99 Sayfada İstanbul Depremi — Chapter-by-Chapter Outline

Author: A. M. Celal Şengör (interviewed by Sefa Kaplan) First published: 2006 Edition covered: First and only edition, İş Bankası Kültür Yayınları, Istanbul, 2006. ISBN 9789754587142. 109 pages. Part of the "99 Sayfada" (99 Pages) series — a publisher initiative presenting expert knowledge on urgent public questions through conducted interviews.

Central thesis

Istanbul sits directly above one of the world's most dangerous seismic gaps: the unruptured segment of the North Anatolian Fault running beneath the Sea of Marmara. A magnitude 7.2–7.6 earthquake is not a remote possibility but a geological near-certainty, and the city's combination of soft-soil geology, dense unregulated construction, inadequate preparedness, and strategic geopolitical vulnerability means that when the earthquake strikes it will not merely be a natural disaster — it could threaten Turkey's national sovereignty.

This book is a conducted interview (söyleşi) between Professor A. M. Celal Şengör, among the world's leading tectonics specialists and scientific co-director of the research expedition that mapped the faults beneath the Sea of Marmara, and journalist Sefa Kaplan. It was triggered by the August 17, 1999 İzmit earthquake, which brought the Istanbul earthquake question into national consciousness. The book presents five escalating questions, each answered in depth, moving from the physics of earthquakes to their political consequences.

How close are we to an Istanbul earthquake, what would it actually destroy, and could it end Turkey's independence?

Chapter 1 — Deprem Nasıl Oluşur? (How Do Earthquakes Form?)

Central question

What is the physical mechanism that produces earthquakes, and why does Turkey — and Istanbul in particular — sit in one of the world's most active seismic zones?

Main argument

Plate tectonics as the root cause

Şengör begins from first principles: the Earth's lithosphere is broken into rigid plates that move relative to one another, driven by convection in the mantle. Where plates collide, one may subduct beneath the other (subduction zone), where they spread apart new ocean floor forms (mid-ocean ridge), and where they slide laterally past each other a strike-slip fault develops. Turkey is caught in a vice between the northward-moving Arabian Plate and the nearly stationary Eurasian Plate. This compressive stress is transmitted through the Anatolian microplate, which is being squeezed westward — like a watermelon seed pinched between two fingers — at roughly 25–30 millimetres per year.

The North Anatolian Fault

The westward escape of Anatolia is accommodated along the North Anatolian Fault (Kuzey Anadolu Fay Hattı, NAFH), a right-lateral (dextral) strike-slip fault approximately 1,100–1,200 km long, extending from the Karlıova triple junction in eastern Anatolia all the way to the Saros Gulf in the Aegean. It is one of the fastest-moving and most seismically productive faults on Earth — analogous in mechanics to California's San Andreas Fault but significantly more lethal in the density of population above it.

Fault mechanics: elastic rebound

Along strike-slip faults, stress accumulates as the two sides of the fault are locked together by friction while the deeper parts of the crust continue to move. When accumulated strain exceeds the frictional strength of the rock, the fault ruptures abruptly, releasing elastic energy as seismic waves. The two sides of the fault can displace horizontally by several metres in seconds. Şengör emphasises that in a right-lateral fault the energy is directed along the strike of the fault — horizontally — which differs from thrust (reverse) and normal faults, where energy radiates more in the dip direction.

The Marmara segment: the loaded gun

The NAFH does not stop at the Izmit Gulf; it continues beneath the Sea of Marmara, where it generates a series of pull-apart basins (çekme havzaları). The Marmara segment passes within 8 km of the Istanbul coastline at Yeşilköy. Şengör was scientific co-director of the French–Turkish expedition that deployed deep-sea sonar to map these submarine faults in detail. That survey identified the fault as a single, through-going structure — the Prince Islands Fault — rather than a diffuse system, confirming that the entire Marmara segment is capable of rupturing as a single event.

