13 animals that can predict the weather — and the science behind it

Folk weather lore outlived the invention of the barometer because it is easy to remember, not because it is accurate. A groundhog's shadow, a cow lying down in a wet field, the width of a caterpillar's brown stripe — these observations were passed down for generations as practical knowledge, and they still circulate every February and every October. Some of them describe something real. Most of them describe a rhyme.

The split matters, because the two categories work in completely different ways. A handful of animals that predict weather have a traceable biological mechanism behind the behavior. Crickets chirp faster as the air warms because they are ectothermic, and their muscle contractions speed up with ambient heat. That relationship is steady enough that a 19th-century physicist turned it into a working formula. Birds shift their perching and feeding patterns as barometric pressure falls, and fish respond to the same pressure drops through gas-filled swim bladders sensitive to fine changes in buoyancy. Frogs call more as humidity climbs because their permeable skin depends on moisture to function, which makes them a live gauge of the air they are sitting in.

Other claims have no mechanism at all. The National Weather Service has stated that a woolly bear caterpillar's banding reflects its age, species, and how much it ate the previous season, not the winter ahead. Sheep huddle in cold, dry weather as readily as they do before rain. Ladybugs cluster where it is warm right now, not where the season is heading.

A third category is worth separating out too: animals responding to something other than weather entirely. The most-cited amphibian study in this space tracked toads abandoning a breeding pond in Italy days before an earthquake, which is seismology rather than forecasting.

This list runs through 13 animals that predict weather, or are said to. Each one is labeled by what the evidence actually supports — measurable biology, plausible behavior, or an inherited folk rhyme with nothing beneath it.

Illustration: Quartz

Groundhog Day is a cultural ritual rather than a forecasting method, and Punxsutawney Phil has never shown predictive skill beyond chance. Each February 2, a crowd gathers at Gobbler's Knob in Punxsutawney, Pennsylvania, to watch a groundhog lifted from his burrow. A shadow is said to mean six more weeks of winter. No shadow is said to mean an early spring.

The ceremony dates to 1887 and descends from Candlemas customs carried to North America by German immigrants, who had attached the same meaning to hedgehogs and badgers in Europe. In Pennsylvania, the groundhog — a woodchuck, Marmota monax — was the available stand-in. The format held, and the holiday now anchors regional spinoffs across the U.S. and Canada, among them Wiarton Willie in Ontario and Staten Island Chuck in New York.

The premise does not survive contact with meteorology. Whether a groundhog casts a shadow depends on cloud cover over one small patch of western Pennsylvania on a single morning, which carries no information about temperatures across a continent over the following six weeks. NOAA has compared Phil's calls against national temperature records and found no meaningful forecasting skill. The groundhog is the marquee name in any roundup of animals said to forecast the weather, and it is also the weakest case in the group.

What groundhogs do have is a real seasonal biology. They are true hibernators, dropping body temperature and heart rate sharply through winter rather than simply denning up, and their emergence is governed by internal circannual rhythms together with rising soil temperature. Males often surface early to scout nearby burrows and locate mates before the breeding season begins, which is part of why a groundhog is awake and visible in early February at all.

That emergence timing is itself a usable climate signal, since the end of hibernation shifts as regional spring warming shifts. It records conditions already underway rather than conditions still to come. The holiday survives on pageantry — the crowd, the top hats, the scroll read aloud at dawn — not on data.

Illustration: Quartz

Cattle turn their hindquarters toward driving wind, and that behavior is the one piece of cow weather lore with a real basis in animal science. Livestock standing in open pasture position themselves to cut exposure, keeping gusts off their faces and out of their eyes, and the herd may swing around as wind picks up ahead of a front.

The rest of the lore is harder to defend. The most repeated version holds that cows lying down in a field mean rain is coming, the idea being that the herd settles in before the sky opens. Another strand keys on which animals walk into a pasture first, with bulls leading the way read as a sign of a storm. A third assigns meaning to the compass direction a cow's tail points, treating one heading as settled weather and the opposite as conditions about to turn.

No mechanism has been documented for any of those three. Cattle lie down for long stretches every day as ordinary routine, resting and ruminating for hours whether or not the weather is changing, which makes recumbency a poor indicator of anything atmospheric. Dairy science treats lying time as a welfare and comfort measure shaped by stall design, flooring, and herd health. Pasture entry order follows social rank within the group. A tail points wherever the animal happens to be standing.

