Natural and Man-Made Hazards
CSIR NET Physical Geography: hazard classes, magnitude–frequency, risk and expected loss, hazard paradigms, PAR model, technological disasters, India's DRR.
- 20 min
On this page · 15 sections
- Prerequisites
- The Hook: The Levee That Raised the Losses
- Hazard, Risk and Disaster
- Classifying Hazards
- Magnitude and Frequency
- How Geographers Explain Disasters
- Man-Made (Technological) Hazards
- Managing Disaster Risk
- Memory Tricks
- Don’t Confuse
- Common Mistakes
- Practice
- Quick Revision Table
- Key Takeaways
- What’s Next
Prerequisites
- The physical mechanisms of earthquakes, tsunamis, volcanoes, landslides, floods, cyclones and droughts, with magnitude scales and return periods, from Natural Hazards (Part B). This page builds the geographer’s layer on top of it.
- Seismicity and the Gutenberg–Richter relation from Earthquakes: Causes, Measurement and Seismicity.
- Perception and human choice in the Man–Land Relationship.
- Probability of independent events from Counting, Probability and Number Properties.
The Hook: The Levee That Raised the Losses
Hazard, Risk and Disaster
| Term | Working definition |
|---|---|
| Hazard | a process or event that may cause loss of life, injury, damage or disruption; it has a magnitude and a probability |
| Exposure | people and assets located in hazard-prone areas |
| Vulnerability | conditions (physical, social, economic, environmental) that make exposed people and assets susceptible to harm |
| Capacity | strengths and resources that reduce harm and speed recovery |
| Resilience | the ability to resist, absorb, adapt to and recover from a hazard event |
| Risk | the likelihood and size of losses: \( \text{Risk} = f(\text{Hazard}, \text{Exposure}, \text{Vulnerability}, \text{Capacity}) \) |
| Disaster | a serious disruption of a community that exceeds its ability to cope using its own resources |
The equation often written in texts is
$$\boxed{\text{Risk} = \frac{\text{Hazard} \times \text{Exposure} \times \text{Vulnerability}}{\text{Capacity}}}$$It is a conceptual statement, not a calculation: it says that a large hazard striking an empty or well-prepared place is a small risk, and a moderate hazard striking a crowded, poor and unprepared place is a large one.
Classifying Hazards
By origin
| Class | Sub-class | Examples |
|---|---|---|
| Natural | geophysical (endogenic) | earthquake, volcanic eruption, tsunami |
| geomorphic (exogenic) | landslide, avalanche, subsidence, coastal erosion | |
| hydrological | riverine and flash floods, glacial lake outburst floods (GLOF) | |
| meteorological | tropical cyclone, storm surge, cloudburst, heat and cold waves, lightning, hail | |
| climatological | drought, wildfire, desertification | |
| biological | epidemics, pest and locust outbreaks | |
| Quasi-natural (socio-natural) | natural processes made more frequent or severe by human action | floods worsened by deforestation and encroachment, landslides from road cutting, subsidence from groundwater pumping, reservoir-triggered seismicity |
| Man-made (anthropogenic) | technological | industrial explosions and toxic releases, nuclear accidents, dam failure, structural collapse, transport accidents, oil spills |
| environmental degradation | air and water pollution episodes, toxic-waste contamination, soil salinisation | |
| social | war, terrorism, stampedes, communal conflict |
A na-tech event is a natural hazard that triggers a technological disaster: the 2011 Tōhoku earthquake and tsunami causing the Fukushima Daiichi nuclear accident is the standard example.
By speed of onset and extent
| Rapid-onset | Slow-onset (creeping) | |
|---|---|---|
| Examples | earthquake, flash flood, cyclone, industrial gas leak | drought, desertification, sea-level rise, groundwater depletion, soil salinisation |
| Warning time | seconds to days | months to decades |
| Typical response | evacuation, rescue | adaptation, policy, land-use change |
Hazards also differ in spatial extent (localised landslide versus a regional drought), frequency and predictability. These attributes, with magnitude, define a hazard’s profile.
graph LR
N[Natural
earthquake, cyclone] --> Q[Quasi-natural
deforestation floods,
pumping subsidence]
Q --> T[Technological
Bhopal, Chernobyl]
T --> S[Social
war, terrorism]
N -.na-tech.-> TThe arrow runs from hazards where people are only the victims to hazards where people are both cause and victim.
