The Purple Report September 2026 | Disaster Forecasting and Predictions

Global 10-Year Hotspot Outlook, 2026–2036

Dated baseline: 1 September 2026
Forecast coverage: 50 global hotspot systems
Planning horizon: September 2026 to September 2036
Current-month context: strong El Niño developing; global temperatures expected to remain at or near record levels in coming years

The Purple Report is moving one step upstream.

Daily Purple reporting asks what happened, what changed, what was repaired and what remained unresolved. This September edition asks a harder question: where are the ingredients for future catastrophe already assembling?

That does not mean we can predict the exact day, place and magnitude of every future disaster. We cannot. Earthquakes, volcanoes, landslides, floods, cyclones, wildfire and coastal failure have different physics, different observing systems and different forecast limits.

The purpose of this report is not to announce disasters before science can. It is to identify places where susceptibility, changing conditions, exposed people, critical infrastructure and cascading consequences are converging strongly enough to deserve attention.

Quick Read

This first September baseline contains 50 hotspot systems: 15 earthquake/tsunami systems, 10 volcano/lahar systems, 8 cryosphere/landslide systems, 10 flood/cyclone/coastal systems and 7 heat/fire/drought compound systems.

The September watch list contains 8 HIGH-CONCERN, 14 ELEVATED and 28 WATCH systems. These are Purple Report research states, not official emergency alerts. National and local authorities retain control of public warnings, evacuation orders and hazard-specific alert levels.

The 10-year horizon is a planning horizon, not a promise that a listed event will happen within ten years. For example, the official Cascadia and Alpine Fault probabilities are published on 50-year horizons. This report preserves those native horizons rather than inventing a ten-year number. Taiwan is different: its official earthquake potential system already publishes 10-, 30- and 50-year maps. Volcanoes are different again: current deformation and alert states can change over days or weeks.

The Forecasting Boundary: Prediction Is Not One Thing

The word prediction is attractive because it sounds precise. Natural-hazard science is usually more disciplined. A useful forecast system must first ask what kind of claim is scientifically available.

Public forecast form	What it means	What it must not become
Official long-term probability	An authority publishes a probability or hazard model over a defined horizon.	A casual conversion into an exact 10-year probability or date.
Current official alert or unrest	Authoritative monitoring reports an active change in a volcano or geophysical system.	An assertion that eruption or failure is certain.
Changing environmental conditions	Rain, heat, water, ice, fuel or other conditions are changing the background hazard.	A deterministic event prediction.
Post-event secondary risk	A recent event leaves unstable slopes, blocked rivers, damaged infrastructure or recovery vulnerabilities.	Treating the first event as finished because rescue has begun.
Seasonal or climate outlook	Climate conditions alter the background likelihood or severity of some hazards.	An exact storm landfall, wildfire ignition or flood date.

Three Clocks, Not One

A ten-year project needs more than a ten-year clock. The same place may be structurally dangerous for decades and temporarily quiet this month.

Near-term conditions. Rainfall, heat, fire weather, glacier motion, volcanic unrest, river levels and similar observations can change quickly.
Seasonal and developing conditions. Climate patterns, recovery weakness, persistent drought, lake growth and infrastructure deterioration can alter risk over time.
Long-term structural hazard. Active faults, volcanoes, low-lying deltas, glacierised valleys and subsiding coasts remain important planning concerns even during quiet periods.

September 2026 Global Loading

The September baseline does not begin from climatic neutrality. The World Meteorological Organization’s August–October 2026 outlook says a strong El Niño is developing and expected to strengthen, with above-normal temperatures expected widely and a positive Indian Ocean Dipole also expected. The same WMO system is explicit that seasonal forecasts are windows for early action, not deterministic local-event forecasts.

The WMO Global Annual to Decadal Climate Update 2026–2035 expects global temperatures to remain at or near record levels in the next five years and gives a 91% chance that at least one year from 2026–2030 temporarily exceeds 1.5°C above the 1850–1900 average. That does not tell us where the next flood, fire or cyclone will occur. It changes the background operating environment for heat, rainfall extremes, ocean heat, glacier loss and sea-level pressure.