Historical earthquake sequence: a westward march

Şengör lays out the well-documented westward progression of major NAFH earthquakes in the 20th century:

• 1939 Erzincan, M7.9 — eastern end of the fault
• 1942–1944 — successive central Anatolian ruptures
• 1967 Mudurnu Valley
• 1999 Gölcük (İzmit), M7.8 — reaching the Marmara coast

Each earthquake transferred stress to the west, loading the next segment. The Marmara segment is the last unruptured link in this chain. Historical records show Istanbul earthquakes of 1509 (the "Küçük Kıyamet" / Little Apocalypse) and 1766, both estimated above M7, both consistent with a ~250-year recurrence interval on this segment. That interval has been exceeded.

Key ideas

• Earthquakes on the NAFH are caused by the westward escape of the Anatolian plate squeezed between the Arabian and Eurasian plates.
• The NAFH is a right-lateral strike-slip fault; energy radiates horizontally along the fault plane, meaning zones directly along-strike receive greatest shaking.
• The Sea of Marmara conceals the most dangerous unruptured segment; Şengör's own sonar survey confirmed a single, continuous fault line running close to Istanbul's European coast.
• The 20th-century westward sequence of major earthquakes culminated at the Marmara coast in 1999, leaving the Istanbul segment as the last loaded link.
• Historical records show major Istanbul earthquakes in 1509 and 1766; the average recurrence interval (~250 years) has been exceeded, meaning the fault is overdue.
• Understanding fault geometry is not academic: the position of the fault relative to Istanbul determines which districts will shake most and which direction the rupture will propagate.

Key takeaway

Istanbul stands over the last locked segment of one of the world's most active faults, and the historical recurrence interval for a major earthquake here has already elapsed.

Chapter 2 — Deprem İhtimali Nasıl Hesaplanır? (How Is Earthquake Probability Calculated?)

Central question

Can science give a meaningful probability for an Istanbul earthquake in a specific time window, and how is such a number derived?

Main argument

The impossibility of exact prediction

Şengör is explicit and emphatic: it is impossible to predict the time, location, and magnitude of an earthquake in advance. This is not a limitation of current technology but a fundamental property of the chaotic, non-linear fracture dynamics of crustal rock. No seismologist — anywhere — has successfully predicted a major earthquake before the event. Claims to the contrary are scientifically unfounded.

Probabilistic seismic hazard analysis (PSHA)

What science can do is calculate probability. The standard method — probabilistic seismic hazard analysis — uses two inputs: (1) the known physical dimensions of the fault segment (length, width, and locking depth) and (2) the historical and paleoseismic record of past earthquakes on that segment.

From fault geometry, the maximum possible magnitude is calculated via the relationship between fault area and seismic moment:

• Fault length of the unruptured Marmara segment: approximately 150–160 km from the Izmit Gulf mouth to the Saros Gulf
• Rock density of the crust: approximately 2.7–2.8 g/cm³
• Shear modulus and expected slip: these physical constants yield a moment magnitude of M7.2 to M7.6, with M7.6 as the theoretical maximum if the entire segment ruptures simultaneously

From historical recurrence, the elapsed time since the last major Istanbul earthquake (1766) exceeds the estimated ~250-year average return period, which shifts the conditional probability upward.

The 67% figure

A landmark 2000 study by Parsons et al. (using Coulomb stress-transfer modelling combined with historical data) calculated that the probability of a M≥7 earthquake striking the Marmara segment within 30 years was approximately 62–67%. Şengör explains that this number should not be read as "there is a one-in-three chance nothing happens" but as an engineering-grade hazard: insurance companies, bridge designers, and hospital planners work with far smaller probabilities as grounds for action.

The "aspérite" problem

Şengör introduces the technical concept of an aspérite (asperite) — a locked patch of the fault where friction is highest and stress concentration is greatest. The precise location and strength of asperities determines whether the fault ruptures in one great event or a series of smaller ones, and how the energy radiates. If the aspérite of the Marmara segment could be located and characterised, the maximum magnitude could be refined. Monitoring the aspérite is therefore a priority for research — which is why Şengör argues for continued investment in seabed instrumentation in the Marmara.

Recurrence interval versus randomness

A common misunderstanding is that a "250-year cycle" means earthquakes come like clockwork. They do not. The recurrence is a statistical average; the actual interval varies because fault strength and stress accumulation are not perfectly uniform. The correct framing is conditional probability: given that the 1766 rupture happened and that stress has been reloading since, what is the probability of rupture in the next N years? This probability increases with time, unlike a coin flip.