Wind orientation holds up because it is a reaction to a condition already present, not a reading of one still forming. A cow turning its back on a rising wind responds to air it can feel on its hide, the same way it moves into shade during heat or crowds a tree line in sleet. The reaction can come minutes before rain, since gust fronts often run ahead of precipitation, but it describes the weather at hand rather than the weather to come.

Anyone watching a herd pivot together was reading something real. What accumulated around it — tails, bulls, lying down — turned a short-range response to wind into a set of rules it cannot support.

Illustration: Quartz

Frog choruses intensify as humidity climbs ahead of rain, and the reason sits in skin biology rather than any ability to sense tomorrow's sky. Amphibian skin is permeable and needs a moist surface to manage hydration, temperature, and part of the animal's oxygen exchange, which leaves frogs unusually responsive to the water content of the surrounding air.

When a humid air mass arrives before rain, frogs can stay active above ground far longer without drying out. Damp conditions also widen the window for breeding calls, since a frog that would otherwise sit tucked into leaf litter or shallow water can hold an exposed calling position. That is the chorus people in warm, wet regions read as a sign that rain is close.

The distinction worth keeping is that this is measurement, not forecasting. A frog is not detecting a low-pressure system building hundreds of miles away. It is reacting to moisture already present in its immediate surroundings, which often precedes rainfall by a few hours. Amphibian calling works as a short-range local cue, not a multi-day outlook.

Species matter here. Tree frogs, which often live well away from standing water and depend heavily on moist skin to get through a day, tend to show the sharpest jump in calling when the air turns heavy. Ground-dwelling species that rarely leave a pond edge have less reason to react, since their hydration does not hinge on atmospheric moisture in the same way.

Breeding season complicates the picture. Many temperate frog species concentrate calling into a few weeks in spring, and rainfall during that stretch triggers explosive breeding in species that spawn in temporary pools. Wood frogs and spadefoot toads are built around exactly that pattern, moving to ephemeral ponds in large numbers after heavy rain. A chorus heard in April carries different information than one heard in August, and seasonal timing is often doing more of the work than the weather.

Illustration: Quartz

The most cited scientific record of toads sensing something coming involves an earthquake, not weather. Biologists Rachel Grant and Tim Halliday documented common toads abandoning a breeding site in central Italy days before the magnitude 6.3 earthquake that struck L'Aquila in April 2009, and published the findings in the Journal of Zoology.

The data existed by chance. Grant was already counting toads at a lake about 46 miles from the eventual epicenter for a study on breeding behavior and lunar cycles, so daily records covered the period before, during and after the quake. Male common toads usually stay at a breeding site until spawning finishes. These left several days early, and numbers stayed low until the last significant aftershocks had passed.

The researchers made no claim about the mechanism. They suggested the toads might have been reacting to pre-seismic changes in their surroundings: shifts in groundwater chemistry, released gases such as radon, or charged particles produced as rock comes under stress. Geophysicists treat all three as candidate earthquake precursors, but none has been established as a dependable warning signal, and the toad work covered one colony during one event rather than a repeatable experiment.

This matters for a list about animal weather lore because the study gets misquoted so often. It turns up in roundups of creatures that sense storms, which is not what the counts showed. Nothing in the research links toad behavior to rainfall, temperature or barometric pressure. The animals were responding to the ground, not the sky.

Toads do react to moisture in ordinary ways, emerging after warm rain and retreating to damp cover during dry stretches. That is routine amphibian physiology rather than prediction. The L'Aquila observations belong to seismology, and they remain one of the few named, peer-reviewed entries in this whole subject area — worth describing for what they actually found rather than borrowed for a claim they never made.

Illustration: Quartz

A woolly bear caterpillar's stripes say nothing about the winter ahead. NOAA's National Weather Service has addressed the folklore directly, explaining that the width of the brown and black bands reflects the caterpillar's age, its species, and how much it ate during the previous growing season.

The legend is specific enough to sound testable, which is part of why it has lasted. A broad rusty-brown middle band is said to promise a gentle winter, while heavy black at both ends is read as a warning of deep cold and snow. Some versions go further, treating the caterpillar's 13 body segments as a week-by-week calendar of the season to come.

What the banding tracks is growth. The familiar banded woolly bear is the larva of the Isabella tiger moth, and it molts repeatedly as it feeds through late summer and autumn. The brown section tends to widen with each successive molt, so a caterpillar crossing a road in October carries a record of how early it hatched and how much food it found — not a reading of the atmosphere. A dry season that cut short its feeding leaves a different-looking insect than a lush one, whatever January turns out to hold.