Magnitude and Frequency
Large events are rare; small ones are common. This inverse relation is the core of hazard assessment.
Return period and exceedance probability
For an event with return period \( T \) years, the annual exceedance probability is \( p = 1/T \). Over a design life of \( N \) years, the probability of at least one exceedance is
$$\boxed{R = 1 - \left(1 - \frac{1}{T}\right)^{N} \approx 1 - e^{-N/T}}$$| \( T \) | \( N \) | \( R \) |
|---|---|---|
| 100 yr | 30 yr | 26% |
| 100 yr | 100 yr | 63% |
| 475 yr | 50 yr | 10% |
A “100-year flood” is not an event that happens once a century: it has a 1% chance every year, and about a 63% chance of occurring at least once in any century. How \( T \) is estimated from a record of annual floods (plotting positions, Gumbel) is in Stream Flow, Hydrographs and Flood Frequency.
Gutenberg–Richter relation
For earthquakes in a region,
$$\boxed{\log_{10} N = a - bM}$$where \( N \) is the annual number of earthquakes of magnitude \( \ge M \). The b-value is usually close to 1, so each unit increase in magnitude makes events about ten times rarer.
Magnitude–frequency and work done
M. G. Wolman and J. P. Miller (1960) asked which events do most geomorphic work. Their answer: events of moderate magnitude and moderate frequency (for rivers, roughly the bankfull flow) transport most sediment over the long term, because very large events are too rare to dominate the total. Disaster losses can follow a different pattern: they are often dominated by the rare, extreme event.
Expected annual damage
For flood and similar hazards, damage \( D \) is plotted against annual exceedance probability \( p \). The expected annual damage (EAD) is the area under this curve:
$$\boxed{\text{EAD} = \int_0^1 D(p)\,dp \approx \sum \frac{(p_i - p_{i+1})(D_i + D_{i+1})}{2}}$$EAD is used to compare mitigation options: the benefit of a scheme is the reduction in EAD, and the benefit–cost ratio is that reduction divided by the scheme’s annualised cost.
How Geographers Explain Disasters
The behavioural (dominant) paradigm
From the 1940s to the 1970s, hazard research was led by Gilbert F. White and colleagues (Ian Burton, Robert Kates). White’s Human Adjustment to Floods (1945) argued that flood losses came largely from human occupation of flood plains, and that engineering control was only one of many possible adjustments. Burton, Kates and White’s The Environment as Hazard (1978) summarised the approach.
Key ideas:
- Hazard perception: people judge risk through experience, not statistics. Common misperceptions: the “gambler’s fallacy” (a flood last year means none this year), denial, and trust in protective works.
- Bounded rationality: people choose satisfactory rather than optimal adjustments, with incomplete information.
- Range of adjustments: a menu of choices, not only dams and embankments.
| Adjustment | Examples |
|---|---|
| Bear the loss | accept, absorb from savings |
| Share the loss | insurance, relief and aid |
| Modify the event | dams, embankments, slope stabilisation, cloud seeding |
| Modify vulnerability | forecasts and warnings, building codes, flood-proofing, raised plinths |
| Change land use | zoning, relocation, setback lines |
- Levee effect (safe-development paradox): protection encourages development in the protected area, raising losses when protection fails.
The structural (vulnerability) paradigm
From the late 1970s, critics argued that the behavioural school blamed victims’ perceptions and ignored why people lived in hazardous places at all. Kenneth Hewitt (ed.), Interpretations of Calamity (1983), and others argued that vulnerability is produced by social, economic and political structures: poverty, marginalisation, insecure land tenure, and development that pushes the poor onto steep slopes, flood plains and unsafe housing. Disasters are therefore not “natural” in their causes of loss.