UNDRR adds the wider impact context. Direct disaster losses tell only part of the story: its Global Assessment Report 2025 estimates annual disaster costs above US$2.3 trillion once cascading and ecosystem impacts are included. That is why this report asks not only where can the hazard occur?, but what else breaks if this critical point fails?

How to Read the Public Outlook

Each hotspot is reviewed against public scientific evidence, current observations, exposed people and infrastructure, preparedness, and uncertainty. The article gives readers enough reasoning to understand the evidence, its significance and its limits.

Public reading	What readers should ask
Hazard evidence	What do authoritative scientific sources say about the physical hazard?
Changing conditions	Are current observations increasing or reducing concern?
People and infrastructure	Who or what could be affected if the hazard develops?
Uncertainty and preparedness	What is not known, and what warning or response capacity exists?

High consequence is not the same thing as high event probability. High susceptibility is not a warning. A hotspot is a reason to observe and prepare, not a prophecy.

The 50-Hotspot Baseline

Earthquake and Tsunami Systems

Hotspot	Forecast mode / Sept state	Why it is included	Main cascade
Nankai Trough, Japan Earthquake / tsunami	Quantified WATCH	Japan’s official long-term evaluation keeps Nankai in its highest probability rank; the key risk is a large subduction earthquake plus tsunami affecting dense Pacific-coast cities, ports, industry and transport.	coastal evacuation, ports, industry, power, transport
Eastern Taiwan / Hualien–Taitung plate boundary Earthquake / landslide / tsunami	Quantified WATCH	Taiwan’s Central Weather Administration publishes 10-, 30- and 50-year potential maps for M6.0, M6.7 and M7.0 earthquakes; steep mountains add landslide and access-loss cascades.	urban shaking, mountain roads, landslides, ports, power
Cascadia Subduction Zone, Pacific Northwest Megathrust earthquake / tsunami	Quantified WATCH	USGS estimates a 10–15% chance of an approximately M9 full-margin rupture in the next 50 years; the 10-year project must preserve that native horizon rather than invent a conversion.	coastal tsunami, ports, bridges, utilities, regional supply chains
Puget Sound / Seattle–Tacoma Deep and crustal earthquake	Quantified WATCH	USGS estimates an 85% chance of M6.5+ deep earthquake and 17% chance of M6.5+ crustal earthquake in 50 years; deep sedimentary basins and dense infrastructure amplify consequences.	bridges, ports, pipelines, power, healthcare, communications
San Francisco Bay Area Crustal earthquake	Structural WATCH	The USGS national seismic model identifies very high California shaking hazard and explicitly models Bay Area basin effects; multiple active faults sit under a dense, highly connected economy.	transit, bridges, water, data centres, ports, housing
Greater Los Angeles / Southern San Andreas system Crustal earthquake	Structural WATCH	USGS hazard modelling treats southern California as a major high-hazard zone with basin amplification and complex multi-fault rupture possibilities.	water imports, freeways, ports, power, dense urban fabric
Alpine Fault, New Zealand South Island Crustal earthquake / landslide	Quantified WATCH	AF8 reports a 75% probability of an Alpine Fault earthquake within 50 years and a four-in-five chance the next rupture will be M8+; mountain access and infrastructure isolation dominate the cascade.	roads, rail, power, communications, isolated communities
Hikurangi–Wellington, New Zealand Subduction / crustal earthquake / tsunami	Structural WATCH	New Zealand’s national hazard work treats the Hikurangi subduction margin and Cook Strait region as major earthquake sources; Wellington’s constrained geography raises lifeline vulnerability.	port, ferries, roads, water, government continuity, tsunami
Sunda Strait megathrust, Indonesia Megathrust earthquake / tsunami	Structural WATCH	BMKG identifies Sunda Strait as a major seismic-gap concern while stressing that this is not a short-term prediction; Java–Sumatra population and infrastructure make the consequences potentially large.	Jakarta/Java access, ports, coastal settlements, industry
Mentawai–Siberut megathrust, Indonesia Megathrust earthquake / tsunami	Structural WATCH	BMKG similarly flags the Mentawai–Siberut seismic gap; western Sumatra faces near-field tsunami exposure and difficult coastal evacuation.	Padang/coastal communities, roads, ports, communications
Metro Manila / West Valley Fault Crustal earthquake	Structural WATCH	PHIVOLCS’ M7.2 West Valley Fault scenario shows potentially catastrophic shaking, liquefaction and building loss across Greater Metro Manila; it is a scenario for planning, not a date prediction.	housing, hospitals, transport, water, governance, economy
Marmara Sea / Istanbul Crustal earthquake / tsunami	Structural WATCH	Türkiye’s official hazard maps show active faulting beneath the Marmara region; enormous exposure and legacy building vulnerability make this a high-consequence structural watch.	buildings, Bosporus logistics, industry, emergency access
Central Himalaya / Kathmandu seismic corridor Crustal thrust earthquake / landslide	Structural WATCH	The India–Eurasia collision keeps the central Himalaya highly seismic; Kathmandu’s basin, dense urban fabric and mountain access routes create strong secondary landslide and logistics risk.	urban shaking, mountain roads, hospitals, heritage, supply access
Central Peru coast / Lima Subduction earthquake / tsunami	Structural WATCH	The Peru–Chile subduction margin repeatedly produces very large earthquakes; Lima combines high exposure, coastal tsunami risk and critical national infrastructure.	housing, port of Callao, water, transport, national economy
Central–southern Chile coast Subduction earthquake / tsunami	Structural WATCH	Chile sits on one of Earth’s most active megathrust margins; strong building standards reduce but do not remove tsunami, port, lifeline and regional isolation risk.	ports, mining supply chains, coastal cities, power, roads

