Weekly Regional Risk Digest, Issue 017

This issue covers a fourteen-day interval rather than the usual seven; no issue was published on 14 August, and all movement is measured against 7 August. Three regions carry developments that reset how I would advise on them: the Middle East, where the Oman transit framework collapsed and the composite reached the highest figure this index has ever recorded; Sub-Saharan Africa, where the Congo outbreak became the deadliest in that country’s history; and North Africa, where the Bab el-Mandeb began killing merchant crews.

Two scores moved downward this issue and both deserve a word, because a falling number is easy to misread. East Asia fell three points because Typhoon Dolphin arrived, did what we forecast, and passed — so we removed the elevation we had added for it. Central Asia fell five points because we corrected our own scoring, not because anything in Afghanistan improved. I would rather tell you that plainly than let the arrow imply something it does not mean.

You will also find a correction to our 24 July issue in the Western Europe section. We got something wrong about the Spanish wildfire deaths and we have fixed it in the open.

The single most useful thing to understand about this region right now is that the risk has migrated almost entirely into movement. Being in Dubai, Doha or Riyadh is not materially more dangerous this week than last. Getting cargo, fuel or yourself in and out of the Gulf is dramatically harder, more expensive and less predictable than at any point since February. The insurance market has priced this more honestly than the political statements have.

Aviation is the practical chokepoint for most travelers. Gulf carriers continue to operate, but the routing environment is unstable, conflict-zone advisories are live, and a single further escalation against a Gulf state produces the kind of airspace closure that stranded travelers in Kuwait and Erbil in July. If you are transiting Doha, Dubai or Abu Dhabi, the exposure is not the destination but the connection.

Lebanon has moved from a fragile ceasefire to open escalation over the past week. Israel resumed strikes after two reservists were killed on 5 August, issued its first forced displacement order in weeks, and killed eleven on 15 and 16 August with the tempo increasing. Southern Lebanon and the Tyre district should be treated as an active combat zone, and Beirut carries meaningful onward-escalation risk.

One genuine caution against over-reading the score: this remains a High-rated region, not a Critical one, and the difference is real. Commercial aviation functions, Gulf state infrastructure is intact, and no Gulf energy facility has been struck. Business travel to the GCC is being conducted successfully every day. What has changed is that the margin for error has narrowed considerably and the contingency planning that felt excessive in June is now proportionate.

▪ A strike on a Gulf energy node. A hit on Saudi, Emirati or Kuwaiti production or export infrastructure is the clearest single trigger for a Critical crossing and would take Brent decisively through $100.

▪ Multi-state Iranian salvos in a single action. The 18 August UAE strike involved one state and two missiles. Simultaneous strikes across several Gulf states would restore the pattern of the war's opening weeks and force region-wide airspace closures.

▪ Restoration of any channel, direct or via Oman. Trump threatened Oman on 17 August. Whether the mediator is restored or punished is the cleanest early signal of which direction the next month runs.

▪ War-risk insurance falling below roughly twenty times pre-crisis. Premiums are a better forward indicator than official statements, because operators are betting their own capital on the answer.

The practical picture for most travelers to this region has changed less than the headline numbers imply, and saying so is not complacency. The CDC continues to assess US-traveler risk from the outbreak as low, no case has been reported outside Africa, and advisory levels have not moved since 4 August. Nairobi, Accra, Cape Town, Kigali and Dakar are functioning normally. The catastrophe is real and it is geographically concentrated.

Where that concentration matters, it matters absolutely. Ituri and Nord-Kivu carry avoid-all-travel guidance; Haut-Uele and Tshopo carry reconsider-nonessential-travel guidance. The outbreak has now penetrated internally displaced persons camps, and it is spreading in areas under contested armed control where contact tracing is hardest. There is no licensed treatment for Bundibugyo virus. If your itinerary touches eastern DRC, the correct answer is not mitigation, it is postponement.

Somalia and the Kenyan border counties deserve more attention than they are receiving. The African Union drawdown is happening on a funding timetable rather than a security one, and al Shabaab's tempo in Mogadishu is the highest in months. Mandera county in Kenya saw attacks surge across June and July. Overland travel in Kenya's north-eastern counties is not advisable.

The Ervebo allocation deserves to be understood precisely. Seventy thousand doses is a meaningful intervention and the first of this outbreak, but 20,000 of those are allocated to a Phase 3 trial specifically because it is not known whether the vaccine protects against this virus in humans. Early laboratory and animal data suggest it may. That is a reason for cautious hope, not for assuming the curve is about to bend.

