How the rating works

Tagged-shark detections, drone/lifeguard sightings and drumline or net captures from SharkSmart in the last 4 days. Full weight within 2 km, fading to zero at 12 km; weight halves every 24 h; repeat pings of one tagged shark count once. A detection or sighting within 3 km in the last 24 h forces HIGH at that beach.

Evidence: Location and recent detections explain about ten times more of white shark presence than all weather variables combined (17.3% vs 1.8% deviance). White sharks cluster: up to five off the same beach on the same day. An alert is strong evidence a shark is present; silence is weak evidence it is absent, because only tagged animals are heard and only within ~500 m of a receiver.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Spaet et al. 2020, MEPS (444 juvenile white sharks, NSW)
• Butcher et al. 2025, Frontiers (same-day white shark clustering)
• NSW SharkSmart program (37+ real-time listening stations)
• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)

Peaks when nearshore water is 16-20.5°C, strongest May-November.

Evidence: NSW drumline catches of white sharks are highest at 15-21°C and decline above 21°C; mesh-net abundance peaks near 17-18°C. This is a good seasonal flag but adds little day to day within the season.

Backtest: On for 49% of 133 past NSW bite days vs 54% of matched days nearby, so bites were about 0.55× as likely when it was on (95% range 0.24–1.10). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Only within-month temperature changes are tested, so most of this factor’s effect is out of reach. Details

• Cappa et al. 2026, Frontiers (NSW SMART drumlines 2022-25)
• Lee et al. 2018, MEPS (25 years of NSW mesh-net data)
• Monteforte et al. 2022 (36,384 drone flights, 42 NSW beaches)
• Spaet et al. 2020, MEPS (444 juvenile white sharks, NSW)

Reduces the white shark score when an upwelling drops water below ~15°C.

Evidence: Tagged juvenile white sharks started searching for warmer water within 10-12 h of upwelling below 14°C and most left within days; Cape Town sightings were 5-8x more likely at 18°C than 14°C. North-easterlies drive this kind of upwelling off Sydney.

Backtest: On for 1% of past bite days, too rarely to estimate an effect. Only within-month temperature changes are tested, so most of this factor’s effect is out of reach. Details

• Spurgeon et al. 2022, Sci Rep (juvenile whites leave upwelling <14°C)
• Weltz et al. 2013, PLoS ONE (Cape Town shark-spotter sightings)
• Swellnet: why NE winds make Sydney water colder (upwelling)

Rises above 20°C and peaks at 22-25°C, October to May. Sydney Harbour mouth makes Manly and Palm Beach end especially relevant.

Evidence: The best-replicated presence relationship on the NSW coast: occurrence on NSW beaches rises above 20°C and peaks at 23°C (233 tagged sharks); in Sydney Harbour only water temperature mattered (52.5% of variability explained). Bull sharks leave once water falls to ~19-20.5°C, and arrivals are now recorded in October.

Backtest: On for 83% of 133 past NSW bite days vs 84% of matched days nearby, so bites were about 0.81× as likely when it was on (95% range 0.24–2.90). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Only within-month temperature changes are tested, so most of this factor’s effect is out of reach. Details

• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)
• Smoothey et al. 2019 (7 years of Sydney Harbour bull shark telemetry)
• Smoothey et al. 2016 (Sydney Harbour bull shark catches)
• Lee et al. 2019, MEPS (bull sharks on the EAC shelf)
• Lubitz et al. 2025 (15 years of Sydney bull shark telemetry)

Bull and tiger sharks come nearshore from late afternoon through the night (strongest ~6pm-1am), with a little extra on dark nights. Full in bull or tiger shark water; white sharks count for 30%.

Evidence: Large tagged bulls on NSW beaches were nearshore from midday to 4am, peaking ~18:00-01:00; the authors recommend "afternoon and low-light" messaging rather than "dawn and dusk". Réunion bulls and tigers were caught more at and after dusk and on dark nights. Juvenile white sharks tagged off NSW showed essentially no daily pattern, so in white shark water this counts for only 30%, as local knowledge.

