How many outbreaks of Highly Pathogenic Avian Influenza in Wild Birds will be reported in the EFSA HPAI dashboard during July 2026?

• My most likely prediction for the number of HPAI outbreaks in wild birds reported in the EFSA dashboard during July 2026 is 40.61, with a 50% chance of falling between 25.53 and 68.75.
• The forecast is primarily driven by the strong seasonal decline of avian influenza typically observed in the Northern Hemisphere during the summer months.
• By early June 2026, major European nations including France and the United Kingdom had already downgraded their HPAI risk levels to ‘negligible’ or ‘insignificant’ following the end of spring migration.
• Historical data from July 2025 shows sporadic weekly outbreaks, supporting a monthly total in the low-to-mid tens.
• A significant tail risk exists (captured in the 80% HDI up to 108.95) due to the record-breaking intensity of the 2025-2026 winter season, which may lead to higher environmental persistence or localized summer die-offs in seabird colonies.
• Data from the EFSA dashboard integrates reporting from ADIS and WAHIS, and the final resolution will depend on these systems’ updates through August 12, 2026.

Key figures

Historical context

Historically, HPAI in Europe follows a strict seasonal cycle. Peaks occur between October and March, driven by the migration of waterfowl. The summer months (June, July, August) are almost always the annual nadir. For example, in July 2025, reports were sporadic, with some weeks showing fewer than 10 total outbreaks across the entire EFSA reporting area. However, the pattern changed slightly in 2022 and 2023, where ‘summer persistence’ was observed for the first time on a large scale. During those years, the virus did not disappear but instead caused significant mortality in summer-breeding seabird colonies, such as Sandwich Terns and Gannets. This historical shift means that while July is expected to be quiet, a ‘zero-outbreak’ scenario is now considered highly unlikely compared to the pre-2021 era. The 2025-2026 season was particularly notable for being the most active on record, setting a high baseline of environmental contamination leading into the summer of 2026.

Tailwinds

• The 2025-2026 winter season was 3 to 5 times more active than previous years, potentially leaving a much larger viral reservoir in the environment.
• Recent May 2026 detections in gulls in Spain and mammals in Norway (Svalbard) prove the virus is still actively circulating in wild populations.
• Historical “summer persistence” seen in 2022 and 2023 suggests that modern H5N1 strains are better adapted to surviving and spreading in summer breeding colonies.
• Improved surveillance and the use of the ‘Avian Check’ app in countries like Ireland may lead to higher reporting rates for the few cases that do occur.

Headwinds

• The ‘negligible’ risk ratings recently assigned by France and the UK suggest that authorities expect almost zero transmission in the coming weeks.
• Higher temperatures and increased UV radiation in July typically reduce the environmental stability and survival of the influenza virus.
• The conclusion of spring migration means bird populations are less mobile and less likely to spread the virus across large geographic distances.
• Many European countries have implemented stricter biosecurity measures and vaccination pilots which may have reduced the overall viral reservoir.

Detailed reasoning

My analysis of the Highly Pathogenic Avian Influenza (HPAI) situation in Europe suggests that July 2026 will be characterized by a significant seasonal recession, though not a total disappearance of the virus. Historical epidemiological data from the European Food Safety Authority (EFSA) consistently shows that HPAI outbreaks in wild birds peak during the autumn and winter months, following migratory patterns, and reach their lowest point in the summer (June through August).

Recent reporting from early 2026 provides a critical baseline. The 2025-2026 winter season was exceptionally severe, with over 2,100 detections in wild birds between November 2025 and February 2026. This was approximately three to five times higher than previous years. However, by April and May 2026, the trend followed the expected seasonal downward trajectory. Leading agricultural authorities, such as the French Ministry of Agriculture, downgraded HPAI risk levels to ‘negligible’ or ‘insignificant’ by early June 2026, citing the end of spring migrations and improving sanitary conditions.

To reach the predicted most likely value of 40.61, I weighted the following factors:

1. Historical July Baselines: Data from July 2025 indicated sporadic weekly reports, with approximately 6 outbreaks noted in a single mid-July week. Extrapolating this suggests a monthly total in the range of 25 to 40 cases under ‘normal’ summer conditions.
2. Residual Viral Load: Because the preceding 2025-2026 season was record-breaking, the environmental viral load is likely higher than in previous years. This residual presence supports a forecast that sits slightly above the absolute historical minimums.
3. Summer Persistence Risk: Historical precedents from 2021-2023 show that HPAI can sometimes persist through the summer, particularly in breeding colonies of seabirds and gulls. While official risk assessments are low, continued detections in Spain (Galicia) and Norway (Svalbard) as late as May 2026 indicate the virus has not been eradicated from the wild population.

The distribution is modeled using a mixture that captures two primary scenarios: a “quiet summer” scenario centered around 35-45 outbreaks, and a “moderate persistence” scenario where localized clusters in seabird colonies push the count toward 75-100. The resulting median of 40.61 reflects a high confidence that July will remain a low-activity month, but the 80% HDI extending to 108.95 accounts for the volatility inherent in wild bird reporting and the potential for unexpected summer flare-ups. This forecast balances the strong seasonal decline observed in late spring with the unusually high baseline established over the previous winter.

Key uncertainties

• The extent of “summer persistence” in seabird colonies is the largest unknown; a single colony outbreak can result in dozens of reports.
• Delays in reporting between national veterinary services and the EFSA dashboard can shift counts between June, July, and August.
• The presence of the H5N5 variant in the Arctic (Svalbard) introduces a new strain with unknown summer transmission characteristics in Europe.
• Potential for “reporting catch-up” where outbreaks from late June are officially entered into the system in early July.

Conclusion

• July 2026 is expected to see a seasonal low in HPAI outbreaks (nadir), though residual virus from a record-breaking 2025-2026 winter season will likely keep numbers above historical minimums.
• Decision-makers should monitor for “summer persistence” events, such as large-scale die-offs in seabird colonies, which could rapidly push the outbreak count toward the upper bound of the forecast.
• The 50% HDI of 25 to 69 reflects a transition from spring recession to a stable summer baseline, with the possibility of localized spikes.