The Pressure and Release (PAR) model
Piers Blaikie, Terry Cannon, Ian Davis and Ben Wisner, in At Risk (1994), set out the Pressure and Release model. A disaster is the meeting of two forces: the hazard on one side and a progression of vulnerability on the other.
graph LR
A[Root causes
limited access to power,
resources; ideologies] --> B[Dynamic pressures
lack of institutions, training,
urbanisation, deforestation]
B --> C[Unsafe conditions
fragile location, unsafe
buildings, low income]
C --> D((DISASTER
Risk = Hazard × Vulnerability))
H[Hazard
earthquake, flood,
cyclone, drought] --> D- Pressure: vulnerability builds up through the chain root causes → dynamic pressures → unsafe conditions.
- Release: risk is reduced by working back along the chain, not only by acting on the hazard.
The authors also proposed an Access model describing how households’ access to resources shapes their vulnerability and recovery.
Resilience and systems approaches
More recent work treats communities as social-ecological systems whose resilience can be built through diversity, redundancy, learning and good governance. This view underlies the “Build Back Better” principle in the Sendai Framework.
| Paradigm | Main cause of loss | Main remedy | Key names |
|---|---|---|---|
| Engineering / technocratic | the physical event | control the event (dams, embankments) | — |
| Behavioural | perception and choice | widen adjustments, education, warnings | White, Burton, Kates |
| Structural / vulnerability | social and political structures | reduce poverty and inequality, rights to land | Hewitt; Blaikie, Cannon, Davis, Wisner |
| Resilience | system rigidity | build adaptive capacity | — |
Man-Made (Technological) Hazards
Technological hazards come from the failure of human-made systems or from the release of hazardous materials. They differ from most natural hazards in that a specific organisation is usually responsible, the hazard is often invisible (gas, radiation), and its effects can be long-lasting.
Landmark technological disasters
| Event | Year | Place | What happened | Lasting outcome |
|---|---|---|---|---|
| London smog | 1952 | United Kingdom | coal smoke and SO\(_2\) trapped under a winter inversion | Clean Air Act 1956 (UK) |
| Minamata disease | 1950s | Kumamoto, Japan | methylmercury in factory effluent bioaccumulated in fish | Minamata Convention on Mercury (2013) |
| Seveso | 1976 | Italy | release of dioxin (TCDD) from a chemical plant | EU Seveso Directives on major-accident hazards |
| Love Canal | 1970s | New York, USA | housing built on a buried chemical-waste dump | US Superfund law (1980) |
| Three Mile Island | 1979 | Pennsylvania, USA | partial reactor meltdown | INES level 5 |
| Bhopal gas tragedy | 2–3 December 1984 | Madhya Pradesh, India | methyl isocyanate (MIC) leak from the Union Carbide pesticide plant | Environment (Protection) Act 1986; Public Liability Insurance Act 1991 |
| Chernobyl | 26 April 1986 | then USSR, now Ukraine | explosion of reactor 4 (RBMK type) | INES level 7; exclusion zone |
| Exxon Valdez | 1989 | Alaska, USA | tanker oil spill | US Oil Pollution Act 1990 |
| Deepwater Horizon | 2010 | Gulf of Mexico | offshore well blowout and oil spill | — |
| Fukushima Daiichi | 2011 | Japan | tsunami-triggered meltdowns (na-tech) | INES level 7 |
| Visakhapatnam styrene leak | 2020 | Andhra Pradesh, India | styrene vapour leak from a polymer plant | — |
The International Nuclear and Radiological Event Scale (INES) runs from 0 (deviation) to 7 (major accident); each step up represents roughly a tenfold increase in severity. Only Chernobyl and Fukushima have been rated 7.
Slow technological and degradation hazards
- Aral Sea: diversion of the Amu Darya and Syr Darya for irrigation from the 1960s shrank the lake drastically, exposing salt- and pesticide-laden beds to wind erosion.
