unit is attached to the animal, often on a collar or harness, and it records the animal’s position at set intervals. Speed is then calculated from how far the animal moved between position fixes.
The accuracy of GPS-based speed measurements depends heavily on how often the device records a position. At low sampling rates, like two fixes per hour, the device only captures where the animal was at those two moments and draws a straight line between them. Any winding, zigzagging, or stopping the animal did in between is invisible. Research on GPS-tracked domestic dogs found that with only two fixes per hour, the actual distance traveled was underestimated by 74%. Even at 12 fixes per hour (one every five minutes), distance was still underestimated by 30 to 64%, with a median position error of about 275 meters.
To get highly accurate path and speed data (position errors under about 8 meters), you need GPS fixes at least once every five seconds. But that kind of frequency drains the battery in roughly a week. Devices recording just two fixes per hour can last over two years. Researchers constantly balance precision against battery life depending on whether they need fine-grained speed data or long-term movement patterns.
One way to improve accuracy without burning through battery life is to combine GPS with accelerometers, which are motion sensors that detect changes in movement and orientation. A technique called dead reckoning uses the accelerometer data to fill in the gaps between GPS fixes, reconstructing a more realistic path. This hybrid approach can bring position errors down to around 26 meters while keeping battery life measured in months rather than days.
High-Speed Video and Camera Tracking
For lab-based studies or smaller animals, video is often more practical than GPS. The idea is straightforward: film the animal moving past a calibrated reference (a grid with known dimensions), then track its position frame by frame. Software identifies the animal’s center point in each frame, and speed is calculated from how far that point moves over time.
Frame rate matters enormously. A standard 30-frames-per-second camera can only detect movement changes longer than about 0.067 seconds, which is too slow to capture the rapid limb movements of a running rodent or the wingbeat of a bird. Research on rat gait analysis found that accuracy broke down at frame rates below 125 fps. Above that threshold, automated tracking software matched the precision of painstaking manual analysis. For very fast movements, researchers use cameras recording at 1,000 fps or higher.
The math behind this relates to a principle from signal processing: your sampling rate needs to be at least twice as fast as the quickest event you’re trying to measure. At 125 fps, the smallest detectable change is about 0.016 seconds. Anything faster than that gets missed.
Outdoors, drones equipped with cameras can follow herds or individual animals across open terrain, recording video that is later analyzed for speed. Stationary trail cameras work for smaller areas but only capture a narrow slice of an animal’s activity.
Wind Tunnels for Birds and Insects
Measuring the speed of a flying animal is trickier than measuring a runner because you need to account for wind. A bird flying into a headwind has a different ground speed (how fast it moves relative to the earth) than airspeed (how fast air flows over its wings). Wind tunnels solve this by controlling the airflow precisely.
In a typical setup, the animal flies in place inside a test section while air moves past it at a known, adjustable speed. Researchers at Lund University in Sweden built a low-turbulence wind tunnel where airspeed varies by less than 1.3% across the test section, with turbulence below 0.04% of wind speed. That level of control means any speed measurement reflects the animal’s actual flight performance, not random gusts.
For insects like mosquitoes, wind tunnels are paired with infrared lighting and multiple cameras recording at 60 fps or higher. The infrared light reflects off the insect’s body and wings without disturbing it, and the cameras capture 3D flight paths that can be analyzed for speed, turning behavior, and response to stimuli like odors or temperature changes. These tunnels are typically small, around 0.25 to 1 meter wide and 1 to 2 meters long, with air velocities between 0.1 and 2 meters per second. Mosquitoes, for example, typically fly in winds of 0.2 to 0.5 meters per second.
Underwater Acoustic Tracking
GPS signals don’t travel through water, so measuring swimming speed requires different technology. Acoustic telemetry is the standard approach. A small transmitter (called a tag) is attached to or implanted in the fish, and it emits sound pulses at regular intervals. Stationary receivers positioned underwater pick up these pulses, and the fish’s position is triangulated from the timing of arrivals at different receivers.
A newer method uses Doppler analysis, the same principle behind the pitch change you hear when an ambulance drives past. As a tagged fish swims toward or away from a receiver, the frequency of its acoustic signal shifts slightly, and that shift reveals its speed. Testing in sea cages showed the technique could clearly distinguish a swimming fish from a stationary reference tag. The stationary tag registered a mean speed of 110 millimeters per second (essentially noise), which was only about 10% of the overall speed range measured for the swimming fish. This matched expected error levels from earlier lab tests, confirming the Doppler readings were genuine.
The practical advantage of Doppler-based speed measurement is that it works with commercially available acoustic tags that need only minor modifications, meaning researchers can integrate it into existing tracking systems without overhauling their equipment.
Accelerometers and Motion Sensors
Accelerometers are small sensors that measure vibrations and changes in motion. When attached to an animal, they pick up the tiny jolts produced by each stride, wingbeat, or tail stroke. Researchers have found that the intensity of this “jiggle,” the high-frequency vibration pattern from the sensor, correlates strongly with forward speed.
In controlled flow-tank experiments on aquatic animals, the correlation between jiggle amplitude and speed was extremely tight, with R-squared values averaging 0.97 (where 1.0 would be a perfect match). In open-water field deployments, accuracy dropped somewhat to an R-squared of 0.82, but that’s still strong enough to reliably estimate swimming speed without needing GPS or visual tracking. When two tags were placed on the same animal, their speed readings differed by only about 0.16 meters per second on average, showing good consistency.
Accelerometers are especially useful for animals that move through environments where GPS and cameras can’t follow: deep-diving whales, burrowing mammals, or birds flying over open ocean. The sensors are lightweight, run on small batteries, and log data continuously for later download.
Cleaning Up Noisy Data
Raw data from any tracking method contains errors. GPS positions can jump around due to satellite geometry or signal interference. Accelerometers pick up vibrations from sources other than movement. Video tracking can lose the animal behind obstacles. Before calculating a reliable speed, researchers run the data through filtering algorithms that separate real movement from noise.
The most common approach for satellite and GPS data uses a Kalman filter, a mathematical technique that compares each new position reading against what it predicts the animal should have done based on previous readings. If a GPS fix suddenly places the animal 500 meters from its last known position in a timeframe that would require impossible speed, the filter flags it as an error and smooths the path to something physically realistic. These corrections are essential for turning raw tracking data into trustworthy speed estimates.