Question: how many wild-bird HPAI detections will EFSA’s dashboard count for July 2026 record (observation-date, dashboard value by 12 Aug)? Forecasters agree this is best treated as a regime-mixture forecast anchored to past summers: very low (~19/month) as in 2024, moderate (~50–60/month) as in 2025, or very high (>200/month) as in the 2022–23 seabird-persistence summers. All teams note the 2025–26 winter wave was unusually large but declined sharply by early June, and the critical near-term determinant is whether infections shift into colony-breeding gulls/seabirds (which would sustain high July counts) versus continuing decay toward a summer lull. They also flag reporting-lag issues (observation date vs submission date, WOAH/GB delays) that could push July counts up by the Aug 12 cutoff. Central estimates cluster around a moderate July count (~50–80) with meaningful probability of a 2024-like lull (roughly 20–25%) and a fat right tail (combined ~25–30% for elevated to high persistence, with ~4–5% explicitly above ~220). The teams largely agree on structure and drivers; their minor differences are mainly in exact bin probabilities and how much mass to place in the upper tail.

Forecast and Reasoning

Forecast: Probability distribution:

• 1.00% chance of value below 20.5
• 5.00% chance of value below 25.0
• 10.00% chance of value below 30.0
• 20.00% chance of value below 40.0
• 30.00% chance of value below 50.0
• 40.00% chance of value below 62.0
• 50.00% chance of value below 80.0
• 60.00% chance of value below 95.0
• 70.00% chance of value below 115.0
• 80.00% chance of value below 150.0
• 90.00% chance of value below 225.0
• 95.00% chance of value below 255.0
• 99.00% chance of value below 270.248

Forecast Reasoning

We need the count of HPAI outbreaks in wild birds on the EFSA dashboard with observation dates during July 2026. All analysts agree July is typically in a seasonal trough compared with autumn/winter, but there is a key disagreement about the probability of an anomalous “seabird-colony summer wave” (seen in some recent years) that could push monthly totals into the 200+ range.

Forecasters 1–3 anchor on recent low/moderate summers (e.g., 2024 low; 2025 moderate) and put the median around ~60–85 with high percentiles mostly below ~200, implying little probability above the platform’s in-range upper cap. Forecaster 4 emphasizes that some summers (notably 2022/2023) produced Europe-wide wild-bird totals around or above ~200/month and argues the extraordinarily large 2025–26 season could seed a renewed large summer wave; they therefore assign substantial open-upper mass and a much higher median.

My synthesis: (1) keep the center of the distribution in the “summer trough” regime (so median closer to the 70–90 band suggested by forecasters 1–3/2), because extreme winter activity does not reliably translate to extreme midsummer counts and EFSA reporting often notes declines toward spring; but (2) retain a meaningful, not dominant, upper tail for a 2022/2023-like seabird amplification regime. Concretely, I model this as roughly ~10–20% chance July 2026 exceeds the in-range cap (so p90 is placed past the upper bound), while most mass remains below ~170.

A typical amateur forecast would either (a) anchor on “summer is always quiet” and place almost no probability on 200+ months, or (b) extrapolate from the huge 2025–26 winter and set an overly high median, implicitly assuming high transmission persists unchanged into midsummer. My forecast explicitly blends strong seasonality (keeping the median moderate) with regime uncertainty from recent history (assigning a non-trivial open-upper tail), which should be better calibrated than a single-regime extrapolation. Confidence: moderate—tail risk is real but hard to quantify without directly observing early-summer 2026 dashboard trends.

Forecast rationale (numeric):

— Iteration 1 — Overall, the forecasts converge on a moderate July 2026 outbreak count, with the center of mass in the mid-range rather than at either extreme. The shared reasoning is that July is usually below the winter HPAI peak, but wild-bird activity remains meaningful in summer, especially through seabird colony outbreaks and other localized events.

Shared reasoning patterns

• Seasonality matters: July is treated as a transitional or lower-activity month compared with the winter peak, which pulls expectations down.
• Background activity persists: None of the forecasts expect very low counts, because HPAI remains established in wild birds and summer detections continue.
• Seabird colonies are the main upside risk: A major recurring theme is the possibility of a large seabird epizootic, which could push counts sharply higher.
• Recent trend is somewhat lower than earlier panzootic years: Several rationales note declines in 2024–2026 relative to 2021–2023, especially in Europe, which supports a more moderate baseline.
• Reporting timing affects the count: One forecast explicitly discounts for the August 12 reporting cutoff, and others implicitly account for reporting lags, making the effective July total somewhat uncertain.

Areas of consensus

• The most likely outcome is not a near-zero month and not an extreme spike.
• The distribution is right-skewed: a typical July is moderate, but there is a meaningful chance of a much larger count if a major seabird event occurs.
• The bounding guidance suggests the count is expected to be well above the low end and usually below the high extreme.

Main differences

• The forecasts differ mostly on the central estimate:
  • one centers around the mid-50s,
  • another around 70–75,
  • another around 82.
• The lower estimate puts more weight on declining trends and reporting cutoffs; the higher estimates emphasize persistent endemic/background activity and the possibility that summer still produces substantial outbreaks.

In short, the collective view is that July 2026 should see a moderate number of wild-bird HPAI outbreaks, likely in the tens to low hundreds, with substantial upside risk from seabird colony events and some downward pressure from seasonal decline and reporting timing.

— Iteration 2 — Across the forecasts, the main reasoning is highly consistent:

• Seasonality is the dominant driver. July is treated as a natural low point for HPAI in wild birds, so all models expect a sharp decline from the very high winter 2025/26 level.
• The recent winter wave sets the starting point. All models anchor on the unusually severe winter intensity, roughly 706 wild-bird outbreaks per month, and then apply a large summer reduction rather than assuming a normal year.
• Summer decline is expected, but not collapse to zero. The shared view is that July should still show meaningful activity because the virus has remained unusually persistent in wild bird populations post-2022.
• Reporting lag matters. Several rationales explicitly lower the July count a bit because the dashboard is observed with a deadline in mid-August, so some July events may not yet be reported.
• Wide uncertainty is warranted. Each forecast uses broad tails to allow for either:
  • a stronger-than-expected summer fade-out or lag-induced undercount, or
  • an anomalous summer persistence event, especially in seabirds or colony settings, similar to prior off-season surges.