Volcano and Lahar Systems

Hotspot	Forecast mode / Sept state	Why it is included	Main cascade
Campi Flegrei / Naples, Italy Caldera unrest / earthquake / eruption	Unrest HIGH-CONCERN	Civil Protection keeps Campi Flegrei at yellow alert; bradyseism continues, and a M4.7 event on 31 July 2026 caused injuries, building damage and evacuations. An eruption is not declared imminent.	dense urban red zone, evacuation, ports, rail, healthcare
Svartsengi–Sundhnúkur / Grindavík, Iceland Volcanic unrest / fissure eruption	Unrest HIGH-CONCERN	The Icelandic Meteorological Office reports continuing uplift and magma accumulation; another intrusion and possible eruption remains the most likely scenario, with potentially short warning.	Grindavík, roads, power, geothermal infrastructure, tourism
Mayon, Philippines Volcano / lahar	Unrest HIGH-CONCERN	PHIVOLCS lists Mayon at Alert Level 2 at the end of August 2026; volcanic activity combines with heavy tropical rainfall to create lahar and evacuation risk.	communities, roads, agriculture, lahar channels
Kanlaon, Philippines Volcano / lahar	Unrest HIGH-CONCERN	PHIVOLCS lists Kanlaon at Alert Level 2; unrest sits inside a densely inhabited agricultural island where ash and rain-remobilised deposits can extend impact beyond the crater zone.	communities, agriculture, roads, air quality
Taal, Philippines Caldera unrest / eruption	Unrest ELEVATED	PHIVOLCS lists Taal at Alert Level 1; the volcano sits inside a densely populated lake-caldera system close to Greater Manila, so even moderate unrest has high exposure significance.	evacuation, lake communities, aviation, roads, Metro Manila ash exposure
Merapi / Yogyakarta, Indonesia Effusive eruption / pyroclastic flow / lahar	Unrest ELEVATED	Indonesia’s BPPTKG reported high effusive activity and SIAGA status in 2026; Merapi’s steep drainages and monsoon rainfall create recurrent pyroclastic-flow and lahar pathways.	Yogyakarta region, villages, agriculture, river valleys
Popocatépetl / central Mexico Volcanic unrest / ash / lahar	Unrest HIGH-CONCERN	CENAPRED keeps Popocatépetl at Yellow Phase 2 in August 2026, with ongoing exhalations, ash and an exclusion zone; heavy rain can remobilise material into ravines.	Puebla–Mexico City region, aviation, ash, roads, lahars
Nevado del Ruiz, Colombia Volcano / lahar	Unrest ELEVATED	Colombia’s Geological Service lists Nevado del Ruiz at yellow activity in late August 2026; its ice/snow and steep valleys make lahars the defining high-consequence mechanism.	downstream towns, roads, rivers, agriculture
Cotopaxi, Ecuador Volcano / lahar	Structural WATCH	Cotopaxi is a heavily glaciated stratovolcano with established lahar hazard to populated valleys; its structural importance remains high even when acute unrest is low.	Latacunga, valleys, roads, water, Quito-region logistics
Mount Rainier / Puget Sound, USA Volcano / sector collapse / lahar	Structural WATCH	USGS classifies Rainier as Very High Threat and says lahars are its greatest danger; densely populated valleys contain highways, bridges, ports and pipelines.	Tacoma–Seattle lowlands, highways, ports, pipelines, communities