Key ideas

• Exact prediction of earthquakes is physically impossible; probabilistic forecasting is the only scientifically valid approach.
• The maximum magnitude of the Marmara segment earthquake is derived directly from fault length and rock physics: M7.2–7.6.
• Post-1999 Coulomb stress modelling raised the 30-year probability of M≥7 on the Marmara segment to approximately 62–67%.
• The recurrence interval concept is a statistical average, not a clock; conditional probability increases as time since the last event grows.
• The aspérite — the most locked patch — is the critical unknown for refining magnitude and ground-motion forecasts.
• Probabilistic hazard figures are explicitly designed for engineering and policy use: they are the basis on which buildings, bridges, and hospitals should be designed.

Key takeaway

Science cannot say when the next Istanbul earthquake will occur, but it can say the probability is high enough — around 62–67% within a generation — to demand immediate engineering and policy action.

Chapter 3 — Depremde Hangi Bölgeler Risk Altında? (Which Regions Are at Risk in Earthquakes?)

Central question

Given the fault geometry and the city's geology, which specific areas of Istanbul and the broader Marmara region face the greatest destruction, and why?

Main argument

Two separate hazard layers: source and site

Şengör distinguishes two independent factors that combine to produce damage. The first is proximity to the fault — shaking intensity decays with distance from the rupture. The second is local site conditions — the geological character of the ground on which buildings stand. A given earthquake at the source is amplified or attenuated differently depending on whether it travels through hard bedrock or soft alluvial sediments. Istanbul is uniquely dangerous because both factors converge in the same densely populated areas.

Source proximity: the coastline danger

The Prince Islands segment of the NAFH runs closest to Istanbul at a distance of approximately 8 km from Yeşilköy (near Atatürk Airport) on the European side. In a M7.4–7.6 rupture, Şengör estimates that Yeşilköy and Tuzla would experience Modified Mercalli Intensity X (MMI X): structural damage to well-built buildings, most masonry buildings destroyed. In the nearest coastal districts — Yeşilköy, Bakırköy, Avcılar, Küçükçekmece, Büyükçekmece on the European side, and Pendik, Tuzla on the Asian side — intensity values would reach MMI IX–X, meaning that on that ground half or more of existing building stock would be severely damaged or destroyed.

Site amplification: soft soil as a force multiplier

Şengör emphasises what engineers call zemin büyütmesi (site amplification): when seismic waves enter soft, water-saturated sedimentary soils from bedrock below, they slow down and their amplitude increases — sometimes by a factor of three to five. Istanbul's coastal lowlands, Golden Horn shores, river valley fills (Kağıthane, Alibeyköy, Küçükçekmece), and the reclaimed coastal strips of Avcılar and Zeytinburnu are built on exactly this type of alluvial and fill material. The 1999 Avcılar anomaly — where districts 60 km from the Gölcük epicentre suffered damage comparable to zones 10 km away — was precisely caused by this effect.

Soil liquefaction (zemin sıvılaşması)

Beyond amplification, water-saturated loose sands can undergo liquefaction: the seismic shaking temporarily causes the soil to behave like a liquid, causing buildings to sink, tilt, or overturn entirely. This risk is concentrated in coastal fill zones, riverbanks, and areas near the Bosphorus shoreline.

The 1766 precedent for damage geography

Historical sources document that the 1766 İstanbul earthquake (estimated M7.1–7.2) destroyed or severely damaged thousands of structures, triggered landslides in the Bosphorus, and created tsunamis along the Marmara coast. This provides a historical calibration for which parts of the city suffered most — broadly matching the modern site-amplification maps.

Beyond Istanbul: the broader Marmara region

Şengör widens the lens: a Marmara earthquake would affect not only Istanbul but the entire industrialised corridor. Heavily populated and economically critical areas at risk include:

• The İzmit–Gebze industrial belt (already damaged in 1999 but rebuilt without adequate seismic upgrading)
• The Bursa plain (built on deep alluvial fill)
• The Balıkesir plain
• The Tekirdağ coast
• The Marmara islands

The combined economic and human geography of the Marmara basin — home to roughly 25 million people and generating the majority of Turkey's GDP — means the earthquake is not an Istanbul problem but a national one.