▪ The case fatality ratio turning downward. It has risen for several consecutive weeks to 47.4 percent. A sustained fall would be the first hard evidence that the response is closing the gap.

▪ A confirmed case outside Africa. An exported case would trigger advisory and screening changes across multiple governments within days, largely independent of the actual transmission risk it represents.

▪ Confirmed transmission in Kinshasa or Goma. Penetration of a major population and transit hub converts this from a regional emergency into a continental one and would change aviation policy quickly.

▪ Al Shabaab holding a district capital through the AU drawdown. The group has taken and lost positions repeatedly. Holding one would signal that the handoff to Somali forces is failing rather than merely strained.

The monsoon is the dominant operational factor in this region for the next six weeks, and it is a planning problem rather than a cancellation trigger. Delhi, Mumbai, Colombo and Male are functioning. Northern Pakistan, Himachal Pradesh, Uttarakhand and Assam are where the disruption concentrates, and it concentrates on roads, mountain passes and short-haul aviation.

The specific hazard travelers underestimate here is the cloudburst-and-landslide sequence in mountain districts. It arrives with little warning, cuts road access rather than damaging destinations, and strands people rather than injuring them. If your itinerary includes hill stations, mountain trekking or overland movement through northern Pakistan or the Indian Himalaya between now and late September, build in slack and confirm road status the morning of travel rather than the week before.

The Afghanistan-Pakistan frontier remains closed to ordinary travel and should be treated as such. Pakistani strikes into Afghan provinces continue, and UN reporting documents 1,187 people killed or injured in those strikes between December 2024 and June 2026. Khyber Pakhtunkhwa and Balochistan carry elevated risk independent of the monsoon.

For business travel to Indian metropolitan centres, this region's score overstates what you will actually encounter. The composite is pulled upward by Pakistan, Afghanistan-adjacent territory and rural flood exposure. Bangalore, Hyderabad and Chennai present a materially lower risk profile than a 58 implies.

▪ The September monsoon peak against an above-normal forecast. Pakistan's NDMA identifies August and September as the climatological peak. A repeat of the 2022 or 2025 flood scale would move this region several points.

▪ Any terror incident inside India attributed across the border. The post-Sindoor ceasefire holds, but Indian doctrine now rejects geographic limits on retaliation. A single mass-casualty incident could escalate this region faster than any other on the board.

▪ Escalation of Pakistani strikes into Afghanistan. The current tempo is receded but continuous. A return to the intensity of early 2026 would raise both South and Central Asia.

▪ Renewed cost-of-living protest across multiple countries. Energy prices have turned upward again. The earlier protest wave spanned six countries and is a proven response to this specific pressure.

Egypt, Morocco and Tunisia remain functional destinations and the score should not be read as a warning against them. Cairo, Marrakech and Tunis are operating normally, tourism infrastructure is intact, and the direct security threat to visitors in those countries has not changed. What has changed sits offshore and upstream, and it reaches travelers through price and availability rather than through danger.

That said, the economic squeeze is the thing to watch in this region, and it is worth being direct about why. Two impaired chokepoints feed straight into domestic fuel and bread prices in countries where subsidy adjustments have historically produced street protest within days. Nothing has happened yet. If it does, it will happen quickly and it will centre on the capitals.

Sudan is not a destination and should not be treated as one. El-Obeid is under active siege with roughly half a million civilians trapped, and both UN and US officials have warned that an RSF ground assault could produce mass atrocities on the scale of el-Fasher. Libya's security situation remains fragmented and its institutional disputes are flaring.

For anyone with shipping, logistics or supply-chain exposure through the Red Sea, the 11 August attack is the operative development of the fortnight. The declared blockade nominally targets Saudi, Israeli and US-linked shipping, but the Tihamah was Egyptian-owned and its crew were third-country nationals. Assume the targeting discrimination is looser in practice than in declaration.

▪ A lethal attack on a genuinely neutral hull. The Tihamah was Egyptian-owned with Pakistani and Indonesian crew. A clearly unaffiliated vessel taking casualties would collapse Bab el-Mandeb transit outright.

▪ Suez transit volumes breaking from their plateau. They have held broadly unchanged through the blockade. A sharp fall is the signal that operators have written the corridor off rather than merely re-priced it.