Cryosphere, GLOF and Landslide Systems

Hotspot	Forecast mode / Sept state	Why it is included	Main cascade
Gyirong–Rasuwa / Shigatse–Nepal corridor Glacier collapse / debris avalanche / flash flood	Secondary HIGH-CONCERN	The 26 August 2026 catastrophe is the main recent case study: extreme Himalayan relief, glacier instability, river confinement and a critical border road produced a multi-system cascade. Secondary instability and reconstruction remain live.	border trade, road access, power, communications, rescue
Sikkim / Teesta basin, India GLOF / landslide / hydropower cascade	Dynamic ELEVATED	ICIMOD-linked research identifies multiple dangerous glacial lakes in Sikkim and models downstream exposure including settlements, bridges and hydropower; steep valleys amplify cascade potential.	hydropower, Chungthang/Teesta settlements, bridges, roads
Uttarakhand–Himachal mountain corridors, India Landslide / GLOF / flash flood	Dynamic ELEVATED	Rapidly changing glaciers, intense rainfall, steep relief and dense road/hydropower construction create recurrent compound mountain-hazard pathways.	pilgrimage routes, hydropower, towns, roads, bridges
Bhutan glacial-lake / hydropower valleys GLOF / landslide	Dynamic ELEVATED	High Mountain Asia glacier retreat and potentially dangerous lakes intersect with narrow valleys, hydropower assets and downstream settlements; monitoring and early warning are central.	hydropower, valleys, settlements, cross-border rivers
Karakoram–Hunza–Chitral, Pakistan GLOF / landslide / flash flood	Dynamic ELEVATED	Glacial lakes, steep valleys, road corridors and exposed communities create repeated GLOF and slope-failure risk; warming and rainfall variability can alter loading quickly.	Karakoram Highway, villages, bridges, hydropower
Nepal Himalayan hydropower corridors GLOF / landslide / river blockage	Dynamic HIGH-CONCERN	The 2026 disaster shows how glacier and landslide hazards can enter tunnel, dam, road and power systems; multiple Himalayan hydropower corridors share steep confined valleys and limited access.	hydropower, workers, roads, power supply, rescue
Cordillera Blanca / Peru GLOF / landslide / earthquake	Dynamic ELEVATED	Rapid tropical-Andes glacier retreat, moraine-dammed lakes, steep valleys and downstream settlements create long-recognised GLOF danger; earthquake shaking can compound instability.	Huaraz-region settlements, roads, water, tourism
Taiwan central/eastern mountain corridors Typhoon landslide / earthquake landslide	Dynamic ELEVATED	Extreme relief, active tectonics, typhoon rainfall and mountain roads combine to produce high landslide susceptibility; the same corridors can be hit by both seismic and rainfall triggers.	cross-island roads, rail, isolated communities, rivers