Construction quality as multiplier

A recurring theme is that the fault itself is only part of the story. Şengör stresses that poor construction quality — the kaçak yapı (unlicensed building) problem and the widespread use of substandard concrete — multiplies the death toll. This is the one variable humans can actually control.

Key ideas

• The fault passes within 8 km of Yeşilköy; nearest European and Asian coastal districts face MMI IX–X intensities.
• Site amplification in soft-soil zones (Avcılar, Bakırköy, Zeytinburnu, Golden Horn shores) can multiply ground motion three to five times compared to bedrock sites.
• Soil liquefaction threatens coastal fill zones and riverbank deposits; buildings on such ground may sink or topple even at moderate shaking levels.
• The Avcılar anomaly in 1999 demonstrated empirically that soft-soil amplification operates at regional scale — districts far from the epicentre suffered disproportionate damage.
• The risk extends across the entire Marmara basin, home to ~25 million people and the core of Turkey's industrial economy.
• Building quality is the only variable humans control; unlicensed construction and substandard concrete convert a geological event into a human catastrophe.

Key takeaway

The most dangerous zones are not simply those closest to the fault but those combining proximity with soft-soil geology — particularly Istanbul's European coastal districts — and the damage will be amplified by decades of poor construction practice.

Chapter 4 — Depremde Tsunami Tehlikesi Var mı? (Is There Tsunami Risk in Earthquakes?)

Central question

Could the anticipated Marmara earthquake generate a tsunami, and if so how large and where would it strike?

Main argument

The tsunami mechanism in semi-enclosed basins

Tsunamis are most commonly associated with subduction-zone earthquakes (like the 2004 Indian Ocean event) that produce large vertical seafloor displacements. The North Anatolian Fault in the Marmara is a strike-slip fault, which produces primarily horizontal motion — not, at first glance, the kind of motion that generates tsunamis. However, Şengör explains two mechanisms by which the Marmara earthquake can still produce dangerous tsunami waves:

Mechanism 1 — Vertical component at fault bends

The NAFH through the Marmara is not perfectly straight. At bends and steps in the fault trace, strike-slip motion generates localised vertical displacements. The deepest depressions in the Marmara — the Central Marmara Basin and the Çınarcık Basin — are pull-apart structures formed at such bends. A major rupture would produce some vertical movement at these bends, sufficient to displace the water column.

Mechanism 2 — Submarine landslides

The more dangerous mechanism, in Şengör's assessment, is submarine mass failure. The flanks of the Marmara basins are steep and covered with unconsolidated sediment. Strong shaking from a M7+ earthquake could trigger large-scale underwater landslides. When a mass of sediment collapses from a basin wall into the deep water below, it displaces an enormous volume of water instantaneously, generating a local tsunami that can be far more energetic than the tectonic displacement alone.

The 7-metre figure

Research conducted at the request of the Turkish Air Force modelled the tsunami scenarios for the eastern Marmara. Şengör cites these calculations: if a significant underwater landslide occurs in the Central Marmara Basin, wave heights near the coasts could approach 7 metres in the worst-case scenario. The areas most exposed include the Marmara coastline of Istanbul's European districts near Yeşilköy and Florya, the coastline toward Marmaraereğlisi, and parts of the Marmara islands.

Historical precedent

Şengör refers to the 1509 earthquake — the "Küçük Kıyamet" — which contemporary Ottoman sources describe as having produced flooding of coastal areas and ships thrown onto land in Istanbul's harbour, consistent with a tsunami. The 1766 event similarly caused coastal inundation. These are not purely geological abstractions; they are documented events in Istanbul's own recorded history.

The scale problem for evacuation

Unlike Pacific-basin tsunamis where deep-ocean propagation gives hours of warning, a Marmara tsunami would strike Istanbul's shoreline within 2–4 minutes of the earthquake — faster than any warning system can operate. The implication is that no evacuation plan can protect people on the Marmara shore after the shaking starts; the only protection is not being on the shore at the moment of rupture, which requires long-term land-use planning, not emergency response.