Species confusion muddies the picture. Several fuzzy, dark-and-rust caterpillars across North America get called woolly bears, and their natural coloring differs. Two insects collected in the same yard on the same afternoon can look nothing alike for reasons unrelated to weather.

The modern version of the myth traces to C. H. Curran, an entomologist at the American Museum of Natural History, who began collecting and measuring woolly bears at Bear Mountain State Park in New York in 1948 and comparing their bands with the winters that followed. His sample sizes were small, and he treated the outings as an informal diversion rather than controlled research. The work was never replicated at any serious scale, and no research since has established a link between band width and winter severity.

Illustration: Quartz

Cricket chirp rate tracks air temperature closely enough to work as a rough thermometer, and the formula for it has been in print since 1897. That year the physicist A. E. Dolbear published "The Cricket as a Thermometer" in The American Naturalist, setting out how warmer air drives a faster call and how to convert one into the other. The relationship carries his name: Dolbear's Law.

The method needs nothing but a watch. Listen to a single cricket for 14 seconds, tally each chirp, and add 40 to the total. The sum is the approximate air temperature in degrees Fahrenheit. A tally of 30 lands near 70 degrees. Picking one cricket out of a full chorus is the awkward part, and overlapping calls from several insects will throw the count off.

The physiology behind it is straightforward. Crickets are ectothermic, producing almost no internal heat, so their body temperature settles wherever the surrounding air sits. A chirp is made by scraping one forewing against the other, a fast repetitive muscle action. Warm muscle contracts and relaxes more quickly, the wings move faster, and the chirps come closer together. Cool the air and the whole mechanism slows in step.

Dolbear worked with the snowy tree cricket, a pale green species found across much of North America and known for an even, metronomic call that is easy to count. Field crickets, the darker ground-dwelling insects most people hear in a yard, chirp in less regular bursts and give looser readings against the same arithmetic. A count taken from the wrong species is one reason a backyard attempt can come out well off the actual temperature.

The distinction worth keeping straight is that this is a measurement, not a forecast. A cricket reports the temperature at the spot where it is sitting, in that moment, and says nothing about the hours ahead. Among the weather claims attached to animals, it is the rare one with a named author, a dated publication, and arithmetic any reader can check.

Illustration: Quartz

A drop in barometric pressure changes how birds perch, feed, and fly, and this is one of the better-grounded links between animal behavior and approaching weather. Birds have a small sensory structure near the middle ear, the paratympanic organ, which is thought to register slight shifts in air pressure — the kind that arrive ahead of a low-pressure system, hours before the first visible cloud bank.

The behaviors that follow are practical rather than mysterious. Birds often feed heavily in the window before a storm, because foraging gets difficult once rain and wind set in, which is why backyard feeders can turn busy on a heavy, still morning. Flight patterns shift as well. Many species fly lower or stay put as pressure falls, since the air becomes less stable and lift harder to maintain. A row of birds sitting out a gray afternoon on a utility wire reflects that trade-off, not a forecast.

Migration runs on a different clock. Long-distance movement in spring and fall is cued primarily by day length, which is fixed year to year, with temperature and local conditions setting the exact departure date inside that window. Birds frequently wait out bad flying weather and launch on a night with following winds and rising pressure behind a departing front. The Cornell Lab of Ornithology's BirdCast project forecasts nightly migration volume across the U.S. using weather radar, which works because bird movement tracks atmospheric conditions closely enough to model.

Plenty of bird lore has nothing behind it. The old rhyme about crows traveling in pairs meaning clear skies, and a lone crow meaning foul weather, has no mechanism supporting it. Crow grouping is governed by season, food supply, and social structure — flocks consolidate around winter roosts and break up in breeding season, which has no bearing on the next day's sky.

The useful distinction is between sensing and predicting. A bird feeding hard at dawn is reacting to conditions already building around it. That reaction can be read as a short-term signal, but it is not an outlook, and it says nothing about the week ahead.

Illustration: Quartz

Fish often feed more aggressively in the hours before a storm, a pattern tied to how falling barometric pressure alters conditions in the water. Anglers in freshwater and coastal fisheries describe the same sequence. Activity picks up ahead of a front, drops off sharply while it passes, then returns as pressure steadies.