▪ The fall of el-Obeid. An RSF ground assault on a besieged city of half a million carries the el-Fasher precedent, and both UN and US officials have warned in exactly those terms.

▪ Fuel or bread subsidy unrest in Egypt or Tunisia. The economic transmission from two impaired chokepoints is direct. Street response is the lagging indicator, and it lags by weeks rather than months.

Colombia is the region's live story and the picture is more specific than a country-level warning would suggest. The violence is concentrated in Cauca, Tolima, Narino, Catatumbo and the Pacific coast, and it is aimed at state targets rather than at foreigners. Bogota, Medellin and Cartagena continue to operate normally. What has changed is the trajectory: a government that has publicly ended negotiations and an armed opposition that has answered with drones and explosives will produce more of this, not less, over the coming months.

The practical consequence for travelers is about overland movement rather than presence. The Pan-American Highway attack on 8 August hit exactly the kind of infrastructure that matters for road travel between Colombian cities. Fly between destinations rather than driving, particularly in the south-west.

Venezuela remains a country in an unsettled transition compounded by disaster recovery. The earthquake toll passed 6,000, the opposition is pressing to remove an unpopular interim government, and the political prisoner question is unresolved. Maiquetia is operating, but this is not a destination for discretionary travel.

The rest of the continent is materially calmer than the regional score implies, and that gap is worth naming. Chile, Uruguay, Argentina and southern Brazil present risk profiles closer to the Moderate band. A traveler to Santiago or Montevideo is not facing a 55.

▪ Escalation of the Colombian government's confrontation campaign. A major military operation producing significant casualties would likely trigger a corresponding armed-group response in urban centres rather than only rural ones.

▪ Attacks moving from state targets to civilian or commercial ones. So far the pattern is toll booths, town halls and police stations. A shift toward soft targets would change the traveler calculus entirely.

▪ A resolution or escalation of Petro's annulment suit. An unresolved legal challenge to the presidency is a standing source of institutional instability and street mobilisation.

▪ Movement on Venezuelan elections. The opposition's push for elections is the most plausible route to either stabilisation or a fresh crisis, and it will be visible well before it resolves.

The single most important development for travelers this fortnight is the smallest one. A NATO member's airspace was violated and an allied fighter shot down the intruder. Four such incursions have now occurred in Romania this year. This is being managed carefully by all parties, but it places Romania, Poland and the Baltic states inside the operational picture of a war they are not party to, and travelers to those countries should understand that context even though it does not currently change day-to-day conditions.

Ukraine is not open to ordinary travel and the mass-casualty strikes on Kyiv on 20 August underline why. Air defence interceptor supply has fallen sharply, and the practical effect of that shortfall is measured in civilian deaths in the capital rather than in territory. Anyone with essential business in Ukraine should be planning around shelter access rather than around schedules.

Russia presents a distinct and separate problem. Beyond the airspace and airport disruption from mass drone assaults on Moscow, the wrongful-detention risk for US and allied citizens remains formally designated, and consular support is minimal. Domestic fuel shortages are affecting overland movement in some regions.

For the rest of the region, and this is the majority of it by population and by visitor volume, conditions are normal. Prague, Krakow, Budapest, Ljubljana, Tbilisi and the Baltic capitals are functioning without disruption. The composite reflects a war at one end of a very large region and should not be read as a caution against the other end.

▪ A NATO airspace incident causing casualties or damage on alliance territory. Four Romanian incursions have been contained. One that lands changes the political situation within hours rather than weeks.

▪ Announced interceptor resupply to Ukraine. The three-fold year-on-year shortfall is the variable most directly connected to civilian casualties in Kyiv, and any resupply announcement is a leading indicator of that changing.

▪ Consolidation of Ukraine's Dnipropetrovsk gains. Roughly 745 square kilometres were recaptured, with ISW conservatively assessing 627. Holding that ground through September would mark a genuine shift rather than a raid.

▪ Sustained closure of Moscow-area airports. Mass drone assaults on the capital produce knock-on effects for anyone transiting Russian airspace, and repeated closures would signal a durable change in air defence performance.

This region's five-point fall is a correction to our scoring and not an improvement on the ground, and it would be misleading to let a falling number suggest otherwise. We had been scoring seven countries as though the worst of them defined all of them. Nothing about Afghanistan has improved. What has changed is that the score now distinguishes between Almaty and Kandahar, which it previously did not.