Limits of certainty

Şengör is careful to note that the 7-metre figure is a modelled worst case, not a guarantee. The actual wave height depends on whether a submarine landslide is triggered at all, and on its volume and location — which cannot be predicted. What can be said with confidence is that tsunami risk on the Marmara coast is real, non-negligible, and substantially underappreciated by the public and by planners.

Key ideas

• The Marmara earthquake will generate some vertical seafloor displacement at fault bends, but the larger tsunami threat comes from earthquake-triggered submarine landslides.
• Worst-case modelling (Turkish Air Force study) suggests wave heights up to 7 metres near Yeşilköy and the western Marmara coast.
• Historical records of 1509 and 1766 Istanbul earthquakes describe coastal flooding consistent with tsunamis, providing empirical precedent.
• Marmara tsunami arrival time is 2–4 minutes after the earthquake — too fast for any post-event warning to enable evacuation.
• The only effective mitigation is land-use planning: critical infrastructure and high-density residential uses should not be placed on Marmara shorelines.
• The risk is modelled and real, but the actual outcome depends on stochastic variables (whether a landslide occurs) that cannot be predicted.

Key takeaway

An Istanbul earthquake could produce tsunami waves of up to 7 metres on the Marmara coastline within minutes of the main shock, driven primarily by submarine landslides rather than direct fault displacement, and no warning system can substitute for prior land-use planning.

Chapter 5 — İstanbul Depremi Türkiye'nin Bağımsızlığını Tehlikeye Atar mı? (Would the Istanbul Earthquake Threaten Turkey's Independence?)

Central question

Why does Şengör argue that a major Istanbul earthquake is not merely a humanitarian catastrophe but a potential threat to Turkey's national sovereignty?

Main argument

The scale of destruction overwhelms national capacity

Şengör's argument proceeds from a realistic damage scenario: a M7.4–7.6 earthquake on the Marmara segment would destroy or render uninhabitable hundreds of thousands of buildings in Istanbul, kill tens of thousands to hundreds of thousands of people, and simultaneously cripple the infrastructure — roads, bridges, water supply, hospitals, electricity, port facilities — of a city of fifteen million that generates roughly 25% of Turkey's GDP. Feeding, housing, and providing emergency medical care to the survivors, let alone beginning reconstruction, would demand resources and logistical capacity that far exceed what Turkey can deploy from its own means.

The forced-aid mechanism

This is the core of Şengör's geopolitical argument: because Turkey cannot recover from such a disaster alone, it will be compelled to accept massive international assistance. The international community — especially European states and the United States — will "seve seve" (readily, eagerly) provide this aid, Şengör says. But aid, especially at the scale and urgency required for a city like Istanbul, does not come without conditions, institutional involvement, and the presence of foreign personnel and organisations in critical decision-making roles. This is the entry point for sovereignty erosion.

The Treaty of Sèvres clause

Şengör delivers his most provocative historical argument: the Treaty of Sèvres (1920), the post-World War I peace treaty imposed on the Ottoman Empire that the Turkish War of Independence fought to overturn, contained a clause stipulating that Istanbul would be placed under joint international administration by the League of Nations member states. The Treaty of Lausanne (1923) superseded Sèvres and restored Turkish sovereignty, but Şengör argues that the Sèvres framework has never been entirely forgotten in certain Western diplomatic and academic circles. A catastrophic earthquake that renders Istanbul ungovernable and Turkey dependent on foreign aid could revive the argument — legally or politically — that international stewardship of the city is necessary.

The Sèvres analogy as warning, not prediction

Şengör is not predicting a military invasion or a formal return to Sèvres terms. His argument is more subtle: dependency creates leverage. A Turkey that cannot feed its own earthquake survivors, cannot reconstruct its infrastructure without foreign loans, and cannot manage a city of fifteen million without foreign logistical expertise is a Turkey whose foreign policy autonomy is severely constrained. The historical memory of Sèvres — a treaty that partitioned Turkey — gives this vulnerability a specific geopolitical shape that Turkish policymakers have historically been acutely sensitive to.

The under-investment critique

Şengör is sharply critical of the Turkish state for under-investing in earthquake preparedness and in scientific research on the Marmara fault system. He contrasts the European contributions to Marmara sea-floor research (he cites European scientific and financial contributions that dwarfed the Turkish state's own funding) with domestic neglect. The implication is that Turkey is allowing its own vulnerability to deepen through inaction — and that this inaction is itself a form of strategic failure, not merely a scientific or engineering shortcoming.