Part of the explanation sits inside the fish itself. Most bony fish carry a swim bladder, a gas-filled organ that holds them in position in the water column without constant swimming. Gas volume responds to pressure. When atmospheric pressure falls ahead of a low-pressure system, a fish makes small buoyancy adjustments to hold its preferred depth. Species that rely on the swim bladder for fine depth control tend to show the more pronounced shifts in behavior. Sharks and rays have no swim bladder at all and manage buoyancy through oil-rich livers and lift generated by their fins.

Food supply changes at the same time. Lower pressure arrives with wind and wave action that disturbs the upper layer of sediment in lakes, ponds, and slow rivers. Fine organic material lifts off the bottom, and insect larvae and small invertebrates get dislodged with it. Baitfish move in to feed on what comes loose, and predatory fish follow them upward. That chain produces the surface activity people connect with coming rain, including the old angling rhyme pairing leaping trout with a wet sky.

The pattern reverses once the storm is overhead. Heavy rain, surface turbulence, and fast temperature swings in shallow water push fish into deeper, calmer layers. Feeding usually slows through the worst of it and resumes afterward.

None of this is forecasting. A fish reacting to a pressure drop is responding to a change already present in its surroundings, minutes or hours ahead of the rain, not sensing a system still a day out. The mechanism is physiology and food supply.

Illustration: Quartz

Sheep bunching together in a field is best explained by thermoregulation and flock instinct, not by any demonstrated ability to sense rain before it falls. An old shepherding rhyme pairs a tightly packed flock with wet ground the next day, and it has survived on rhythm rather than evidence. No one has systematically matched huddling episodes against the rainfall that did or did not follow.

Huddling is routine behavior for a flock species. Sheep are highly social animals that group closely for warmth, and wool-covered bodies pressed together lose heat more slowly than animals standing apart in open pasture. Clustering serves an older purpose too: packing tightly reduces any single animal's exposure to predators, a reflex embedded in prey species that evolved on open ground. Both drivers operate year-round, in clear weather and foul.

The apparent link to rain probably comes from a shared trigger. Temperature often falls ahead of a front, and wind strengthens before precipitation arrives. Sheep answer cold and wind by closing ranks and, where the terrain allows, moving to the sheltered side of a hedge, wall, or slope. A shepherd who watches the flock bunch and then sees rain an hour later has observed a real sequence. What goes uncounted are the cold, dry afternoons when sheep huddle and nothing falls.

That asymmetry is the weakness in the folklore. A behavior that occurs often and is followed by rain only some of the time does not function as a forecast. No agricultural or veterinary research has established sheep huddling as a precipitation-specific indicator rather than a general response to discomfort.

What the behavior does reliably indicate is current conditions. A tightly packed flock in an exposed field means the animals are cold, windblown, or both — useful information for a farmer deciding whether to move stock to shelter. It says nothing about tomorrow.

Illustration: Quartz

Honeybee foraging drops off in the hours before a storm, with workers returning to the hive early and outbound traffic at the entrance slowing to a trickle. Beekeepers have long described this pattern, and the usual explanation involves two things happening at once: falling barometric pressure and rising wind.

Flight is the expensive part of a honeybee's job. A forager leaves the hive, navigates to a nectar source using polarized light, landmarks and an internal sense of direction, and comes back carrying a load that can be a substantial share of her body weight. Gusting wind ahead of a front makes that round trip harder and riskier, and a bee blown off course burns energy she cannot recover from a flower. Staying put is the efficient choice.

Colony-level behavior reflects that calculation. Hive entrances often look busy in the short window before a storm, because foragers already in the field are arriving home at once, and then go quiet as the pressure keeps dropping. Guard bees remain at their posts. Nurse bees inside carry on with brood care, which is insulated from the weather entirely. Once conditions settle after the rain passes, outbound flights usually resume the same day if there is still daylight and enough warmth for flight muscles to work.

Nectar itself responds to weather, which complicates the picture. Rain dilutes or washes nectar out of open flowers, and humid air makes it harder for a colony to cure honey down to a storable moisture level. A colony that stops collecting before a storm is not only protecting its foragers, it is also skipping a low-value shift.

None of this amounts to forecasting in the sense of detecting a system days out. A honeybee is reacting to the wind, light and air pressure around the hive at that moment. What sets bees apart from other insects here is how closely they are watched: commercial and hobbyist beekeepers track daily hive activity, so the pattern has been recorded repeatedly rather than anecdotally.

Illustration: Quartz

Butterflies settle into shelter and cut back on flight as wind picks up and air pressure falls ahead of a storm. The behavior comes down to flight mechanics and body temperature, not any forward-looking sense.