Afghanistan is why this region remains rated High. The conflict with Pakistan continues along the Durand Line, ISKP retains operational capacity and is assessed as the primary threat to the wider region, and the Taliban is now facing armed internal opposition in Badakhshan from the Sepahiyan-e Mihan Front, the Afghan Freedom Front and the National Resistance Front. This is a structural vulnerability for Taliban governance and a deteriorating environment for anyone inside the country.

For the Central Asian republics themselves, the picture is considerably more workable than the composite implies. Uzbekistan sits at the lowest US State Department advisory level. Kazakhstan, Kyrgyzstan and Mongolia are accessible destinations with functioning infrastructure and a growing tourism sector, and inter-state coordination among them has improved. The principal traveler risks there are road safety, seismic exposure and, in some countries, the arbitrary-detention profile common to authoritarian systems.

The one genuine regional concern that spans borders is ISKP's recruitment and operational reach. Russia's foreign minister has characterised it as the primary security threat to Central Asia, and the group's ambition is regional rather than confined to Afghanistan. That is a reason for situational awareness in the republics, not a reason to avoid them.

▪ Escalation along the Durand Line. The current tempo is receded but continuous. A return to early-2026 intensity would raise both this region and South Asia.

▪ Consolidation of the Badakhshan resistance. Three separate groups are now operating against the Taliban in one province. If any holds territory, it marks the first durable internal challenge to Taliban rule.

▪ An ISKP attack inside a Central Asian republic. The group's stated ambition is regional. An attack in Tashkent, Almaty or Bishkek would change the risk picture for the republics immediately.

▪ Regional seismic activity. Tajikistan, Kyrgyzstan and northern Afghanistan carry significant earthquake exposure with limited response capacity, and this is the most likely source of a sudden change.

This region's score held flat because two genuine forces cancelled each other out, and both are worth understanding separately rather than as a net figure. Security deteriorated in Haiti. Environmental risk fell materially across the whole basin. A traveler heading to Jamaica or the Dominican Republic in September is facing a better proposition this year than in most; a traveler with business in Port-au-Prince is facing a worse one.

The hurricane point is the practically useful one for most subscribers. September is normally the month to avoid for Caribbean travel, and this year the basin is exceptionally quiet with a strong El Nino suppressing development through what is usually the peak. That is not a guarantee, and one destructive storm remains entirely possible. But if you have been deferring Caribbean travel on seasonal grounds, this is an unusually favourable year to reconsider.

Haiti is a separate category and should be treated as one. Gang control of the capital is near-total, the elections are being organised in conditions that invite violence, and the international force intended to address this remains far under its authorised strength. There is no version of discretionary travel to Haiti that is currently defensible.

Mexico continues to require the same geographic discrimination it has needed for years. The resort corridors, Mexico City and the major business centres operate normally and safely for visitors who use sensible practices. Specific states carry cartel violence that is not aimed at tourists but that tourists can be caught inside. Check the state-level advisory rather than the country-level one.

▪ Haiti's election proceeding or collapsing. Either outcome is consequential. A collapsed process would likely accelerate gang consolidation; a contested result would produce open conflict.

▪ Any storm developing despite the suppression. The season is quiet, not closed. A single major hurricane making landfall would reverse the environmental easing immediately.

▪ Reinforcement of the UN gang suppression force in Haiti. The force is well under authorised strength. Meaningful reinforcement is the only plausible route to improvement.

▪ Cartel violence spilling into a tourism corridor in Mexico. The pattern has held for years that tourism zones are largely spared. A significant breach of that pattern would move this region several points.

The earthquake is the operative development and its impact is geographically narrow. East Nusa Tenggara is not a high-volume destination, and Bali, Java and the principal Indonesian tourism corridors were unaffected. The aftershock sequence, including a magnitude 6.2, is the reason this remains a live rather than a closed event, and anyone travelling to eastern Indonesia over the coming month should factor that in.

The wider regional picture this fortnight was weather, and it was widespread rather than severe. Six ASEAN states recorded flooding or storm damage in a single week, driven by the Southwest Monsoon compounded by two tropical cyclones. This is normal for the season, it disrupts domestic aviation and road travel rather than threatening safety, and the practical response is schedule flexibility rather than avoidance.

Eight volcanoes across Indonesia and the Philippines are at elevated alert levels simultaneously. That is not unusual for these archipelagos, but it is worth knowing if your itinerary includes volcano trekking or if you are transiting airports whose operations are ash-sensitive. Check alert status before committing to those specific activities.