Preparedness as sovereignty

The chapter's positive argument is the inversion of its negative one: preparedness is sovereignty. A Turkey that seismically retrofits its buildings, enforces construction codes, plans for mass-casualty response, and invests in understanding its own fault system is a Turkey that can recover from the earthquake on its own terms. The earthquake itself cannot be prevented; Turkey's dependence on foreign aid in its aftermath can be. Şengör frames earthquake preparedness not as a technical nicety but as a matter of national survival.

Key ideas

• The post-earthquake humanitarian and economic demands will exceed Turkey's independent capacity, forcing large-scale acceptance of foreign aid.
• The Treaty of Sèvres (1920) contained an internationalization clause for Istanbul; while superseded by Lausanne (1923), Şengör sees it as a template that could be revived in the context of post-earthquake dependency.
• Dependency on foreign aid creates political and institutional leverage for external actors, constraining Turkish foreign policy autonomy even without formal territorial claims.
• Turkey's under-investment in earthquake science and preparedness — contrasted with European financial contributions to Marmara research — is itself a strategic failure.
• Preparedness is framed as sovereignty: the ability to recover without dependency is the ability to remain politically independent.
• The geopolitical argument is a warning about incentive structures and historical precedent, not a prediction of invasion or formal treaty revision.

Key takeaway

A major Istanbul earthquake would overwhelm Turkey's independent recovery capacity, force dependency on foreign aid, and thereby create conditions — echoing the unresolved internationalization clauses of the Treaty of Sèvres — that could compromise Turkish sovereignty; earthquake preparedness is therefore not a technical issue but a matter of national independence.

The book's overall argument

1. Chapter 1 (Deprem Nasıl Oluşur?) — Establishes the geophysical foundation: Turkey is squeezed westward by the collision of the Arabian and Eurasian plates; this motion is released along the North Anatolian Fault, the last unruptured segment of which lies directly beneath the Sea of Marmara, 8 km from Istanbul's coast; historical earthquakes in 1509 and 1766 demonstrate that this segment produces major events, and the 20th-century westward sequence of ruptures has loaded it to breaking point.
2. Chapter 2 (Deprem İhtimali Nasıl Hesaplanır?) — Moves from physics to probability: exact prediction is impossible, but probabilistic seismic hazard analysis based on fault geometry (length, rock physics → maximum M7.2–7.6) and historical recurrence yields a roughly 62–67% probability of a M≥7 earthquake within a 30-year window — a number that demands engineering and policy action even if it is not certainty.
3. Chapter 3 (Depremde Hangi Bölgeler Risk Altında?) — Translates the hazard into geography: proximity to the fault sets the baseline shaking; site amplification in soft alluvial soils multiplies it; soil liquefaction threatens coastal fills; the most dangerous zones are Istanbul's European coastal districts (Yeşilköy, Avcılar, Bakırköy) and the entire Marmara basin; poor construction quality is the human-made factor that converts shaking into death.
4. Chapter 4 (Depremde Tsunami Tehlikesi Var mı?) — Adds a second lethal mechanism: earthquake-triggered submarine landslides in the deep Marmara basins can generate tsunamis up to 7 metres; these arrive within 2–4 minutes, making warning-based evacuation impossible; the 1509 and 1766 historical records confirm the precedent; the only mitigation is long-term land-use planning.
5. Chapter 5 (İstanbul Depremi Türkiye'nin Bağımsızlığını Tehlikeye Atar mı?) — Elevates the argument from natural science to geopolitics: the earthquake's aftermath will exceed Turkey's independent recovery capacity; the resulting dependency on foreign aid creates sovereignty risk, amplified by the dormant Sèvres internationalization clause for Istanbul; preparedness is therefore a matter of national independence, not just public safety.

Common misunderstandings

Misunderstanding: The Istanbul earthquake is a distant future possibility

The book directly refutes this framing. The fault has been loading since 1766, the historical recurrence interval has been exceeded, the 1999 earthquake transferred additional stress to the Marmara segment, and probabilistic analysis yields a 62–67% probability within a generation. "Distant future" is not a scientifically supportable characterisation.