Areas of consensus

• A substantial drop from winter to July is expected.
• The most likely July count is well above minimal levels, but far below the winter peak.
• The distribution should be wide and asymmetric, with meaningful probability in both the low and high tails.
• The plausible range is broadly bounded by the guidance near ~20 to ~220 outbreaks.

Main differences

• Central estimate: one forecast leans higher at about 100, while the others cluster lower, around 70 or a core range centered roughly in the mid-double digits to low hundreds.
• Strength of summer persistence assumption: the higher estimate places somewhat more weight on continued elevated activity; the lower estimates emphasize a more typical seasonal trough with only moderate carryover from winter.

Unified takeaway

The collective reasoning expects July 2026 to be a seasonal trough following an exceptionally severe winter, but with enough ongoing circulation and reporting uncertainty that the outcome could still vary widely. The forecasts converge on a moderate-to-low outbreak count relative to winter, with broad uncertainty and a nontrivial chance of both much lower and much higher values than the center estimate.

— Iteration 3 — The forecasts converge on a moderate July 2026 outbreak count, shaped primarily by strong seasonal decline in HPAI during summer but tempered by the fact that wild-bird H5N1 does not disappear in warmer months.

Main reasoning patterns

• Seasonality dominates: July is treated as a low-activity month compared with winter and spring, so counts are expected to be substantially below peak-season levels.
• Recent trend matters: the reported decline in detections through early 2026 pushes the center of the forecast downward.
• Summer seabird risk prevents a near-zero view: all rationales note that seabird or gull colony outbreaks can produce sudden summer spikes, sometimes substantially elevating counts.
• Uncertainty is asymmetric: the distribution is seen as right-skewed, with a modest baseline but a meaningful chance of unusually large outbreak clusters.
• Reporting window effects: the August cutoff/reporting lag may mean some late-July cases are not fully reflected, slightly constraining the observed July total.

Areas of consensus

• A central estimate in the roughly 55–75 range is the shared view.
• The 80% range is wide, reflecting uncertainty around whether July stays quiet or experiences a seabird-driven surge.
• Most reasoning implicitly uses the provided bounding guidance as a rough anchor, with lower tails near ~20 and upper tails extending well above 200 in extreme scenarios.

Main differences

• The forecasts differ mainly in how much weight they assign to extreme summer outbreaks:
  • some lean closer to the high-50s/low-60s,
  • others slightly higher around the mid-70s,
  • but all retain a substantial upper tail for rare multi-country colony events.

Unified takeaway

July 2026 is expected to be a low-to-moderate HPAI month in wild birds, lower than winter levels due to seasonality and the early-2026 decline, but still vulnerable to episodic seabird outbreak clusters that can cause sharp upward deviations.

Synthesis of Rationale Reasoning

To forecast the number of Highly Pathogenic Avian Influenza (HPAI) outbreaks in wild birds reported on the EFSA dashboard for July 2026, the rationales universally weigh competing epidemiological, environmental, and administrative factors. The consensus highlights a tension between the virus’s shift to year-round endemicity and severe structural reporting constraints.

Epidemiological Trajectory and Host Immunity The rationales agree that while the preceding winter season experienced an exceptionally severe, record-breaking wave of infections, the overall trajectory saw a precipitous seasonal decline throughout the spring. Because the H5N1 virus (clade 2.3.4.4b) is now endemic, summer cases no longer drop to zero but are instead largely driven by localized seabird colonies. However, evidence of acquired herd immunity and an observed lack of early-summer mass mortality events suggest that an explosive, panzootic summer wave is unlikely, keeping the true baseline of cases relatively moderate.

Environmental Drivers: The Heat and Drought Paradox Forecasters heavily emphasize the dual, conflicting impacts of the extreme summer weather forecast for Europe:

• Viral Suppression: Unprecedented heatwaves and high UV radiation act as environmental disinfectants, rapidly degrading the virus outside of a host and suppressing broad geographic spread.
• Host Aggregation: Conversely, severe drought conditions shrink inland wetlands, forcing waterfowl into dense aggregations around limited remaining water sources. This clustering creates ideal conditions for highly localized, rapid transmission spikes among susceptible birds.

Surveillance Practices and Structural Aggregation The raw count of outbreaks will be influenced by shifts in testing protocols. Several rationales note that “surveillance fatigue” and updated protocols have led authorities to cap the testing of individual dead birds and to spatially group cases (e.g., regional aggregation), which artificially suppresses the number of distinct outbreaks listed. Conversely, new European active surveillance networks systematically swabbing live birds provide a counterweight, ensuring a persistent baseline of asymptomatic or mild detections.

The Reporting Lag Bottleneck The most critical limiting factor identified across all rationales is the strict August 12 resolution deadline. Because wild bird passive surveillance requires carcass discovery, laboratory sequencing, and bureaucratic processing through international databases (ADIS/WAHIS), there is an inherent reporting delay of several weeks. The rationales universally agree that this narrow 12-day window following the end of July will severely truncate the data, meaning a substantial fraction of true late-July outbreaks will simply not be processed and visible on the EFSA dashboard in time to be counted.

SUMMARY

Question: How many outbreaks of Highly Pathogenic Avian Influenza in Wild Birds will be reported in the EFSA HPAI dashboard during July 2026? Final Prediction: Probability distribution:

• 10.00% chance of value below 30.77
• 20.00% chance of value below 40.203333
• 40.00% chance of value below 60.27
• 60.00% chance of value below 86.27
• 80.00% chance of value below 138.603333
• 90.00% chance of value below 191.936667

Total Cost: extra_metadata_in_explanation is disabled Time Spent: extra_metadata_in_explanation is disabled LLMs: extra_metadata_in_explanation is disabled Bot Name: extra_metadata_in_explanation is disabled

Report 1 Summary

Forecasts

Forecaster 1: Probability distribution:

• 10.00% chance of value below 34.51
• 20.00% chance of value below 45.51
• 40.00% chance of value below 69.51
• 60.00% chance of value below 100.51
• 80.00% chance of value below 167.51
• 90.00% chance of value below 238.51

