5 Calculation or estimation of risk

Note: items in italics throughout the report are those for which further definition and / or information can be found in the glossary section.

The risk posed by natural hazards can be calculated (or estimated) by the combined consideration of hazard analysis and consequence analysis.

Hazard analysis	The in-depth study and monitoring of hazards to determine their potential, origin, hazardimpact characteristics and behaviour, including their magnitude-frequency behaviour, historic performance and initiating (triggering) factors.
Consequence analysis	The identification of the type of impact or loss expected from a given hazard or hazards by determining the elements at risk and their vulnerability.

This section of the report will cover each of these components in turn, and then consider how they are combined in order to calculate or estimate risk.

5.1 Hazard analysis

Hazard analysis aims to obtain a reliable record of the type of hazard, and its frequency and magnitude within a specified area. Data limitations or management decisions may dictate that susceptibility rather than hazard is assessed. Susceptibility analysis identifies areas prone to the occurrence of hazard; it does not go as far as characterising an area on the basis of frequency and magnitude of the potentially damaging event.

Approaches to frequency-magnitude analysis are outlined in this section, based on information sources and methodology employed. These information sources and methods are not mutually exclusive and can be complementary.

Also addressed in this section are hazard impact characteristics assessment, which considers factors other than frequency and magnitude, and an overview of hazard analysis outputs.

5.1.1 Magnitude and frequency analysis

Several options exist for undertaking analysis of magnitude and frequency of natural hazards. The following highlights potential information sources and methodologies to use in this analysis.

Information sources

Instrumental records

Hazard can be determined from the instrumental record of the frequency and magnitude of potentially damaging events. In New Zealand however this is constrained by the length of record, density and location of recording devices.

Meteorological information has the longest record, while hydrological and seismic record are shorter, and marine and volcanic monitoring records are even more constrained. It needs to be remembered that an assessment of a return period event that exceeds the length of record is statistically unreliable, e.g. if there are only 20 years of flood record, little reliance can be placed on the calculated magnitude for the 50 year return period flood.

For many hazards there are no instrumental or standardised recording systems. Other techniques must therefore be used to obtain a measure of susceptibility or hazard. Increasingly, satellite remote sensing coverage is providing a valuable source of information.

Historical record

Records of previous events can be used for hazard identification and to establish the frequency and magnitude of occurrence for hazard analysis. These can include technical reports, and other documentary and oral records including Maori lore.

Comparative sequential air photo coverage, dating in New Zealand from the early 1940s, is an invaluable source for establishing the frequency and magnitude of certain hazards. Blakely and Mosley (1984) discuss the use of air photo coverage and other historical records such as catchment authority records, bed profile surveys at bridge crossings, and newspaper articles and photos for establishing river erosion and sedimentation hazards in New Zealand. Gibb (1984) has also used sequential air photo coverage to establish rates of coastal erosion and sedimentation throughout New Zealand. Sequential air photo coverage has also been used extensively in New Zealand to establish the magnitude and frequency of multiple-occurrence landslide events, such as those associated with tropical cyclones and other extreme climatic conditions (e.g. Wilson and Crozier, 2000).

Geological and geomorphic investigations

Geological and geomorphic investigations are by far the most commonly used information sources for hazard analysis in New Zealand. They provide comprehensive information on a wide range of hazards, and there is a high level of geological and geomorphological expertise available within New Zealand.

A large amount of hazard relevant information is contained within the New Zealand geological map series, and on GIS databases (http://www.gns.cri.nz/what/earthact/modelling/index.html).

Evidence of past hazards and hazard behaviour can be obtained from the interpretation of landforms, with respect to the processes responsible for their formation. Alternatively a theoretical approach may be used to identify, map, and integrate factors that are known to be associated with hazard occurrence. This approach has generally been used to identify the susceptibility to hazard.

However, a measure of hazard can be provided by using dating techniques, both relative and absolute. Dating science is improving very rapidly and new techniques and refinements have been occurring every few years.

Geotechnical investigations

These have high data requirements, are relatively expensive, and are generally used for high risk sites. The analytical techniques used indicate the degree of stability in earth materials and sites, the factors that may lead to failure, and the most appropriate treatment options. There is a high level of geotechnical expertise available within New Zealand.