Backtest: On for 73% of 133 past NSW bite days vs 74% of matched days nearby, so bites were about 0.79× as likely when it was on (95% range 0.29–1.79). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Only within-month temperature changes are tested, so most of this factor’s effect is out of reach. Details

• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)
• Niella et al. 2021 (Réunion bull/tiger catches: turbidity, dusk, dark nights)
• Spaet et al. 2020, MEPS (444 juvenile white sharks, NSW)

Summer-autumn visitor in ~22-26°C water. A small share of NSW nearshore records.

Evidence: Tiger sharks were 43 of 494 SMART drumline captures and absent from 36,384 drone flights; their east-coast movements follow water-temperature change.

Backtest: On for 61% of 133 past NSW bite days vs 59% of matched days nearby, so bites were about 1.41× as likely when it was on (95% range 0.70–3.30). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Only within-month temperature changes are tested, so most of this factor’s effect is out of reach. Details

• Butcher et al. 2025, Frontiers (same-day white shark clustering)
• Monteforte et al. 2022 (36,384 drone flights, 42 NSW beaches)
• Niella et al. 2021, Oecologia (tiger shark movements and SST)

Satellite SST anomaly inshore versus the coast 20-100 km either side, or versus the last 30 days. Applies within the white shark temperature band.

Evidence: The Macquarie thesis behind the "cool pockets" claim is real but its signal was ~0.3°C on a 100 km monthly grid, validated at 66.67% on a balanced test set, and it never linked cool pockets to river mouths or prey. Réunion bite rates were far higher in cooling months. This project tested it at daily 5 km resolution (see Backtest): pockets cooler than the coast either side made no difference to NSW bites, and SharkSmart white shark detections in 2026 leaned slightly higher on 0.1°C+ cooler days but were lower on strongly cooler (0.5°C+) days. Days cooler than the last 30 had fewer bites, probably because cold snaps keep people out of the water. Kept small.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Lynch 2017, Macquarie MRes thesis (SST anomalies and bites)
• Ryan, Lynch et al. 2019, MEPS (environmental bite models, Australia)
• Lagabrielle et al. 2018 (Réunion bites per surf-hour; cooling months)
• NASA JPL MUR SST + anomaly via NOAA ERDDAP
• IMOS OceanCurrent (SST and anomaly maps)

Today's satellite water temperature off the beaches compared with the long-term normal for this date, counted only while the water is in the white shark range. A Hunter shark fisherman's warning: unusually warm water pushes white sharks onto the beaches.

Evidence: In this project’s tests, white shark bites were about 1.8× as likely in weeks 0.5°C+ warmer than the same week in other years (85 bites, 95% range 1.07–3.04), the clearest result among the ocean ideas. But a warm spring also brings more people into the water, which that comparison cannot remove. SharkSmart white shark detections in 2026 showed no day-to-day difference, and fell as the warm water arrived in October. The Macquarie thesis found the opposite (bites at slightly cooler anomalies, on a coarse grid).

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Newcastle Herald, Oct 2026: Hunter shark fisherman warns warm, fast currents push white sharks inshore
• This project’s ocean-idea tests (bites 1986-2022; SharkSmart 2026)
• Lynch 2017, Macquarie MRes thesis (SST anomalies and bites)

Starts at 20 mm and is strong from 45 mm (local, or harbour/Hawkesbury catchment) over the past 1-8 days. Mostly a bull shark factor, so it is scaled by water temperature.

Evidence: Bull sharks were more frequent on NSW beaches after >45 mm of rain; in Sydney Harbour heavy rain moved bulls within a day and their prey within 2-7 days. Not significant for white sharks on NSW drumlines. Tiger bite risk rose with rainfall in national data.