Flood, Cyclone and Coastal Systems

Hotspot	Forecast mode / Sept state	Why it is included	Main cascade
Bangladesh Ganges–Brahmaputra–Meghna delta and coast Flood / cyclone / storm surge	Climate-conditioned ELEVATED	Low elevation, enormous river discharge, cyclone exposure and dense population create one of the world’s clearest compound flood systems; strong warning and shelter capacity reduce mortality but do not remove exposure.	people, agriculture, ports, embankments, freshwater, cities
Pakistan Indus basin River flood / heat / GLOF / drought	Climate-conditioned ELEVATED	Monsoon extremes, glacier-fed headwaters, heat and large irrigation dependence create alternating flood and water-stress risk across a nationally critical basin.	food, irrigation, cities, roads, power, health
Mekong Delta, Vietnam Flood / salinity / subsidence / sea-level rise	Climate-conditioned WATCH	A low-lying food-producing delta is exposed to river-flow change, salinity, coastal erosion, subsidence and sea-level rise; disaster risk can accumulate slowly before a threshold is crossed.	rice/aquaculture, settlements, roads, freshwater, ports
Jakarta and north Java coast Urban flood / tidal flood / subsidence	Climate-conditioned ELEVATED	Extreme rainfall, coastal and tidal flooding and land subsidence interact with very high urban and infrastructure density; local improvements can move rather than eliminate the bottleneck.	housing, transport, drainage, industry, ports, utilities
Metro Manila / Luzon typhoon–flood system Typhoon / extreme rainfall / urban flood	Climate-conditioned ELEVATED	Tropical cyclones, monsoon rainfall, river and drainage limits meet one of Asia’s largest urban regions; earthquake risk makes this a genuine multi-hazard city rather than a one-hazard problem.	housing, drainage, roads, power, healthcare, economy
Pearl River Delta / Greater Bay Area Typhoon / storm surge / extreme rainfall	Climate-conditioned WATCH	A dense low-lying coastal megaregion sits in a major western North Pacific typhoon corridor; ports, manufacturing and urban transport create global supply-chain consequences.	ports, manufacturing, airports, housing, power
Florida and U.S. Gulf Coast Hurricane / storm surge / extreme rainfall	Climate-conditioned WATCH	Long-term coastal exposure remains high even when a particular season is below normal; NOAA explicitly warns seasonal basin activity is not a landfall forecast.	coastal housing, refineries, ports, power, insurance
Caribbean island belt Hurricane / extreme rainfall / landslide / coastal surge	Climate-conditioned WATCH	Rapidly intensifying hurricanes, steep island terrain and limited infrastructure redundancy can turn one storm into long-duration power, water and health crises.	power, water, tourism, ports, healthcare, isolated islands
Mozambique Channel / Madagascar Tropical cyclone / flood / landslide	Climate-conditioned WATCH	Warm southwest Indian Ocean cyclone exposure combines with river flooding, vulnerable housing and constrained recovery capacity across Madagascar and Mozambique.	housing, roads, crops, ports, health, humanitarian access
South-West Pacific small-island systems Cyclone / sea-level / coastal flood	Climate-conditioned WATCH	WMO reports rising ocean heat, sea-level pressure and severe cyclone impacts across the South-West Pacific; small islands often have limited alternative ports, power and freshwater systems.	ports, airports, freshwater, power, housing, national continuity