Myanmar remains the region's genuine no-go and the situation there is getting worse rather than plateauing. Airstrikes are killing civilians in three states, the humanitarian picture is deteriorating sharply, and there is no diplomatic process to point to. Nothing about the country is currently suitable for discretionary travel.

▪ The East Nusa Tenggara aftershock sequence. 341 aftershocks including a magnitude 6.2 indicate an active sequence. Decay over the coming weeks would allow this elevation to reverse.

▪ An eruption from any of the eight volcanoes on elevated alert. Ash disruption to regional aviation is the practical consequence, and it can close airports across Indonesia and the Philippines at short notice.

▪ A direct coast guard confrontation in the South China Sea. The dispute is currently legal rather than physical. The carrier's departure from the theatre is the variable that could change that.

▪ Further escalation of Myanmar's air campaign. Civilian casualties across Shan, Rakhine and Chin States are rising with no diplomatic constraint in sight.

This region's three-point fall is exactly what it looks like: a storm we scored on the forecast arrived, did what was expected, and left. Dolphin made landfall on 9 August and dissipated by 11 August. Shanghai's aviation has normalised, evacuees have returned, and there is no reason to carry the elevation forward. We raise scores on forecast events when the timing warrants it, and we lower them when the event passes, to the same level rather than to somewhere in between.

Nothing else about this region eased, and it would be a mistake to read the falling score as a general improvement. Chinese maritime pressure at the Senkakus ran six consecutive days, PLA activity around Taiwan continued at its daily tempo with median-line crossings, and the US moved its only regional aircraft carrier to the Middle East, which regional analysts read as creating a perceived vacuum.

The Korean Peninsula carries the nearest-term specific risk. The Ulchi Freedom Shield exercises run through 27 August, North Korea has explicitly threatened response, and two ballistic missile tests preceded that threat. This pattern recurs annually and has never produced conflict, but it does produce missile tests and rhetoric that can disrupt regional aviation briefly and unsettle travelers who are not expecting it.

For the ordinary business or leisure traveler, Tokyo, Seoul, Taipei, Osaka and Hong Kong are operating entirely normally, and this region's composite of 34 is one of the lower figures on the board for good reason. The friction here is between states, not directed at visitors. The exception worth flagging is the standing wrongful-detention designation for US citizens travelling to mainland China, which warrants genuine consideration of your device and data practices.

▪ North Korean response to Ulchi Freedom Shield through 27 August. The exercises run to the end of the month and Pyongyang has threatened stern steps. Expect tests; watch for anything beyond them.

▪ Sustained Chinese presence at the Senkakus beyond a week. Six consecutive days is already an escalation of pattern. Continuation would represent a deliberate change in posture rather than a routine assertion.

▪ Consequences of the carrier redeployment. Regional analysts read the US carrier's move to the Middle East as a vacuum. Whether China tests it is the question worth watching.

▪ The next Pacific system. Typhoon season runs through October, and the Shanghai and Taipei corridors are the ones that matter for travel disruption.

Nothing in this region affects travel this month, and that is worth saying clearly before the rest. It is winter in Australia and New Zealand, conditions are normal, and both countries are among the most straightforward destinations on this board. The reason this region's score is not lower is entirely about what is coming.

The bushfire outlook released on 19 August is the most significant forward signal in this issue for anyone planning Southern Hemisphere summer travel. Elevated fire danger in every jurisdiction but one is unusual, and the underlying driver is an El Nino the Bureau of Meteorology warns may become the strongest on record. If you are booking Australian travel for December through February, this is the moment to build in flexibility rather than the week before departure.

The practical implications for summer travel are specific rather than general. Fire risk affects road access, national park closures and regional aviation more than it affects city safety. Melbourne, Sydney, Brisbane and Adelaide will function; the drives and the parks between them are where plans break. Smoke exposure in cities is a secondary effect that has affected air quality significantly in past El Nino years.

For the Pacific island states, the standing considerations are unchanged and are about distance rather than danger. Thin aviation, weather-dependent connections and long medical evacuation timelines are the operative risks. Cyclone season begins in November and this year's El Nino pattern will shape it, which is worth watching as bookings for that period are made.