Backtest: On for 29% of 133 past NSW bite days vs 26% of matched days nearby, so bites were about 1.19× as likely when it was on (95% range 0.75–1.86). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Details

• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)
• Niella et al. 2022, Estuaries & Coasts (harbour rain moves bulls and prey)
• Cappa et al. 2026, Frontiers (NSW SMART drumlines 2022-25)
• Ryan, Lynch et al. 2019, MEPS (environmental bite models, Australia)

Beaches near the harbour, a lagoon or Broken Bay score higher when that catchment has had heavy rain, because murky, prey-rich water flushes out along the beach.

Evidence: Distance to a river mouth was a significant predictor of white shark sightings and of white and whaler bites; harbour flushing after heavy rain moves bulls and teleost prey. For one fixed beach the distance is constant, so only the rain-driven plume changes day to day.

Backtest: On for 29% of 133 past NSW bite days vs 27% of matched days nearby, so bites were about 1.16× as likely when it was on (95% range 0.73–1.79). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Details

• Monteforte et al. 2022 (36,384 drone flights, 42 NSW beaches)
• Ryan, Lynch et al. 2019, MEPS (environmental bite models, Australia)
• Niella et al. 2022, Estuaries & Coasts (harbour rain moves bulls and prey)

Satellite Kd490 (light attenuation) nearshore versus its recent baseline. Murkier water favours bull sharks.

Evidence: Bull sharks were more frequent at higher Kd490 in Réunion and water visibility affected KwaZulu-Natal net catches; not significant for white shark drumline catch in NSW. Satellite ocean colour within 1-2 km of shore is noisy, so confidence is low.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Niella et al. 2021 (Réunion bull/tiger catches: turbidity, dusk, dark nights)
• Tate et al. 2021, Fish Mgmt & Ecol (NSW drumline covariates)
• NOAA VIIRS Kd490 / chlorophyll via CoastWatch ERDDAP

Chlorophyll-a well above its recent baseline: a productive-water signal, the kind of input the SharkChess satellite model uses.

Evidence: Productive water feeds plankton and baitfish, which draw every kind of predator, so this applies to all sharks. White sharks searched intensively near productive fronts at shelf scale. The one bull shark study points the other way (bulls on the shelf favoured low chlorophyll), and satellite chlorophyll is unreliable within a few km of the coast. The "90% accurate" South Australian satellite AI (ESpy Ocean) is unpublished and measures movement direction, not beach risk.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Lee/Smoothey et al. 2021, MEPS 659 (white sharks search at 19-23°C fronts)
• Lee et al. 2019, MEPS (bull sharks on the EAC shelf)
• ESpy Ocean "SharkChess" (unpublished; ~90% movement-direction claim)
• NOAA VIIRS Kd490 / chlorophyll via CoastWatch ERDDAP

A small bull shark bonus in 1.8-2.8 m swell.

Evidence: Bull sharks were more frequent on NSW beaches at 1.8-2.8 m swell; swell was not significant for white or bull sharks on drumlines.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)
• Cappa et al. 2026, Frontiers (NSW SMART drumlines 2022-25)

Recent fishing reports of salmon, tailor, mullet, pilchards, jewfish (mulloway) or bait schools close to shore, weighted by recency (3 weeks), place and strength.

Evidence: SharkSmart advises watching for baitfish, diving birds and dolphins, and prey density is one of the biggest drivers of fish behaviour. But no study has quantified prey runs against shark presence on Sydney beaches, and angler reports measure where people fish as much as where fish are.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• SharkSmart safety advice: baitfish, diving birds, dolphins, fishing activity
• Stoner 2004 (what drives fish catchability)
• Lee/Smoothey et al. 2021, MEPS 659 (white sharks search at 19-23°C fronts)

Humpback migration season, plus reported whales, diving birds or bait balls, and dolphin pods near shore, weighted by recency (3 weeks), place and strength. Whale carcasses are a separate, much heavier factor.

Evidence: Official SharkSmart advice lists diving birds and dolphins as warning signs, and white sharks are known to shadow the humpback migration, but no quantitative link to beach presence has been established near Sydney. The ranking (whales, then birds working bait, then dolphins) is local judgement.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• SharkSmart safety advice: baitfish, diving birds, dolphins, fishing activity

A dead or stranded whale reported on or off the Northern Beaches in the last 2 weeks. Full for 5 days, then fading; full at the beach it is on, less with distance along the coast.