Heat, Drought, Wildfire and Compound Systems

Hotspot	Forecast mode / Sept state	Why it is included	Main cascade
Iberia and southwest France Heat / drought / wildfire	Climate-conditioned HIGH-CONCERN	Western Europe experienced record June–July 2026 heat and dryness, with major wildfire conditions; Europe had record fire impacts in 2025 and long fire seasons are a persistent structural concern.	health, forests, tourism, power, agriculture, transport
Greece–Türkiye / eastern Mediterranean Heat / drought / wildfire / flash flood	Climate-conditioned ELEVATED	The Mediterranean is a rapid-warming region where severe heat, drought and wildfire can be followed by intense rainfall on burned or dry landscapes, creating compound erosion and flood risk.	tourism, settlements, power, forests, water
California wildfire–atmospheric-river belt Wildfire / post-fire debris flow / flood / heat	Climate-conditioned WATCH	The same mountain and coastal systems can move from drought and fire to extreme rainfall and debris flow; large urban–wildland interfaces and critical utilities raise cascade potential.	housing, power, roads, water, insurance, post-fire slopes
Southeast Australia Heat / bushfire / extreme rainfall	Climate-conditioned WATCH	CSIRO and the Bureau of Meteorology report more extreme fire weather, longer fire seasons and more intense heavy rainfall, especially relevant to southern and eastern Australia.	communities, forests, power, agriculture, roads
Amazon–Pantanal Drought / heat / wildfire / flood reversal	Climate-conditioned WATCH	Latin America faces a more extreme water cycle, record heat and fire; drought and fire can damage ecosystems and transport before later flood phases reverse the hydrological state.	ecosystems, smoke health, river transport, agriculture, carbon
Greater Horn of Africa Drought / flood / heat / food-system stress	Climate-conditioned WATCH	WMO reports extreme weather affecting millions while early-warning coverage remains incomplete; drought and flood can alternate and compound food, livestock and displacement pressures.	food, livestock, water, health, displacement
Sahel and West African urban flood–heat belt Extreme heat / intense rainfall / urban flood	Climate-conditioned WATCH	Rapid warming, extreme heat and flood-dominant African disaster statistics combine with fast urban growth and incomplete drainage and early-warning coverage.	health, informal housing, drainage, transport, food prices

The Eight HIGH-CONCERN Systems for September 2026

HIGH-CONCERN does not mean ‘most likely catastrophe’. It means the September baseline contains unusually strong live evidence, active unrest, current loading or recent failure that justifies closer observation.

Campi Flegrei / Naples, Italy

Why now: Civil Protection keeps Campi Flegrei at yellow alert; bradyseism continues, and a M4.7 event on 31 July 2026 caused injuries, building damage and evacuations. An eruption is not declared imminent.

Cascade if triggered: dense urban red zone, evacuation, ports, rail, healthcare.

Evidence route: Italy Civil Protection — Campi Flegrei Italy Civil Protection — Campi Flegrei earthquake 31 July 2026 update

Svartsengi–Sundhnúkur / Grindavík, Iceland

Why now: The Icelandic Meteorological Office reports continuing uplift and magma accumulation; another intrusion and possible eruption remains the most likely scenario, with potentially short warning.

Cascade if triggered: Grindavík, roads, power, geothermal infrastructure, tourism.

Evidence route: Icelandic Met Office — Reykjanes/Svartsengi situation

Mayon, Philippines

Why now: PHIVOLCS lists Mayon at Alert Level 2 at the end of August 2026; volcanic activity combines with heavy tropical rainfall to create lahar and evacuation risk.

Cascade if triggered: communities, roads, agriculture, lahar channels.

Evidence route: PHIVOLCS — current volcano bulletins WMO State of the Climate in the South-West Pacific 2025

Kanlaon, Philippines

Why now: PHIVOLCS lists Kanlaon at Alert Level 2; unrest sits inside a densely inhabited agricultural island where ash and rain-remobilised deposits can extend impact beyond the crater zone.

Cascade if triggered: communities, agriculture, roads, air quality.

Evidence route: PHIVOLCS — current volcano bulletins WMO State of the Climate in the South-West Pacific 2025

Popocatépetl / central Mexico

Why now: CENAPRED keeps Popocatépetl at Yellow Phase 2 in August 2026, with ongoing exhalations, ash and an exclusion zone; heavy rain can remobilise material into ravines.

Cascade if triggered: Puebla–Mexico City region, aviation, ash, roads, lahars.

Evidence route: CENAPRED — Popocatépetl monitoring WMO State of the Climate in Latin America and the Caribbean 2025

Gyirong–Rasuwa / Shigatse–Nepal corridor

Why now: The 26 August 2026 catastrophe is the main recent case study: extreme Himalayan relief, glacier instability, river confinement and a critical border road produced a multi-system cascade. Secondary instability and reconstruction remain live.

Cascade if triggered: border trade, road access, power, communications, rescue.