- Groundwater contamination: arsenic in the Ganga–Brahmaputra delta alluvium and fluoride in hard-rock aquifers are natural (geogenic) contaminants, and human use of the aquifer can raise exposure to them; nitrate comes from fertiliser.
- Mining: subsidence and long-burning underground coal fires in the Jharia coalfield; acid mine drainage.
- Soil salinisation and waterlogging in canal-irrigated tracts.
Quasi-natural hazards in India
Several recent Indian disasters show human action amplifying a natural trigger:
| Event | Natural trigger | Human amplifiers often discussed |
|---|---|---|
| Kedarnath, Uttarakhand (June 2013) | extreme rainfall and a moraine-dammed lake breach | construction on river terraces and along the pilgrimage route |
| Mumbai floods (July 2005) | exceptional rainfall in a day | encroached drains and the Mithi river, reclaimed wetlands |
| Chamoli, Uttarakhand (February 2021) | rock–ice avalanche and debris flow | hydropower works in the valley |
| Joshimath, Uttarakhand (early 2023) | town built on old landslide debris | unplanned construction, drainage, tunnelling debated |
| Sikkim (October 2023) | glacial lake outburst flood from South Lhonak lake | dams downstream on the Teesta |
| Koyna, Maharashtra (1967) | earthquake | reservoir loading (reservoir-triggered seismicity) |
Causes of individual events are still debated; Part C questions test the category (quasi-natural, na-tech, cascading) more than blame.
Managing Disaster Risk
The disaster management cycle
graph LR
M[Mitigation] --> P[Preparedness]
P --> R[Response]
R --> Rc[Recovery and
reconstruction]
Rc --> MPre-disaster phases (mitigation, preparedness) reduce risk; post-disaster phases (response, recovery) reduce loss. “Build Back Better” uses recovery to reduce future risk.
International frameworks
| Period | Framework |
|---|---|
| 1990–1999 | International Decade for Natural Disaster Reduction |
| 1994 | Yokohama Strategy and Plan of Action |
| 2005–2015 | Hyogo Framework for Action |
| 2015–2030 | Sendai Framework for Disaster Risk Reduction: 7 global targets and 4 priorities |
Sendai priorities: (1) understanding disaster risk; (2) strengthening disaster risk governance; (3) investing in disaster risk reduction for resilience; (4) enhancing preparedness for effective response and to “Build Back Better” in recovery, rehabilitation and reconstruction.
India’s system
- Disaster Management Act, 2005: National Disaster Management Authority (NDMA, chaired by the Prime Minister), State authorities (chaired by Chief Ministers), District authorities (chaired by the District Collector or Magistrate), the National Disaster Response Force (NDRF) and the National Institute of Disaster Management (NIDM).
- National Disaster Management Plan (first issued 2016), aligned with Sendai.
- Early warning: IMD (cyclones, heavy rain, heat waves), INCOIS Hyderabad (tsunamis, from 2007), Central Water Commission (river floods), Geological Survey of India (nodal agency for landslides), National Center for Seismology (earthquake monitoring).
- Hazard zonation: seismic zones of IS 1893, flood-prone area maps, cyclone and landslide hazard maps; GIS and remote sensing underpin them.
India’s exposure is high: large parts of the country lie in moderate-to-high seismic zones, a long coastline faces cyclones and storm surges, and large areas are flood-prone or drought-prone. The physical side of each hazard is covered in Natural Hazards.
Memory Tricks
Don’t Confuse
| Pair | Difference |
|---|---|
| Hazard vs disaster | potential damaging event vs the actual disruption that exceeds coping capacity |
| Vulnerability vs exposure | susceptibility to harm vs simply being in the hazard zone |
| Quasi-natural vs na-tech | human action intensifies a natural hazard vs a natural hazard triggers a technological failure |
| Behavioural vs structural paradigm | perception and choice vs social and political production of vulnerability |
| Return period vs time to next event | average recurrence interval vs a random variable; events can come in consecutive years |
| Hyogo vs Sendai | 2005–2015, five priorities vs 2015–2030, four priorities and seven targets |
| Mitigation vs preparedness | reducing the hazard’s impact in advance (codes, zoning) vs readiness to respond (warnings, drills, stockpiles) |
| Bhopal (1984) vs Chernobyl (1986) | MIC gas, chemical plant vs reactor explosion, radiation |
Common Mistakes
Practice
All questions below are original practice questions written for this page. They are not from any CSIR NET paper.