Evidence: Whale carcasses reliably draw white and tiger sharks: up to 40 white sharks fed on one carcass in a single day in False Bay, and NSW beaches near carcasses have been closed after many sharks gathered (Royal National Park). Unlike most intel, this is direct evidence of a food source that holds sharks for days, but the pull on beaches several km away has not been measured.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Fallows et al. 2013, PLoS ONE (up to 40 white sharks on one whale carcass in a day)
• City Hub: Royal National Park beaches closed as sharks feed on a whale carcass
• SharkSmart safety advice: baitfish, diving birds, dolphins, fishing activity

Within about an hour of sunrise or sunset.

Evidence: Standard precautionary advice, and dawn/dusk is a real feeding window for many fish. But NSW telemetry shows juvenile whites have essentially no daily pattern and bulls peak from late afternoon into the night. Only 12% of WA white shark attacks fell within an hour of sunrise/sunset, about what chance predicts. Kept as precautionary local knowledge with a modest weight.

Backtest: Can’t be tested day by day, because bite days and their matched days cover the same hours. Separately, 39% of timed NSW bites were within an hour of sunrise or sunset. If people were in the water at the hours surfers use a Gold Coast break (hourly camera counts), you’d expect about 15%; if they were spread evenly, 28%. Either way, more bites happen in these hours than the number of people in the water explains, though the exposure data come from one Queensland break, not the Northern Beaches. Details

• Hansen & Beauchamp, USGS (dawn/dusk antipredation window)
• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)
• WA correlation study of white shark attack risk factors (1996)
• Riley et al. 2022 (Australian Shark-Incident Database)

Local belief that even light rain stirs up runoff and fish near shore.

Evidence: Rain effects in the shark-presence literature come from plume-scale events (45-100 mm+), roughly 20-50x this 2 mm threshold. But in this project's backtest of past NSW bites, days after 2 mm+ of rain were bite days more often than matched days at the same place, and the lean was the same for white and bull/whaler bites. Rain usually keeps people out of the water, so beach crowds are unlikely to explain it. Bite records are a weak yardstick, so this is low confidence, not proof.

Backtest: On for 45% of 133 past NSW bite days vs 37% of matched days nearby, so bites were about 1.43× as likely when it was on (95% range 0.99–2.11). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Details

• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)
• Niella et al. 2022, Estuaries & Coasts (harbour rain moves bulls and prey)

A change of 3 hPa or more over 6 hours, an angling belief about fish feeding.

Evidence: Fails on physics (3 hPa equals a fish moving ~3 cm vertically) and the one robust shark result runs the other way: blacktips moved deeper as pressure fell. The pro-pressure claims trace to angling blogs.

Backtest: On for 75% of 133 past NSW bite days vs 76% of matched days nearby, so bites were about 0.98× as likely when it was on (95% range 0.62–1.59). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Details

• Heupel et al. 2003 (blacktips moved deeper as pressure fell)
• Huveneers et al. 2016 (wind and noise reduce receiver range)

Within about an hour of high tide.

Evidence: No tide effect on large NSW bull sharks, trivial for juvenile whites, and California white sharks were fewer at higher tide. On a sandy beach, high tide mostly moves the waterline. Flood vs ebb at an estuary mouth may matter, untested.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)
• Spaet et al. 2020, MEPS (444 juvenile white sharks, NSW)
• McCauley et al. 2024 (California white sharks: fewer at higher tide)
• USC surf-zone BRUV study (habitat > tide for fish)

Moon more than ~85% illuminated.

Evidence: NSW net catches of all three species fall as moonlight rises, and sightings were 4x higher at new moon in Cape Town. A global analysis of 2,156 bites found about as many lunar effects as chance predicts. The evidence-based version is the small dark-night bonus inside the bull shark night factor.