Forecaster 2: Probability distribution:

• 10.00% chance of value below 30.3
• 20.00% chance of value below 41.6
• 40.00% chance of value below 64.8
• 60.00% chance of value below 93.8
• 80.00% chance of value below 149.8
• 90.00% chance of value below 202.8

Forecaster 3: Probability distribution:

• 10.00% chance of value below 27.5
• 20.00% chance of value below 33.5
• 40.00% chance of value below 46.5
• 60.00% chance of value below 64.5
• 80.00% chance of value below 98.5
• 90.00% chance of value below 134.5

Research Summary

Research summary (from searches completed up to 6 June 2026):

The research found that, as of early June 2026, HPAI activity in Europe was reported to be declining from the peak of the 2025–2026 season: several national reports and news items noted reductions in wild-bird detections and actions such as France downgrading risk to “negligible” (4 June 2026) and Spain lifting some poultry confinement measures in spring 2026. National reports referenced continuing but fewer wild-bird detections in countries that experienced heavy activity in 2025–2026 (examples in the searches included Poland and Galicia/Spain). EFSA’s HPAI dashboard aggregates reported outbreaks from ADIS and WOAH/WAHIS and was used as a primary target for historical counts; one aggregate number found in the reviewed material cited 4,089 wild-bird cases reported across 29 European countries for the 1 August 2025–4 January 2026 interval. Seasonal context from the sources indicates most HPAI activity historically concentrates in autumn–winter migration periods, with lower detection rates reported in summer months as northward migration ends.

The searches for prediction markets and open forecasting markets (Metaculus, Manifold, Polymarket and related fora) did not turn up markets with clear, high-liquidity contracts specifically tied to “number of HPAI wild-bird outbreaks reported in the EFSA HPAI dashboard in July 2026.” The research therefore focused on base-rate and reference-class material from official surveillance dashboards and national reports to establish historical frequencies and seasonal patterns; while EFSA/WAHIS/ADIS contain outbreak-level historical data that can be queried for July counts in prior years, no single published summary figure for “July outbreaks per year” was identified in the reviewed material without extracting raw dashboard records.

Sources used (sites searched / referenced during this research):

• EFSA HPAI dashboard (European Food Safety Authority): https://www.efsa.europa.eu/en/interactive-pages/hpai-dashboard
• WOAH WAHIS (World Organisation for Animal Health disease information system): https://wahis.woah.org/
• ECDC — avian influenza page (European Centre for Disease Prevention and Control): https://www.ecdc.europa.eu/en/avian-influenza
• Metaculus (forecasting community): https://www.metaculus.com/
• Manifold Markets (prediction market platform): https://manifold.markets/
• Polymarket (prediction market platform): https://polymarket.com/
• Various national ministry and news reports referenced in searches (e.g., French Ministry of Agriculture press items, Polish/Spanish regional reports and national disease updates; these were consulted via news search results and national websites).

RESEARCH

Report 1 Research

I’ll help you research this HPAI forecasting question. Let me gather relevant news, historical data, base rates, and prediction market information through systematic searches.

Forecast — Post 432

Modal: 45-70 (42.0%) • partial_shift • 10m14s

Interpretation

Reading: strict

The resolution criteria are highly specific: exact dashboard filters are enumerated (bird category, subcategory, virus subtype, date type, and precise date range), and the resolver explicitly requires a discrete integer visible on the named source — qualitative or inferred values trigger Ambiguous resolution. This precision leaves little room for interpretive latitude, warranting a strict reading.

Time bounds

• observation_window: HPAI outbreaks with observation dates 2026-07-01 to 2026-07-31 inclusive, as visible on the EFSA HPAI dashboard as of 2026-08-12T12:00:00+00:00
• resolution_date: 2026-08-12

Edge cases identified

1. Dashboard may show aggregated total or per-country breakdown — summing all country rows is required if no single aggregate is shown.
2. Data may be marked ‘preliminary’ or ‘provisional’ at the resolution deadline — this is acceptable for resolution.
3. If both raw and adjusted/final figures are displayed, the adjusted/final figure takes precedence.
4. If multiple conflicting integers appear for the same metric without clear distinction, the largest integer is used.
5. Wild bird subcategory ‘All’ must be selected — stricter subcategory filters (e.g., waterfowl only) would yield different counts.
6. Observation date vs. report/notification date distinction is critical — only outbreaks with observation date within July 2026 are counted, not those merely reported in July.
7. UK (excluding Northern Ireland) data sourced from WOAH rather than ADIS — it is unclear whether the dashboard automatically integrates this or if manual consultation is needed.
8. If the dashboard is unavailable at the resolution deadline, ADIS, WAHIS, or official EFSA reports serve as alternative sources, but their count definitions may differ slightly.
9. Late-reported outbreaks with July observation dates could appear on the dashboard after July 31 and before the August 12 deadline, increasing the count.
10. The answer range (19.5–220.5) suggests the resolver expects values in the tens to low hundreds, consistent with historical HPAI wild bird surveillance in Europe.

Research (3/3 variants, shared evidence pool)

Total evidence registered (shared pool): ?

Research Brief

Evidence confidence: medium

Scenario 1: Very low summer trough — comparable to summer 2024 (roughly 15–30 outbreaks) [medium evidence]

Conditions favoring

If the 2025–2026 season follows 2024 summer seasonality (only 59 wild bird detections across all of Europe June 15–Sept 20, 2024 = ~15–20/month), July 2026 would fall in the same very-low range. The sharp decline trajectory from March–April 2026 (sources 10, 11, 13) continuing into summer supports this. WOAH global July 2025 data showed only 38 non-poultry outbreaks worldwide (source 3), further supporting the summer floor scenario.

Conditions against

Summer 2025 was notably higher than 2024 (source 7: ~52/month vs ~15–20/month in 2024). If the virus continues at an elevated baseline due to persistent strains, counts could exceed the 2024 floor. The 2025–2026 season has been far more intense than 2024–2025.