▪ El Nino strengthening toward the record forecast. The Bureau of Meteorology has flagged the potential for the strongest event on record. Confirmation would raise this region's environmental exposure substantially for the coming summer.

▪ Early-season fires in Queensland or New South Wales. Fires arriving before the conventional season would confirm the outlook's warning and indicate a severe summer ahead.

▪ Drought declarations in eastern Australia. Soil-moisture deficits expanded through July. Formal drought declarations would confirm the trajectory and carry water-restriction consequences for regional travel.

▪ The El Nino signal for the November cyclone season. Pacific island cyclone risk is shaped by the same pattern, and early indications will emerge over the coming two months.

The most useful thing to tell a subscriber about Western Europe right now is that the last week of August is a better proposition than the first three were. The fires that dominated this region's summer are stabilising or out, temperatures are dropping roughly ten degrees, and rain is returning to some areas. Iberian travel in particular has improved materially in the past fortnight.

That improvement should not obscure how severe the season actually was. France recorded its worst wildfire year on record at roughly 115,000 hectares, more than three times its average. Spain has burned over 217,000 hectares. Germany reported a record 14,000 heat-related deaths this year. The season is not over, this is still peak fire period, and the underlying conditions that produced it have not changed.

We are also correcting our own record here. In our 24 July issue we reported that the France and Spain wildfires had produced no deaths in either country. That was wrong as to Spain. The Los Gallardos fire in Almeria on 10 July killed at least twelve people, caught while attempting to escape along a river bed, with the season total for Spanish civilian wildfire deaths now standing at fifteen. We would rather correct that publicly than leave it standing.

For everything other than fire and heat, this region continues to be one of the most straightforward on the board. Schengen entry processing remains slow at major hubs during peak season and is the main practical annoyance. The security baseline is unchanged. If you are travelling to Western Europe in September, the outlook is good.

▪ A late-August or September heat wave reigniting the fire season. The fires are stabilising, not out, and fuel conditions across Iberia remain exceptional. One more heat episode would reverse the easing.

▪ Completion or reversal of the fire season's close. Whether the current stabilisation holds through September determines whether this region continues easing toward the Low band.

▪ Entry-Exit System queue conditions into the autumn. Peak-season friction should decline after August. If it does not, it indicates a structural rather than seasonal processing problem.

▪ Any acute security event. The regional threat baseline has been stable for months, which is precisely why a single incident would move this score more than incremental drivers would.

We are closing a gap in our own coverage this issue and it is worth being explicit about it. This region's environmental score has been anchored almost entirely on the Atlantic hurricane season across every issue of this index. Meanwhile a wildfire season running at 146 percent of its ten-year average acreage, with smoke producing hazardous air quality across six states, was not being captured at all. That was too narrow a view of what environmental risk means in North America, and the score now reflects both.

The practical guidance that follows is specific. If you have a respiratory or cardiac condition and you are travelling in the Pacific Northwest, the Mountain West or central Canada between now and the autumn, check air quality on the morning of travel rather than the week before. Conditions shift daily with wind direction, and a destination that was clear yesterday can be in the hazardous category today. Fine particulate matter is the hazard, and indoor air is meaningfully better than outdoor air on bad days.

Hurricane Lala was a reminder that Hawaii sits in a different basin from the Atlantic and is not covered by the Atlantic season's quiet. It passed near the islands on 15 and 16 August with 70 mile per hour winds and up to three feet of rain on windward slopes, killed one person, cut power to more than 130,000 households and forced a search-and-rescue operation after flash floods displaced occupied homes on the Big Island.

The Atlantic basin itself continues to be the good news. The season is the slowest-starting since 1988 by accumulated cyclone energy with three named storms totalling four and a half days, and El Nino shear is expected to suppress the September peak. For Gulf and East Coast travel through the autumn, this is a materially better year than average.

▪ Wildfire season trajectory into September and October. The season is already at 146 percent of average acreage with El Nino forecast to extend drier conditions. A continued run would sustain smoke exposure well into autumn.

▪ A major smoke event reaching eastern population centres. Canadian smoke has reached the eastern seaboard repeatedly in recent years, and it produces air quality alerts across a much larger population than western fires do.

▪ Whether the Atlantic remains suppressed through the September peak. The basin has been remarkably quiet, but September is historically the peak and one landfalling storm would reverse the environmental easing.

▪ Midterm campaign security incidents. Political tension is rising through the cycle. Incidents at campaign events would be the observable indicator of that translating into risk.