Such investigations are applicable to landslides, earth deformation associated with earthquakes, liquefaction, weak foundation material, swelling clays, and other hazards relating to earth materials.

Modelling

Physical and numerical modelling can be used to determine the behaviour of various hazards, e.g. flooding, climate change, the movement of volcanic ash plumes, lahars (Hancox et al, 1997), tsunami (http://www.gns.cri.nz/news/release/william.html), and the occurrence and runout of landslides (Brooks et al, 2004).

Triggering agent analysis

The incidence of hazards (using historical sources or other evidence) can be correlated with values from the record of the triggering process to establish the threshold required for hazards. This threshold can then be applied to the process record to obtain the frequency with which that threshold is likely to be surpassed, hence providing a measure of hazard. This has been achieved in New Zealand for rainfall-triggered landslides (Glade et al, 2005), earthquake-triggered landslides, and other earthquake related hazards (Hancox et al, 1997).

The Modified Mercalli Scale (adapted for New Zealand) gives a guide to the relationship between earthquake shaking and consequences, particularly to structures and land (Dowrick and Rhoades, 1999; Dowrick and Cousins, 2002).

Expert judgements and systems

An individual or group of experts may use their experience, knowledge and professional judgement to assess hazard and risk. This approach has maximum validity in areas familiar to the experts. It is important that the basis and criteria used in judgments is recorded, as these factors may be subject to change in the future.

Methodologies

One or more of the information sources listed above may be used to estimate hazard. In some circumstances it may be possible to calculate hazard and express it quantitatively; in other cases, qualitative estimation may need to be made. Ultimately, quantitative estimation provides an objective, reproducible measure of hazard that can be compared and evaluated along with other similarly estimated hazards. In an initial scoping study, where investigation resources are limited, or where there is insufficient data, it may only be possible to produce a qualitative estimation, based on analogy with similar areas or by using expert judgement. Initial qualitative estimates can be augmented by calculated risk as more scientific evidence becomes available.

Even when quantitative estimates are available, these can assigned a verbal qualitative ranking to assist in communication with non-scientists

Quantitative

Frequency with respect to natural hazards is calculated from the number of events of a given size per unit time (e.g. the number of earthquakes of magnitude greater than 5.9 per 100 years). The reciprocal of frequency is period, the average interval between events of a given size. For example the return period (or recurrence interval) of a magnitude 6 earthquake may be 75 years (i.e. the frequency is 1.33 per 100 years). This represents an annual probability of 1.3%, often expressed in order of magnitude probabilities e.g. 1.33 x 10-2. As average values are used, the measure takes no account of irregularity of occurrence. Highly irregular events do not easily give an accurate measure of frequency and period.

Smith (2004) provides excellent examples of quantitative hazard calculation.

Changing environmental conditions such as deforestation, urbanisation, and climate change should be taken into account when establishing hazard from the historic record.

Qualitative

The likelihood of occurrence can be estimated and ranked into classes using various sources of qualitative evidence, including experience and expert judgement (Table 2). The indicative probabilities used in this table are more realistic for natural hazards than those found in other similar table (e.g. AS/NZS (2004b) tables 6.4 and 6.5). The likelihood table presented here is almost identical to that in SAA/NZS (1999), example 1, page 28.

Table 2. Qualitative measures of likelihood (Australian Geomechanics Society, 2000)

Level	Descriptor	Description	Indicative annual probability
A	Almost certain	The event is expected to occur	>=10-1
B	Likely	The event will probably occur under adverse conditions	=10-2
C	Possible	The event could occur under adverse conditions	=10-3
D	Unlikely	The event might occur under very adverse circumstances	=10-4
E	Rare	The event is conceivable but only under very exceptional circumstances	=10-5
F	Not credible	The event is inconceivable or fanciful	<=10-6

5.1.2 Hazard impact characteristics

As well as the frequency and magnitude of a hazard occurrence, other hazard characteristics are important in influencing hazard and risk management decisions. These include: threat to life; duration; areal extent; speed of onset; predictability; and controllability. These are set out in Table 3.