Scenario 2: Moderate summer activity — comparable to summer 2025 (roughly 40–70 outbreaks) [high evidence]

Conditions favoring

Summer 2025 averaged ~52 wild bird detections/month across 15 European countries (source 7), with 112 cases across Europe in approximately 10 weeks (May 12–July 21, 2025, source 6). If 2026 follows 2025’s elevated summer baseline (attributable to persistent H5N1 strains, year-round seabird activity), July 2026 could fall in the 40–70 range. The 2025–2026 season has shown a higher overall intensity, which may keep the summer floor elevated.

Conditions against

Summer 2024 was notably lower (15–20/month), suggesting 2025 was an outlier. The exceptional 2025–2026 winter season could exhaust susceptible wild bird populations, potentially reducing summer 2026 counts below 2025 levels.

Scenario 3: Higher residual activity — above typical summer baseline (roughly 70–130 outbreaks) [low evidence]

Conditions favoring

The 2025–2026 season cumulative total of ~5,905–7,067 wild bird cases by mid-May 2026 (sources 9, 12) is substantially above prior seasons. If residual outbreak activity is slow to decline or if new introductions occur in summer 2026, July counts could remain well above the summer norm. The spring 2026 decline trajectory is steep but started from a very high baseline.

Conditions against

Historical summer troughs are consistent across multiple years. Even elevated seasons (2025 summer) plateaued at ~52/month. April 2026 already showed rapid decline (78/week → much lower), and UK activity was halving monthly (source 10). The summer migratory pause typically suppresses transmission sharply.

Key uncertainties

• The EFSA HPAI dashboard has not yet published any data for July 2026 (resolution window has not opened as of the evidence fetch date of 2026-06-06), so no direct authoritative count exists yet.
• Whether summer 2026 follows the low-activity pattern of summer 2024 (~15–20/month) or the elevated pattern of summer 2025 (~52/month) is unknown; the 2025–2026 season has been unusually intense.
• The rate and completeness of reporting lag — late-reported July outbreaks that appear only between August 1 and August 12 (the resolution deadline) — could add materially to the count.
• Whether persistent H5N1 strains circulating in seabird colonies (which drove elevated summer 2025 activity) continue to circulate into summer 2026 at similar intensity.
• How rapidly the steep spring 2026 decline trajectory continues into June–July 2026; April data showed weekly counts declining rapidly but starting from a very high base.

Forecast variants (3/3)

Variant 0: inside_view (inside_view_v1)

Inside View Causal Analysis

Causal chain: The 2025–2026 HPAI season in Europe peaked at >300 reports/week in late February 2026, then declined sharply: ~184 cases/week in late March, ~78 cases/week in mid-April (source 11, 13). UK detections were halving monthly from peak (source 10). This steep decline trajectory, if continued, implies that by July 2026 the seasonal trough should be reached.

Key mechanism: HPAI wild bird activity in European summer is suppressed by seasonal factors (migratory patterns, reduced waterfowl congregation, warmer temperatures). This is a robust, repeated historical pattern:

• Summer 2024: only 59 total wild bird detections across all Europe over ~14 weeks = ~15–20/month (source 5)
• Summer 2025: elevated at ~52/month due to persistent H5N1 in seabird colonies (source 7)

Evidence strength weighting:

• Scenario 2 (“Moderate summer activity ~40–70 outbreaks”) tagged relative_evidence_strength: high — this gets the most mass
• Scenario 1 (“Very low trough ~15–30 outbreaks”) tagged medium — second most mass
• Scenario 3 (“Higher residual ~70–130 outbreaks”) tagged low — minimal mass

Specific projection: The spring 2026 decline was from an exceptionally high base (~703/month in winter). April weekly rates of 78 imply ~312/month in April, declining further. The weekly halving trajectory from UK data suggests by July the rate is very low. However, the 2025–2026 season has been more intense overall than 2024, which elevated the summer 2025 floor relative to 2024. The best analogue is summer 2025 (~52/month) rather than summer 2024 (~15–20/month), given the season intensity.

Central expectation: ~40–65 outbreaks in July 2026. This falls squarely in 45–70 (bin_1), with meaningful probability also in 20–45 (bin_0) if the decline is steeper than 2025’s pattern, and modest probability in 70–95 (bin_2) if seabird colony activity persists.

Distribution rationale:

• 20–45 (bin_0): ~30% — summer 2024 analog + steep spring 2026 decline continuing (medium-strength scenario)
• 45–70 (bin_1): ~40% — summer 2025 analog, high-strength scenario, most supported
• 70–95 (bin_2): ~18% — low-strength residual activity scenario, upper tail
• 95–120 (bin_3): ~7% — very low probability, only if unusual persistence
• bin_4+ (>120): ~5% combined — essentially requires anomalous summer outbreak, against strong seasonal suppression

Variant 1: outside_view (outside_view_v1)

Reference class: monthly counts of HPAI wild-bird outbreaks in Europe for summer months (July in recent years). Base-rate evidence: summer 2024 was very low (~59 wild-bird detections across June–Sept 2024 → ~15–20/month; source 5) while summer 2025 was elevated (~156 across Jun–Sep 2025 → ~52/month; source 7). Adjustment: I shift most mass toward the 45–70 bin because the 2025–2026 season has been exceptionally intense (cumulative ~5,905–7,067 wild-bird cases by mid‑May 2026; sources 9 and 12) and winter 2025–26 peaks were very large (≈703 wild‑bird outbreaks/month Nov–Feb; source 8), which raises the plausible summer floor above the 2024 nadir. Countervailing downward pressure comes from the steep spring 2026 decline (weekly counts fell from >300 in Feb to ~78 in mid‑April; sources 13 and 11) and UK detections halving monthly (source 10), so a strong concentration in the moderate summer range is warranted but with a nontrivial tail to lower and higher bins. Final allocation (outside‑view weighting): 65% to 45–70 (bin_1) — reflecting the brief’s high‑evidence “Moderate summer activity” scenario (labelled high; sources 7, 6, 8, 9, 12); 20% to 20–45 (bin_0) — reflecting the medium‑evidence “Very low summer trough” scenario (sources 5, 3, 10–13); 10% to 70–95 (bin_2) and small tails (4% bin_3, 1% bin_4) to capture lower‑probability persistence or late reporting that would push counts higher (the brief’s “Higher residual activity” scenario is evidence‑light and thus given much less mass). Reading: strict — counts must be the discrete integer(s) visible on the EFSA HPAI dashboard when filtered to Wild birds / All / Observation date 2026-07-01–2026-07-31; UK (excl. NI) is included via WOAH integration per the dashboard (source 15). Confidence in evidence: medium (the dashboard had no July data yet as of the evidence snapshot and reporting lags and late additions to the dashboard before the resolution cutoff are important uncertainties).