Backtest: On for 31% of 133 past NSW bite days vs 32% of matched days nearby, so bites were about 0.96× as likely when it was on (95% range 0.65–1.37). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Details

• Lee et al. 2018, MEPS (25 years of NSW mesh-net data)
• Weltz et al. 2013, PLoS ONE (Cape Town shark-spotter sightings)
• French et al. 2021 (moon illumination and 2,156 bites)

Prevailing wind over the last 24 h (speed-weighted average direction) from 45-135°, stronger above 20 km/h.

Evidence: No shark-presence study supports it. At Sydney a north-easterly blows alongshore and drives upwelling of colder water (pushing water offshore), rather than pushing bait onto the beach, and wind reduces acoustic receiver range, so telemetry "wind effects" are partly artefacts. In this project's backtest of past NSW bites, onshore-wind days were bite days more often than matched days, driven by white shark bites (bull/whaler bites leaned the other way). Warm north-east sea-breeze days also bring more people to the beach, which the backtest cannot rule out.

Backtest: On for 27% of 133 past NSW bite days vs 21% of matched days nearby, so bites were about 1.53× as likely when it was on (95% range 0.93–2.58). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Details

• Swellnet: why NE winds make Sydney water colder (upwelling)
• Weltz et al. 2013, PLoS ONE (Cape Town shark-spotter sightings)
• Huveneers et al. 2016 (wind and noise reduce receiver range)

The "grey day feels sharky" heuristic.

Evidence: No study isolating cloud as a shark predictor was found. The feeling is probably driven by murky water after rain, which the heavy rain and turbidity factors capture more directly.

Backtest: On for 27% of 133 past NSW bite days vs 23% of matched days nearby, so bites were about 1.21× as likely when it was on (95% range 0.76–1.92). Not confirmed: the range still includes 1× (no difference), so chance can’t be ruled out. Details

• Tate et al. 2021, Fish Mgmt & Ecol (NSW drumline covariates)

Belief that sharks come closer in when the swell is small.

Evidence: Bull sharks were more frequent in 1.8-2.8 m swell; drones and drumlines only operate in calmer conditions, so small-swell "effects" are partly when the gear can see. Small swell does put more people in the water.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Smoothey et al. 2023 (233 tagged bull sharks, NSW ocean beaches)
• Cappa et al. 2026, Frontiers (NSW SMART drumlines 2022-25)

Satellite water temperature 35-65 km offshore vs right off the beaches, counted only while the inshore water is in the white shark range. The fisherman's idea: warm East Australian Current water offshore squeezes white sharks into the cooler strip along the coast.

Evidence: White sharks have been tracked searching along temperature fronts at shelf scale. In this project’s tests the bite data pointed the other way (bites about half as likely on these days, for white and bull sharks alike), most likely because cooler water at the beach keeps people out of it. SharkSmart white shark detections in 2026 leaned the fisherman’s way (about 1.3× on these days) but not clearly. Kept as local knowledge with a small weight.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Newcastle Herald, Oct 2026: Hunter shark fisherman warns warm, fast currents push white sharks inshore
• This project’s ocean-idea tests (bites 1986-2022; SharkSmart 2026)
• Lee/Smoothey et al. 2021, MEPS 659 (white sharks search at 19-23°C fronts)

The strongest southward surface current up to ~95 km off the beaches today (modelled), with more when the water is also warmer than usual, counted only while the water is in the white shark range. The fisherman's warning: currents far stronger than usual, arriving all at once, crowd white sharks onto the beaches.

Evidence: Satellite current data covers too few past bites to test this (about 30), and 2026 SharkSmart white shark detections showed no difference on strong-current days, or on strong-and-warm days. The daily figure comes from a current model; the tests used satellite altimetry, which arrives too late for a daily rating. Kept as local knowledge with a small weight.

Backtest: Not tested against past bites: the records don’t go back far enough. Details

• Newcastle Herald, Oct 2026: Hunter shark fisherman warns warm, fast currents push white sharks inshore
• This project’s ocean-idea tests (bites 1986-2022; SharkSmart 2026)