It also plays out differently than it does in colony insects. A butterfly has no hive to return to, so shelter means whatever cover happens to be within reach.

Wing surface area is the central constraint. A butterfly carries large wings relative to its body mass, which makes controlled flight costly in gusty air and close to impossible in sustained wind. As a front approaches and surface winds strengthen, butterflies tend to fly shorter distances and stay lower to the ground. They then tuck themselves under leaves, into tall grass, beneath eaves, or along the sheltered side of a shrub. Resting with the wings closed on a leaf underside keeps those surfaces out of the rain, which matters for an insect that cannot fly well on waterlogged wings.

Temperature compounds the effect. Butterflies are ectothermic, and their flight muscles will not produce enough power until the body has warmed, usually by basking in direct sun. Cloud cover thickening ahead of a storm removes that heat source. The temperature drop that often precedes a front pushes flight muscles below working range. A butterfly grounded on a dim, blustery afternoon may be responding to lost sunlight as much as to anything in the barometer.

Migration runs on a separate clock. Eastern monarchs time their autumn movement toward the oyamel fir forests of central Mexico using day length and seasonal temperature trends built up over weeks, not any single pressure drop. That makes monarch migration a broad seasonal marker rather than a short-range storm signal.

What a resting butterfly indicates is the state of the air it already occupies — the wind, light, and warmth in that one patch of garden. It reads the present moment rather than the hours ahead.

Illustration: Quartz

Ladybug clustering tracks the temperature at that moment, not the weather still to come. The folk saying holds that a swarm of ladybugs means a warm day is on the way, while a shortage of them in the garden signals a cold front moving in. The same behavior drives both halves of the rhyme, and it has nothing to do with forecasting.

Ladybugs are beetles, and most temperate species overwinter as adults. As autumn temperatures fall, they look for sheltered spots to wait out the cold. Loose tree bark, leaf litter and rock crevices all work, as do the wall cavities, attics and window frames of houses. They gather in groups while doing it, sometimes in the hundreds, because clustering helps them hold moisture and heat through months of dormancy. Several species also use chemical cues to find one another, which is why the same sheltered corner of a house can fill with beetles year after year.

The timing of their reappearance is what makes the folklore look convincing. A mild, sunny afternoon in late autumn warms the surface of a south-facing wall, and the beetles sheltering inside become active and visible again. An observer sees a swarm on a warm day and reads it as a prediction. The sequence runs the other way. The warmth came first, and the beetles responded to it.

The scarcity half of the rhyme has an equally ordinary explanation. Ladybugs are ectothermic, and they stop flying once temperatures drop below roughly 55 degrees Fahrenheit. A cold snap can empty a garden of them within hours. Population size also swings from year to year with how much aphid prey the larvae had months earlier. A thin ladybug season may say more about last spring than about next week.

No entomological research establishes ladybug numbers as a seasonal weather indicator. This one belongs firmly on the rhyme side of the list.

Illustration: Quartz

The claim that spiders heading indoors signals an approaching cold snap has no research supporting it, and the autumn jump in sightings has a plainer cause: most common house spiders reach adult size in late summer and early fall. A bigger, more mobile spider is easier to notice, whatever the temperature is about to do.

Males of several species that live alongside people turn much more active in this same window as they wander in search of mates. That wandering is what sends one across a kitchen floor or into a bathtub after dark in September. The timing is set by the species' life cycle, which keys to the season already underway rather than to conditions several days out.

Regional folklore pushes the claim further, holding that black widows in particular retreat indoors ahead of the first hard freeze. Black widows do shelter in protected spots — woodpiles, garages, crawl spaces, the undersides of outdoor furniture — and they handle sustained cold poorly. A black widow in a dark basement corner marks where a warm, undisturbed cavity happens to exist, not a forecast.

The underlying logic holds up on its own terms. Spiders cannot generate their own body heat, so a heated building offers steady temperatures and a break from wind. Houses and garages also stand in for the rock crevices and gaps under bark that spiders would otherwise use. Indoor insect activity runs later into the year than outdoor activity, keeping a food supply available after the yard has gone quiet. An animal following warmth and prey tends to end up inside.

The predictive part is what goes missing. No controlled study has shown that the number or timing of indoor spider sightings tracks a coming cold spell with measurable accuracy. Nor has one shown that spiders register atmospheric cues before those conditions arrive. The common house spider also lives indoors year-round, breeding and feeding in the same walls in every season. That makes a tally of sightings a poor instrument for reading the weather.