Level 1: Foundation
- What is a na-tech event?
- Name the three links in the progression of vulnerability in the PAR model.
- What is the annual exceedance probability of a 50-year flood?
Answers
- A natural hazard that triggers a technological disaster, such as the tsunami-driven Fukushima accident.
- Root causes → dynamic pressures → unsafe conditions.
- 0.02 (2%).
Level 2: Part B Style
Practice Q1. Which of the following is best classified as a quasi-natural (socio-natural) hazard?
(a) A volcanic eruption on an uninhabited island (b) Land subsidence caused by over-pumping of groundwater (c) A methyl isocyanate leak from a pesticide plant (d) A geomagnetic storm
Answer and explanation
(b). Subsidence is a natural process (compaction of sediments) that human abstraction triggers and accelerates. (a) and (d) are natural; (c) is technological.
Practice Q2. A bridge has a design life of 30 years and is designed for the 100-year flood. The probability that this flood is equalled or exceeded at least once during its life is closest to:
(a) 0.3% (b) 3% (c) 26% (d) 30%
Answer and explanation
(c). \( 1 - 0.99^{30} = 1 - 0.740 = 0.260 \). (d) adds 1% thirty times, which overcounts years with more than one flood.
Practice Q3. The view that disaster losses are rooted in poverty, marginalisation and unequal access to resources, rather than in the hazard itself, is associated with:
(a) the behavioural paradigm of Gilbert White (b) the structural (vulnerability) paradigm associated with Hewitt and the authors of At Risk (c) environmental determinism (d) the Wolman–Miller magnitude–frequency concept
Answer and explanation
(b). The structural paradigm and the PAR model locate the causes of loss in social and political structures. White’s paradigm emphasised perception and choice of adjustments.
Practice Q4. For a region, \( \log_{10}N = 5 - M \), where \( N \) is the annual number of earthquakes of magnitude at least \( M \). The return period of earthquakes of magnitude 7 or more is:
(a) 7 years (b) 10 years (c) 100 years (d) 1,000 years
Answer and explanation
(c). \( N = 10^{5-7} = 0.01 \) per year, so \( T = 1/N = 100 \) years.
Level 3: Part C Style
Practice Q5. Flood damages for a town (₹ crore) are estimated for four floods: annual exceedance probability 0.2, damage 0; 0.1, damage 20; 0.04, damage 80; 0.01, damage 200 (ignore floods rarer than 0.01). An embankment would eliminate damage for all floods up to and including the 25-year flood (\( p \ge 0.04 \)) and costs ₹1.5 crore a year (annualised). Which statement is correct?
(a) EAD falls from ₹8.2 crore to ₹3.0 crore a year; the benefit–cost ratio is about 3.5 (b) EAD falls from ₹300 crore to ₹200 crore; the benefit–cost ratio is about 67 (c) EAD falls from ₹8.2 crore to zero; the benefit–cost ratio is about 5.5 (d) The scheme is not worthwhile because the 100-year flood still causes ₹200 crore of damage
Answer and worked solution
(a).
Without embankment: \( 0.1\times\frac{0+20}{2} + 0.06\times\frac{20+80}{2} + 0.03\times\frac{80+200}{2} = 1.0 + 3.0 + 4.2 = 8.2 \).
With embankment (damages 0, 0, 0, 200): only the last segment remains, \( 0.03\times\frac{0+200}{2} = 3.0 \).
Benefit \( = 5.2 \) crore a year; \( 5.2/1.5 = 3.47 \).
(b) sums damages instead of integrating over probability; (c) forgets that the embankment does not stop the rarest flood; (d) confuses the size of a rare loss with the expected loss. The levee effect would erode the benefit if new building raised the 200 figure.