Table 3. Hazard impact characteristics (Crozier and Paterson, 1999)

Threat to life
Serious = several deaths possible
Concern = isolated deaths possible
Low = death highly unlikely
Duration The length of time the direct effects last (as opposed to prolonged indirect effects).
Long = weeks - seasons
Medium = hours - days
Short = seconds - minutes
Areal extent
Wide = whole district may be affected
Local = single community or region
Restricted = individual site
Speed of onset Time from first indication of hazard to the onset of effects.
Instantaneous = seconds - minutes
Medium = hours - days
Prolonged = weeks - months
Predictability The extent to which forecasts can be made of place, time, and magnitude.
High = successful procedures exist
Medium = indications only
Low = no means of forecasting or experimental procedures only
Controllability The extent to which the hazard agent itself can be modified (this does not refer to mitigating measures that reduce vulnerability).
Complete = hazard can be removed
Partial = damaging elements may be reduced
None = no way of changing physical occurrence or energy impact

Table 4 presents an example of how to apply a hazard impact characteristics assessment, in relation to fault rupture.

Table 4. Assessment of hazard impact characteristics for fault rupture (Crozier and Paterson, 1999)

Hazard Element: FAULT RUPTURE	Assessment
Threat To Life	Concern
Duration	Short
Areal Extent	Restricted
Speed Of Onset	Instantaneous
Predictability Time	Low
Predictability Place	High
Predictability Magnitude	High
Controllability	None

5.1.3 Hazard analysis outputs

Maps are frequently used as the outputs of hazard analysis to identify areas of hazard or susceptibility to hazard. The following describes the likely outputs of hazard analysis according to different natural hazard types.

Flooding

Maps showing areas of inundation, in some cases for particular return periods events e.g. 50 year and 100 year. These may be qualified by depth and velocity expected within the affected areas.

Volcanic hazards

Maps showing the extent of area affected by different processes involved in volcanic eruptions e.g. lahars and floods, lava flows, pyroclastic fall deposits, pyroclastic flows, debris avalanches and gas plumes. Usually several maps for each process are provided, based on scenarios of credible events. The areas affected may be zoned by properties of the process e.g. depth of expected pyroclastic fall or, in the case of lahar, travel times and discharge.

Seismic hazards

Maps showing the extent of area, or specific location, affected by different processes involved and may cover fault rupture, ground shaking, liquefaction, earthquake-triggered landslides, tsunami and seiche. Such maps include the location of faults with different degrees of activity and credible vertical and horizontal displacement values (fault rupture hazard). Maps can also be provided showing isoseismals i.e. ground shaking zones for scenario earthquakes, such as a rupture on a given fault segment. Scenario earthquakes may be assigned appropriate return periods to provide a measure of hazard. Expected shaking may be refined based on microzoning of ground conditions that influence enhancement of ground shaking.

Tsunami

Tsunami inundation maps based on hydrodynamic modelling of tsunami waves using detailed bathymetry and topography and in some cases building and infrastructure data (especially in built up areas).

Resonance and wave height is ideally modelled at large regional scales to allow for offshore bathymetry and shoreline shape/alignment, while inundation modelling and mapping is usually undertaken at smaller scales, but utilising information of wave heights from larger scale models. Mapping can be limited by a lack of detailed elevation data to allow accurate modelling.

Coastal hazards

Maps denoting coastal hazard areas, including areas that are affected by coastal erosion, coastal flooding (inundation) and cliff instability.

Maps may indicate the inland extent of areas likely to be affected. In a number of instances in New Zealand these have been based on the Coastal Hazard Zone Index (Gibb, 1994). The coastal hazard zone may be represented by a range of conditions (e.g. worst case, expected, and least likely) based on the range of input variables used in the assessment. Maps usually refer to the choice of a range of variables available showing worst case, most likely, least likely, rather than return periods for inundation.

Coastal erosion areas (usually on soft sandy/gravel coasts) can be defined in a number of ways. A Current Erosion Risk Zone (CERZ) for instance, takes into account natural shoreline change due to storm events, seasonal changes and longer term shoreline change trends. Effects of predicted sea level rise (e.g. 50 and 100 year planning horizons) can also be assessed by a landward shift in shoreline position with increasing water levels. These areas are usually a distance landward from a baseline (such as the toe of dune or vegetation line at a particular time) and represented on a map with lines.