Variant 2: contrarian_v2 (contrarian_v2)

Adversarial reading

The implicit consensus in the brief is that July 2026 will see moderate summer activity comparable to summer 2025 (~52/month, roughly 40–70 outbreaks) because the 2025–2026 season has been exceptionally intense overall and 2025 demonstrated elevated summer levels driven by persistent H5N1 strains in seabirds.

Counter-evidence found (or absence)

The brief explicitly notes that summer troughs are historically consistent across years, that April 2026 already showed a steep decline trajectory (78 cases in one week mid-April dropping rapidly), that UK detections were halving monthly from the winter peak, and that the summer migratory pause typically suppresses transmission sharply. The 2025 summer elevation may therefore be an outlier rather than the new baseline, and the very high winter 2025–2026 peak could have exhausted susceptible populations, favoring reversion toward the lower 2024 summer pattern (~15–20/month).

Distribution implications

This shifts material probability mass leftward relative to naive consensus: bin_0 receives the largest share because the decline dynamics and historical summer consistency outweigh the single elevated 2025 summer observation (high-evidence moderate scenario is discounted). Bin_1 is still substantial but reduced; bins 2+ receive only residual mass given the rapid spring decline already visible and the consistent summer nadir pattern.

Consolidate

Strategy: logit_mean • Drafts: 3/3

Reviewers (3/3)

anchoring → shift_lower (low magnitude, medium confidence) anchoring_v1

The draft rationale anchors heavily on the “Moderate summer activity ~40–70 outbreaks” scenario (tagged high-evidence) by treating the summer 2025 figure of ~52/month as the modal expectation. However, summer 2025 was itself identified as a notable outlier above summer 2024 (~15–20/month), so using the single elevated 2025 observation as the primary anchor represents optimistic selection of the upper historical analog rather than the median of the two-year range. The brief explicitly states “Summer 2024 was notably lower, suggesting 2025 was an outlier” — yet the draft places 47% of mass in 45–70 (bin_1), which corresponds to treating 2025 as the expected baseline. Furthermore, the rapid spring 2026 decline trajectory (78 cases/week in mid-April, UK halving monthly) implies a July entry rate far below even the summer 2025 floor, suggesting the modal expectation for July 2026 is closer to the lower 20–45 range than 45–70.

Flagged concerns

• Optimistic analog selection: The draft selects summer 2025 (~52/month) as the primary modal anchor despite the brief explicitly labeling it an outlier above summer 2024 (~15–20/month). The median of the two available summer analogues is ~35/month, placing the modal expectation in 20–45 (bin_0), not 45–70 (bin_1).
• Decline trajectory underweighted: April 2026 already showed ~78 cases in one week, but that was still declining rapidly; UK detections were halving monthly from the winter peak. Extrapolating the observed halving rate from April through July 2026 (3 additional halving periods) implies July totals well below 50, consistent with bin_0 rather than bin_1 as the mode.
• Best-case scenario treated as modal: The draft assigns 47% to bin_1 and 30% to bin_0, concentrating modal mass on the upper analog. A flat-prior average of the two anchors (2024: ~20/month, 2025: ~52/month) gives ~36/month as prior expectation, which falls squarely in 20–45 (bin_0). The current distribution is shifted ~one bin too high relative to a symmetric prior over the two historical analogs.

ceiling → shift_lower (low magnitude, medium confidence) ceiling_v1

The draft places ~67% of mass in 45–70 (bin_1) and 20–45 (bin_0), which is largely consistent with the structural evidence. However, there is a relevant ceiling constraint the forecaster underweights: the brief’s own best historical analog for July specifically (not the full June–September average) is the UK Defra July 2025 figure of 53 wild bird cases in Great Britain alone plus ~112 across all Europe over ~10 weeks ending July 21 2025 — implying a European July 2025 total of roughly 40–60. That is already a ceiling signal for a “normal elevated summer,” and the steep spring 2026 decline trajectory (78 cases in one week mid-April, halving monthly in UK per source 10) structurally compresses the upper tail. The week of April 16–22, 2026 had 78 cases — if that weekly rate continued halving through May, June, and July (~4 halvings over 8–10 weeks), the July weekly rate would be ~5–8 cases/week or ~20–35 for the month. This arithmetic strongly reinforces the floor-and-ceiling range of 20–70, making 70–95 (bin_2) and above over-weighted at ~20%+ combined. The high-end bins (bin_3 through bin_8) represent structural impossibility under the halving trajectory unless a wholly new outbreak wave emerges.

Flagged concerns

• Trajectory ceiling — halving-rate arithmetic: Source 10 states UK detections were halving each month from Nov/Dec 2025 peak. April 2026 showed ~78 wild bird cases in one week (source 11). Projecting 4 halvings from mid-April through July gives ~5–8 cases/week or ~20–35 for July — this provides a structural ceiling near 35–50 for the ‘continuation of spring trajectory’ scenario. The draft assigns ~20% to bins 70+ which is inconsistent with this arithmetic unless the halving trend reverses substantially.
• Upper-bin over-weighting relative to analog ceiling: The brief’s best analog for a full July European count is ~40–60 cases (summer 2025 rate of ~52/month across Europe, with July 2025 Europe-wide count implying ~40–60 from the UK Defra 112 across ~10 weeks figure). This places a soft ceiling around 60–70 for the ‘elevated summer’ scenario, yet the draft places ~20% mass in bins above this (bin_2 through bin_8 combined = ~22.5%). This is modestly inflated given the two-scenario structural ceiling.

math → shift_higher (low magnitude, medium confidence) math_v1

The forecast distribution sums to 1.0000 and bins are correctly keyed. However, there is a significant internal inconsistency: the rationale explicitly states the forecast should concentrate ~40% in 45–70 (bin_1), ~30% in 20–45 (bin_0), and ~18% in 70–95 (bin_2), but the published draft assigns 46.7% to bin_1, 30.4% to bin_0, and 14.2% to bin_2. The rationale does not justify the 32 percentage-point shift from the stated “Moderate summer activity ~40–70 outbreaks” high-evidence scenario (which should anchor the distribution) to the observed 46.7%, nor does it explain the reduction of bin_2 from 18% to 14.2%. The outside-view model (65% to bin_1) appears to drive the forecast away from the inside-view and brief-stated allocation, but this dominance is not made explicit or reconciled in the rationale text. The distribution shape is plausible in isolation, but does not match the stated reasoning.