Evidence route: eduKateSG — Shigatse and the Gyirong Disaster WMO State of the Climate in Asia 2025 NASA Landslide Hazard Nowcast and Exposure (LHASA)

Nepal Himalayan hydropower corridors

Why now: The 2026 disaster shows how glacier/landslide hazards can enter tunnel, dam, road and power systems; multiple Himalayan hydropower corridors share steep confined valleys and limited access.

Cascade if triggered: hydropower, workers, roads, power supply, rescue.

Evidence route: ICIMOD library — increasing GLOF risk in Sikkim WMO State of the Climate in Asia 2025 NASA Landslide Hazard Nowcast and Exposure (LHASA)

Iberia and southwest France

Why now: Western Europe experienced record June–July 2026 heat and dryness, with major wildfire conditions; Europe had record fire impacts in 2025 and long fire seasons are a persistent structural concern.

Cascade if triggered: health, forests, tourism, power, agriculture, transport.

Evidence route: WMO European State of the Climate 2025 Copernicus — July 2026 heat/dryness and European wildfire conditions WMO Global Seasonal Climate Update Aug–Oct 2026

Why Gyirong Changes the Forecasting Project

The 26 August Gyirong–Rasuwa catastrophe is not included here merely because it is recent. It is the case study that changed this research approach.

Before the event, earlier regional research had already documented that the wider Shigatse region connects south through Gyirong towards Nepal, and that extreme relief, glaciers, river valleys, roads and border logistics occupy the same corridor. What the research had not yet done was turn that accumulated evidence into a live public question: where are the failure ingredients converging?

After the event, the Shigatse research was updated to show how the disaster affected an already-known regional corridor. This September project asks the inverse question: can the same accumulated knowledge help identify similar convergence elsewhere before failure?

To check whether the concern was genuinely useful rather than hindsight, the review compares only evidence that was available before the event with similar Himalayan corridors where no catastrophe occurred. A useful forecast must distinguish meaningful warning signs from conditions that are common across a much wider region.

The False-Positive Laboratory

Forecasting systems become dangerous when they are rewarded only for finding events that happened. A model can always look clever after the fact.

Every hotspot therefore needs controls: comparable places where the hazard did not occur; periods when an alert rose and then fell; seasons that looked dangerous but remained quiet; and signals that proved irrelevant.

False positive: concern rose, but no material event followed within the declared window.
False negative: a material event occurred without the forecast raising concern.
Wrong mechanism: the place was correctly identified as dangerous, but the event arrived through a different pathway.
Right hazard, wrong consequence: the physical event occurred but the expected human or infrastructure cascade did not.
Data-gap miss: a needed observation or source was absent rather than the analysis being wrong.
Successful prevention: a hazard occurred but warning, evacuation or engineering prevented the expected catastrophe.

A forecast that never records its misses is not a forecasting system. It is a storytelling system.

The Early-Warning Gap Is Itself a Hotspot

Observability is not uniform. Japan, Taiwan, the United States, New Zealand, Iceland, Italy and several volcano observatories publish sophisticated hazard or unrest products. Other regions contain equally serious physical hazards but much thinner sensor networks, communications, maintenance budgets or cross-border data-sharing arrangements.

WMO reports that only about 40% of African countries have multi-hazard early-warning systems. The Himalayan Gyirong case exposed another type of gap: a physical hazard system crosses political borders, while detailed observation and warning information may not cross them with the same speed.

This means the forecast must sometimes say HIGH UNCERTAINTY / OBSERVABILITY GAP rather than lower the risk simply because fewer data exist.

Singapore and Southeast Asia

For Singapore, the most consequential disasters may occur outside the island while still entering daily life through food, energy, haze, shipping, insurance, aviation, digital infrastructure and regional supply chains.

The Southeast Asian watch field therefore matters disproportionately: Sunda Strait and Mentawai seismic risk; Philippine earthquake and volcanic systems; Indonesia’s volcanic and landslide belt; Jakarta subsidence and flooding; the Mekong Delta; Pearl River typhoon exposure; and the South-West Pacific climate field.

Singapore itself should be treated as a highly connected country with important external dependencies rather than labelled ‘safe’ simply because it lacks local volcanoes or major active faults. A distant port, strait, power source or food region can become part of a Singapore disaster through a chain of dependencies.