Misunderstanding: Earthquake prediction is almost possible, and scientists just need more data

Şengör is unequivocal: earthquake prediction — specifying time, location, and magnitude — is not merely technically difficult but involves chaotic, non-linear dynamics that are fundamentally unpredictable. More monitoring improves hazard assessment and site characterisation but does not make exact prediction achievable. Expecting a warning before a major earthquake is not a realistic basis for preparedness.

Misunderstanding: The North Anatolian Fault in the Marmara is a diffuse zone, not a single line

This was a live scientific debate before the Franco-Turkish Marmara survey that Şengör co-led. Some researchers had proposed that the Marmara fault was branched and distributed. The sonar mapping confirmed a single, through-going, narrow fault zone — the Prince Islands fault — capable of rupturing as a single M7+ event. The "distributed fault" model was effectively ruled out by this research.

Misunderstanding: A strike-slip earthquake cannot generate a tsunami

The dominant energy release in a strike-slip earthquake is horizontal, not vertical, which limits direct tsunami generation from fault motion. However, the Marmara earthquake can generate tsunamis via a different mechanism: submarine landslides triggered by the strong shaking. The 1509 and 1766 historical tsunamis confirm this. The tsunami risk is real regardless of fault type.

Misunderstanding: Earthquake preparedness is purely a technical/engineering matter with no political dimension

Şengör's final chapter explicitly challenges this. The scale of a Marmara earthquake disaster exceeds what Turkey can manage independently; the forced acceptance of foreign aid, combined with the dormant Sèvres internationalization clause, creates sovereignty risk. Preparedness is therefore a national security and political question, not only an engineering one.

Misunderstanding: The danger is limited to Istanbul proper

The Marmara region is home to roughly 25 million people, most of Turkey's industrial base, and the country's main port infrastructure. The earthquake danger extends to the İzmit–Gebze industrial corridor, the Bursa and Balıkesir plains, Tekirdağ, and the Marmara islands — making this a national economic disaster, not a local urban one.

Central paradox / key insight

The book's central paradox is this: the earthquake that will strike Istanbul is an entirely natural, geologically determined event beyond human control — and yet the catastrophe it produces is almost entirely a product of human choices. The fault will rupture when it must; the death toll, the economic collapse, and the sovereignty risk are determined by decisions made now. Poor construction codes, unlicensed buildings, inadequate emergency planning, underinvestment in fault monitoring, and political short-termism are not caused by the earthquake — they are caused by governments and citizens who treat the earthquake as someone else's future problem.

Şengör's most unsettling implication is that Turkey's vulnerability is self-inflicted and reversible. Unlike the fault itself, the cascade from earthquake to national catastrophe to sovereignty crisis is not inevitable — it is the product of choices that could still be reversed by investment in preparedness, enforcement of building standards, and land-use planning. The earthquake cannot be stopped; Turkey's dependence in its aftermath can be.

The earthquake is geological fate; the disaster is political choice.

Important concepts

Kuzey Anadolu Fay Hattı (North Anatolian Fault, NAFH)

A right-lateral (dextral) strike-slip fault approximately 1,100–1,200 km long, extending from Karlıova in eastern Anatolia to the Saros Gulf in the Aegean. It accommodates the westward escape of the Anatolian microplate, squeezed between the northward-moving Arabian Plate and the stationary Eurasian Plate. Among the fastest-moving and most seismically productive continental strike-slip faults on Earth.

Sağ yanal doğrultu atımlı fay (Right-lateral strike-slip fault)

A fault where the two sides move horizontally past each other, with the right side moving toward the observer. In the NAFH, the southern block (Anatolia) moves west relative to the northern block (Eurasia). Energy is radiated primarily along the fault's horizontal strike direction, rather than vertically.

Asperite (Aspérite)

A locked patch on the fault plane where friction is especially high and stress concentration is greatest. The rupture of an asperite releases a disproportionate amount of seismic energy. The location and size of the Marmara segment's aspérite is a critical unknown for hazard forecasting.