Flagged concerns

• Allocation inconsistency: Rationale states inside-view should assign ~40% to 45–70 (bin_1) and ~30% to 20–45 (bin_0), but draft assigns 46.7% and 30.4% respectively. The 6.7 percentage-point shift to bin_1 is not explained.
• Missing reconciliation: Rationale mentions three independent models (inside-view, outside-view, contrarian) but does not explain how their outputs were combined. Outside-view allocates 65% to bin_1 vs inside-view 40%, yet the draft shows 46.7%—implying a weighted average or logit-mean, but the exact weighting and methodology are not stated in the rationale.
• High-evidence scenario underweighting: The brief tags Scenario 2 (‘Moderate summer activity ~40–70 outbreaks’) as relative_evidence_strength: high and cites four high-quality sources. The rationale correctly identifies this as the anchor, but the final distribution (46.7% bin_1, 30.4% bin_0) does not clearly reflect high-confidence concentration on the 45–70 range—the 30.4% to 20–45 (bin_0) is substantial and dilutes the signal from the high-evidence scenario.

Reconcile: partial_shift

Reviews consumed: 3

The anchoring reviewer flagged that summer 2025 (~52/month) was treated as modal when summer 2024 (~15-20/month) is an equally valid analog, and that the steep April 2026 halving trajectory implies July rates well below the bin_1 floor; I partially accepted this by shifting a small amount of mass from bin_1 to bin_0. The ceiling reviewer made consistent halving-arithmetic arguments (4 halvings from April through July → ~20-35 cases) supporting a lower mode but also noted bin_1 remains plausible as the elevated-summer analog; I integrated by keeping bins_0 and bin_1 close in mass with bin_1 still slightly modal. The math reviewer suggested shifting slightly higher (more weight on bin_1, bin_2) to better honor the high-evidence “moderate summer” scenario; I partially honored this by keeping bin_2 at ~15%, but I weighed the two shift_lower reviewers more heavily since they agree on direction and the spring decline trajectory is the strongest forward-looking signal. Net effect: a modest leftward shift from the draft’s 30.4/46.7/14.2 to 36/42/15, splitting the difference between anchoring/ceiling reviewers and the math reviewer’s hold-higher position.

Rejected critiques

• [math] Suggested shifting higher to give more mass to bin_1/bin_2 to honor the high-evidence ‘moderate summer’ scenario.

  • Why rejected: Two other reviewers (anchoring and ceiling) independently flagged shift_lower based on the same decline trajectory evidence. The halving arithmetic from April 2026 (~78/week) projected to July implies the spring decline outweighs the 2025 summer-floor analog. I kept bin_1 modal but did not shift mass higher into bin_2/bin_3.

• [math] Allocation inconsistency between stated inside-view rationale (40/30/18) and consolidated output (46.7/30.4/14.2).

  • Why rejected: This is an artifact of logit-mean consolidation across three variants, not a bug. The consolidated draft is a valid aggregation; recovering the exact inside-view percentages would override the outside-view and contrarian signals, which is not warranted.

Final Distribution

Evidence Sources (18)

1. Auto-prepended resolver source — Auto-prepended source-query: https://hpai.efsa.aus.vet/ (fetched 2026-06-06T08:04:26.356887Z). value=nil. Excerpt: EFSA HPAI dashboard Highly pathogenic avian influenza virus detection in Europe

Highly pathogenic avian influenza virus detection in Europe

Map - Subtypes

Time series

SOURCE: ADIS updated on 08/05/2026 and WAHIS updated on 11/05/2026

*United Kingdom (except Northern Ireland) data are sourced from ADNS up to 31/12/2020 and then from WOAH. United Kingdom (Northern Ireland) data are from ADIS.