What We Will Update Every Month

September 2026 is the baseline. We do not overwrite it. Every later Purple Forecast records what changed.

Monthly status	Meaning
STRONGER	New observations increase concern or consequence.
WEAKER	Loading or unrest falls; evidence reduces concern.
UNCHANGED	No material new state.
RESOLVED	A temporary concern returns to background.
EVENT OCCURRED	A material hazard event occurs in the declared system.
WRONG MECHANISM	The hotspot was relevant but the causal pathway differed.
FALSE POSITIVE	A time-bounded elevated watch expires without event.
MISSED EVENT	A material event occurred outside the forecast field.
DATA GAP	The system could not judge because critical observation was missing.

This is how the forecast becomes a persistent dated record rather than a monthly opinion column.

What This Report Cannot Do

It cannot tell you the exact date, place and magnitude of a future major earthquake.
It cannot replace national volcano observatories, meteorological services, geological surveys, civil-defence agencies or tsunami-warning centres.
It cannot convert a 30- or 50-year official probability into a trustworthy 10-year probability without the underlying statistical model.
It cannot infer that every strong El Niño year produces the same regional disasters.
It cannot infer that a glacier lake, steep slope or active volcano will fail merely because it is dangerous.
It cannot treat low observability as low risk.
It cannot use a hotspot label in this report as evidence that a place is objectively dangerous; independent external evidence remains essential.

The 2026–2036 Research Programme

This master report is the entrance to a larger set of supporting forecasting articles. Each article will retain the same baseline and use the scientific evidence appropriate to its hazard rather than forcing unlike hazards into a single measure.

How Disaster Forecasting Works — what can be forecast, what cannot, and why.
Earthquake and Tsunami Forecast 2026–2036 — probabilities, seismic gaps, shaking, tsunami and infrastructure.
Volcano and Lahar Forecast 2026–2036 — unrest, deformation, gas, seismicity, eruption pathways and lahars.
Glacier, GLOF, Landslide and Mountain-Corridor Forecast 2026–2036 — ice, lakes, slopes, rainfall, roads, hydropower and early warning.
Flood, Cyclone and Storm-Surge Forecast 2026–2036 — river, coast, rainfall, ocean heat, landfall uncertainty and evacuation.
Heat, Drought and Wildfire Forecast 2026–2036 — fuel, soil moisture, heat, water stress, smoke and compound reversal.
Coastal Subsidence and Sea-Level Risk Forecast 2026–2036 — slow hazards that can cross thresholds suddenly.
Infrastructure Cascade Forecast — where one failed road, port, bridge, grid, hospital or communications link multiplies damage.
The Early-Warning Gap — dangerous places we cannot observe well enough.
The False Positive Laboratory — calibration, misses and where the model worried for nothing.
Singapore and Southeast Asia — local and regional dependency implications.
Purple Forecast Scorecard — monthly recalibration and the public record of what we got right and wrong.

Primary and High-Authority Evidence Base

The baseline is deliberately source-diverse because no single organisation owns every hazard. The issuing authority retains ownership of its own data and warning language.

Final September Reading

The useful question is not: Which disaster is next?

The useful question is: Where is the world becoming less forgiving, and do we have enough information, redundancy and response capacity before the next failure arrives?

Across the 50 systems in this baseline, the same structures recur: active tectonic boundaries under dense cities; glacierised valleys carrying roads and power; volcanoes whose most dangerous effect may be a lahar rather than lava; deltas where subsidence, sea level and storm surge combine; and climate systems where heat, drought, fire and flood can arrive in sequence rather than isolation.

Some of these places will remain quiet for the entire decade. Some events will occur somewhere we did not list. Some elevated states will prove false positives. The project is successful only if those outcomes are retained and used to recalibrate the next forecast.

Prediction earns credibility by surviving the future, not by sounding convincing in the present.

Baseline: 50 hotspot systems under continuing review
Publication state: September 2026 baseline
Next formal update: October 2026, with interim updates only when a material state change justifies one.