Practice Q6. A 20-year earthquake catalogue for a region contains 120 events of magnitude \( \ge 4 \) and 12 events of magnitude \( \ge 5 \). Assuming the Gutenberg–Richter relation and a Poisson process, which statement is correct?
(a) b = 1; events of \( M \ge 6.5 \) have a return period of about 53 years; the probability of at least one in the next 50 years is about 61% (b) b = 10; events of \( M \ge 6.5 \) have a return period of about 5 years (c) b = 1; events of \( M \ge 6.5 \) have a return period of about 17 years; the probability of one in the next 50 years is 100% (d) b = 0.1; the catalogue is too short to estimate anything
Answer and worked solution
(a).
Annual rates: \( M \ge 4 \): 6 yr\(^{-1}\); \( M \ge 5 \): 0.6 yr\(^{-1}\). So \( b = \log_{10}(6/0.6) = 1 \), and \( a = \log_{10}6 + 4 = 4.778 \).
\( M \ge 6.5 \): \( N = 10^{4.778 - 6.5} = 0.0190 \) yr\(^{-1}\); \( T = 52.7 \) years.
Poisson: \( 1 - e^{-50\times 0.0190} = 1 - e^{-0.949} = 0.613 \).
(c) uses \( M \ge 6 \) (return period 16.7 years) and treats probability as certain; (b) takes the ratio 10 as b.
Practice Q7. Consider these statements:
- Gilbert White’s work showed that flood losses in the USA could rise even as spending on flood-control works increased.
- The Sendai Framework covers 2015–2030 and has four priorities for action.
- The Bhopal gas tragedy involved a leak of methyl isocyanate and preceded the Environment (Protection) Act, 1986.
- On the INES scale, Three Mile Island and Chernobyl were both rated level 7.
Which are correct?
(a) 1, 2 and 3 only (b) 1 and 4 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4
Answer and worked solution
(a).
- 1 true: the levee effect and continued flood-plain occupance.
- 2 true: 2015–2030; seven targets, four priorities.
- 3 true: December 1984; the Act followed in 1986.
- 4 false: Three Mile Island was level 5; Chernobyl and Fukushima are the level-7 events.
Quick Revision Table
| Idea | Fact or formula |
|---|---|
| Risk | \( \text{H}\times\text{E}\times\text{V}/\text{C} \) (conceptual) |
| Classes | natural; quasi-natural; technological; social; na-tech |
| Onset | rapid (quake, cyclone) vs slow (drought, desertification, sea-level rise) |
| Exceedance | \( R = 1 - (1 - 1/T)^N \approx 1 - e^{-N/T} \); \( N = T \) gives 63% |
| Gutenberg–Richter | \( \log N = a - bM \), \( b \approx 1 \) |
| Wolman–Miller | moderate, frequent events do most geomorphic work |
| EAD | area under damage–probability curve; benefit = ΔEAD |
| Behavioural paradigm | White (1945), Burton, Kates, White (1978); perception, adjustments, levee effect |
| Structural paradigm | Hewitt (1983); vulnerability is socially produced |
| PAR | Blaikie et al. (1994): root causes → dynamic pressures → unsafe conditions + hazard |
| Technological landmarks | Bhopal 1984 (MIC); Chernobyl 1986 (INES 7); Fukushima 2011 (na-tech, INES 7); Seveso 1976; Minamata |
| Frameworks | IDNDR 1990s; Yokohama 1994; Hyogo 2005–15; Sendai 2015–30 |
| India | DM Act 2005; NDMA (PM), SDMA, DDMA; NDRF; NIDM; NDMP 2016 |
| Warnings | IMD, INCOIS, CWC, GSI, NCS |
Key Takeaways
What’s Next
Many man-made hazards are pollution events; the final topic looks at how ecosystems keep their balance and how pollution disturbs it.
→ Ecological Balance and Pollution
Related: Natural Hazards · Climate, Sea-Level Change, Pollution and Ozone