Zemin büyütmesi (Site amplification)

The increase in ground-motion amplitude when seismic waves travel from hard bedrock into overlying soft sediments. Waves slow down and grow taller. Amplification factors of three to five are documented in Istanbul's coastal alluvial zones. The 1999 Avcılar anomaly is the canonical local example.

Zemin sıvılaşması (Soil liquefaction)

The process by which water-saturated loose sand loses shear strength during strong shaking and temporarily behaves like a liquid. Buildings on such soil can sink, tilt, or overturn. Concentrated in coastal fill zones, riverbanks, and reclaimed shorelines.

Olasılıksal sismik tehlike analizi (Probabilistic Seismic Hazard Analysis, PSHA)

A methodology that combines fault geometry, recurrence rates, and attenuation relationships to estimate the probability that ground shaking above a specified level will occur at a given site within a given time window. The 62–67% probability of M≥7 on the Marmara segment within 30 years is a PSHA output.

Coulomb gerilme transferi (Coulomb stress transfer)

The mechanism by which one earthquake changes the static stress on nearby fault segments, advancing or delaying their rupture. The 1999 Gölcük earthquake transferred positive Coulomb stress to the western Marmara segment, increasing its conditional probability of rupture.

Çekme havzası / pull-apart basin

A depression formed where a strike-slip fault bends or steps, creating local extension. The Marmara Sea's deep basins (Çınarcık, Central Marmara, Tekirdağ) are pull-apart structures; they are steep-walled and sediment-draped, creating the conditions for earthquake-triggered submarine landslides.

Sevr Antlaşması (Treaty of Sèvres, 1920)

The post-WWI treaty imposed on the Ottoman Empire, containing a clause for internationalisation of Istanbul under the League of Nations. It was superseded by the Treaty of Lausanne (1923) after Turkey's War of Independence. Şengör invokes the Sèvres clause as a dormant precedent that could be revived if Turkey is rendered dependent on foreign assistance after an earthquake.

Söyleşi formatı (Interview format)

The book is a conducted interview between Şengör and journalist Sefa Kaplan, part of İş Bankası's "99 Sayfada" series. Each section is structured as a substantive question-and-answer exchange, making expert geological knowledge accessible to a general Turkish readership. The format is not traditional academic exposition but a dialogic popularisation.

References and Web Links

Primary book and edition information

• Şengör, A. M. Celal (interviewed by Sefa Kaplan). 99 Sayfada İstanbul Depremi. İş Bankası Kültür Yayınları, Istanbul, 2006. ISBN 9789754587142. 109 pages.
  • Open Library: İstanbul depremi — Şengör, 2006
  • Nadir Kitap listing

Author background

• Celâl Şengör — Wikipedia (English)
• Ali Mehmet Celal Şengör — ResearchGate profile

The North Anatolian Fault

• Kuzey Anadolu Fay Hattı — Wikipedia (Turkish)
• The North Anatolian Fault: A new look — ResearchGate PDF
• The North Anatolian Fault and the North Anatolian Shear Zone — Springer

Şengör's public statements on Istanbul earthquake risk

• Şengör: İstanbul depremi olursa Türkiye bağımsızlığını kaybedebilir — T24
• Şengör: Marmara'da 7 metre tsunami dalgası olabilir — Hürriyet
• Şengör: İstanbul için depremin eli kulağında — Cumhuriyet
• Şengör'den Sevr hatırlatması: İstanbul depreminde bağımsızlığı kaybedebiliriz — Onedio

Seismic hazard and probability research

• TARİHİ VERİLER IŞIĞINDA BEKLENEN İSTANBUL DEPREMİ — 21. Yüzyıl Türkiye Enstitüsü
• Deprem Olgusu, Türkiye'de Deprem ve İstanbul Depremi — İTÜ Vakfı
• Marmara Denizi Fay Haritası — TÜBİTAK Bilim ve Teknik (PDF)

Istanbul Metropolitan Municipality earthquake risk resources

• IBB Deprem Zemin — Olası Deprem Kayıp Tahminleri

Additional chapter summaries and study resources

These are secondary summaries and should be used alongside, rather than instead of, the original book.

• Kitapaloku book page — 99 Sayfada İstanbul Depremi
• Sputnik Türkiye: Şengör İstanbul depremi Türkiye'nin bağımsızlığını kaybetmesiyle neticelenebilir