Timeseries type: Bird categories Subtypes

2. EFSA HPAI Dashboard — The EFSA HPAI dashboard (authoritative resolution source) returned HTML with no discrete integer value for any specific filtered query; the dashboard is interactive and requires filter application to display counts. ADIS data updated 08/05/2026 and WAHIS updated 11/05/2026.
3. WOAH Facebook post - July 2025 avian influenza statistics — According to WOAH (World Organisation for Animal Health), July 2025 saw a relatively low number of avian influenza outbreaks globally: 10 outbreaks in poultry and 38 outbreaks in non-poultry birds and mammals, with 12,410 birds dead/culled.
4. Avian Influenza annual report 2024 - EFSA Journal — In 2024 (full year), 23 EU countries and 8 non-EU countries reported a total of 910 HPAI wild bird outbreaks through ADIS, with the highest HPAIV activity occurring in the first quarter. This averages approximately 75.8 outbreaks/month over the full year.
5. Avian influenza overview June–September 2024 - PMC — Between 15 June and 20 September 2024 (approximately 3 months), only 75 HPAI A(H5) and A(H7) virus detections were reported in Europe: 16 in domestic birds and 59 in wild birds. This confirms the summer period (June-September) is a low season for HPAI in wild birds in Europe.
6. 21 July 2025: HPAI in Great Britain and Europe - Defra — UK Defra report dated 21 July 2025: In Great Britain, wild bird HPAI cases (by test date) declined from 109 in April 2025 to 60 in May, 60 in June and 53 in July 2025 so far. Across Europe, 112 wild bird cases reported between 12 May and 21 July 2025 (about 10 weeks). The Netherlands was most affected with 23 cases, all other European countries had 8 or fewer cases in this period.
7. Avian influenza overview June–September 2025 - EFSA Journal — June–September 2025 (13 weeks): 156 HPAI wild bird detections across 15 European countries, approximately 52 per month. Predominantly seabirds (herring gulls), contrasting with summer 2024 which had only 59 wild bird detections in the same period—2025 summer was notably higher than 2024.
8. EFSA: Avian influenza detections in birds decline across the EU — Between 29 November 2025 and 27 February 2026, 406 HPAI outbreaks in domestic birds and 2,108 in wild birds were reported across 32 countries in Europe. This is approximately 703 wild bird outbreaks per month during winter 2025-2026.
9. Avian influenza, Europe - BEACON — As of 10 May 2026, Europe had reported 843 poultry outbreaks, 226 outbreaks in captive birds, and 7,067 cases in wild birds during the 2025-2026 season (starting approximately October 2025).
10. HPAI in Great Britain and Europe - Defra 10 April 2026 — UK Defra report (10 April 2026): wild bird HPAI cases in Great Britain have been halving each month since the peak in November/December 2025. Since peak, detections declined sharply into spring 2026. Since 10 March 2026, only 15 new detections based on collection date in Great Britain. Total wild bird cases in GB since 1 October 2025 season: 908.
11. ADIS weekly update: avian disease outbreaks in Europe (16-22 April 2026) — ADIS weekly update (16-22 April 2026): In wild birds, HPAI H5N1 outbreaks were reported in Austria (9), Germany (39), Denmark (9), Poland (7), Sweden (7), Finland (2), France (3), Norway (2) = approximately 78 wild bird outbreaks in just one week in mid-April 2026, showing activity was still significant in spring 2026.
12. FACE Animal Health Updates for May 2026 — FACE Animal Health update (last updated 19 May 2026): Total avian influenza cases in Europe = 6,721 total outbreaks; wild bird cases = 5,905; farms/captive/backyard = 962; affected countries = 35. This is the cumulative 2025-2026 season total as of mid-May 2026.
13. BEACON: HPAI Europe report week 23-29 March 2026 — During the week of 23-29 March 2026, 184 new wild bird HPAI cases were reported across Europe (~26 countries). Levels had risen to over 300 positive reports per week through late February 2026 before decreasing to under 150 reports per week in early April 2026. This confirms spring 2026 decline trajectory.
14. EFSA HPAI dashboard (auto-prepended fetch) — EFSA HPAI dashboard HTML requires interactive filter application to display counts; the pre-fetched dashboard page (fetched 2026-06-06) did not expose a discrete July-2026 wild-bird outbreak count in the static HTML.
15. EFSA HPAI Dashboard — The EFSA HPAI dashboard (hpai.efsa.aus.vet) explicitly states: “United Kingdom (except Northern Ireland) data are sourced from ADNS up to 31/12/2020 and then from WOAH.” This means WOAH-WAHIS data for UK (excl. NI) is directly integrated into and displayed on the EFSA HPAI dashboard automatically — no separate manual WOAH consultation is required.
16. Avian influenza overview September–November 2025 (EFSA Scientific Report, doi: 10.2903/j.efsa.2025.9834) — The EFSA Avian Influenza overview report (September–November 2025) confirms: “United Kingdom (excluding Northern Ireland) data are from WOAH-WAHIS, whereas United Kingdom (Northern Ireland) data are from ADIS.” The source note also states: “Source: ADIS, EFSA and WOAH-WAHIS (data extraction carried out on 28 November 2025).” This confirms EFSA actively extracts and integrates WOAH-WAHIS data for UK (excl. NI) into its own reporting pipeline.
17. Avian influenza overview September–November 2025 – Appendix B Data and Methodologies (EFSA Scientific Report) — The EFSA report methodology section (Appendix B) describes two parallel data streams: “B.1.1 Overview of avian influenza outbreaks in Europe” (covering EU/EEA countries via ADIS) and “B.1.2 Overview of avian influenza outbreaks in other countries not reporting via ADIS” — the latter explicitly covers countries like UK (excl. NI) that report via WOAH-WAHIS, confirming WOAH data is incorporated at the same data-extraction step, not post-hoc or manually.
18. Avian influenza overview March–June 2025 (EFSA Journal) — The EFSA March–June 2025 Avian Influenza overview confirms: “From 1 January 2021 onwards, the data source was WOAH for the United Kingdom (excluding Northern Ireland) and ADNS/ADIS for the United Kingdom (Northern Ireland).” This is consistent across multiple EFSA quarterly reports, indicating a stable, systematic integration of WOAH data for UK (excl. NI).

wooley_rhino • run 2bbc82fb • $2.94 • 10m14s

Seasonality limits likely pathways: waterfowl-driven winter waves and colony-driven summer waves follow different dynamics.

Geographic reporting boundaries and surveillance rhythms compress observable outbreaks and introduce persistent delay effects.

Recent transmission among migratory waterfowl has declined sharply from winter peaks.

Signals from colony-breeding seabirds and gulls remain weak so far, lowering the immediate amplification pressure.

If seabird-colony amplification begins in late June, reported July counts can increase sharply and concentrate regionally.

Absent seabird amplification, July activity is more likely to remain at low-to-moderate seasonal levels with scattered local events.

Key uncertainties include reporting lags and backfilled observation dates that can move the final July tally upward after the month ends.

Sensitivity to species composition, regional surveillance intensity, and the biological potential for colony outbreaks preserves a nontrivial upper tail in outcomes.

Forecast (median July 2026): 63.28 outbreaks

• Summer is usually low: EFSA reporting shows a consistent seasonal “trough” in wild-bird HPAI during summer, so July is expected to be far below autumn/winter levels.
• July can be a local peak: In the June–September 2025 period, detections were specifically noted to peak in July, even though the overall season is lower than autumn.
• Last season matters: The run-up into 2026 followed an unusually intense autumn–winter 2025–2026, which raises the chance that July 2026 stays elevated rather than reverting immediately.
• Best historical anchors are limited: The reasoning relies heavily on the contrast between summer quarters in 2024 vs 2025 (with July singled out in 2025), but there’s no direct, clean “July-only” history available.
• Big unknown = spring-to-summer trajectory: How active March–June 2026 ends up being (not fully observed/available in the analysis) is the main factor that could push July well above or below the median.