Storm inundation levels are determined by calculating water levels from tide, storm surge, wave and sea level rise predictions. Inundation areas are produced using topographic data and identifying areas below the calculated inundation level.

Hazard areas for coastal cliff instability are either a line or an area along a stretch of coast determined by analysing cliff height, slope, cliff material strength, long-term erosion rates and/or other site-dependant criteria.

Landslides

Regional susceptibility maps are often created to assist in identifying areas more susceptible to landslips than others. These maps offer a broad assessment of:

• slope gradient (steepness);
• basement geology;
• vegetation; and
• known history of susceptible areas.

These maps provide a regional scale assessment only, and are often of insufficient detail to represent risk on specific sites. This is particularly the case for those sites that have been modified by earthworks, or those affected by localised variations in subsurface geology or concentrations of surface and ground water.

Regional susceptibility maps provide an indication of where specific geotechnical assessment is likely to be required to effectively manage the risk. Outside of the identified hazard areas, each site should be assessed on a case-by-case basis.

Meteorological hazards

Maps of historical cyclone paths and frequency. Because of the difficulties in predicting specific locations for meteorological hazards, these hazards are represented by probabilities of occurrence on a regional basis. NIWA's High Intensity Rainfall Design System (HIRDS) is a computer-based programme that can estimate rainfall frequency at any point in New Zealand (http://www.niwascience.co.nz/ncc/tools/hirds).

Other hazards

Susceptibility maps can be produced for many other hazards, e.g. areas likely to experience severe bush fire, snow and ice avalanches, expansive soils, and wind erosion.

5.2 Consequence Analysis

Consequence analysis includes determining or defining the elements at risk from a given hazard and their vulnerability.

Elements at risk refers to the people, buildings and structures, infrastructure, economic activities, public services, or any other defined values exposed to hazards in a given area.

The elements at risk can be obtained by survey, or by using records such as census statistics or rating databases (particularly useful for getting a quick idea as to economic values). If a measure of economic risk is required, then a value must be established for the elements (e.g. re-sale value or replacement cost). This may be obtained from valuation records or other sources. Where elements at risk are mobile, such as population within a central business district, then temporal risk needs to be taken into account. This may be provided for by expressing risk for different periods, e.g. day-time risk and night-time risk.

Vulnerability is an assessment of the proportion of damage (degree of loss) expected from exposure to hazards, and can be a function of physical or social conditions. It is directly but inversely related to the resilient qualities of the element at risk (e.g. building standards). Vulnerability can be expressed on a scale from 0 (no loss) to 1.0 (total loss), or 0% damage to 100% damage. It is referred to as the damage ratio in cases of quantifiable loss.

Vulnerability values, with some exceptions, are unavailable for most hazards. In New Zealand the most robust values available are for buildings subject to seismic shaking e.g. damage ratios (Dowrick, 2001). A few vulnerability values are available for landslide hazard (Glade et al, 2005: table 2.6). The paucity of data on vulnerability means that, in most cases, risk has to be expressed qualitatively.

The derived risk from consequence analysis can be expressed as the number of elements lost (e.g. fatalities), proportion of elements affected (e.g. 25% of the roading network), or in economic terms. Risk is generally expressed as annual expected loss or for a specified period such as the design life of a structure.

5.3 Risk Calculation

Risk calculation involves integrating the results of hazard analysis with consequence analysis (as described in the preceding sections) to derive an overall measure of risk.

The basis of risk calculation is expressed in the hazard-risk model shown below. This section of the report then addresses the two primary approaches to risk calculation - quantitative (where risk is expressed as probability) and qualitative (where risk is expressed as likelihood).

5.3.1 Hazard-Risk Model

The fundamental concepts determining the degree of risk from a hazard are reflected by the three factors represented in the hazard-risk model:

Hazard x Elements at risk x Vulnerability = Risk

Where the following definitions apply:

Hazard: The probability or likelihood of a potentially damaging event occurring in a unit of time. Often expressed as the probability of occurrence of a given magnitude of event.

Defined in this way, hazard represents a state or condition, and is assessed and applied to a particular place e.g. site, unit area of land surface, region or object (e.g. lifelines and hydro dams).

Refer also to section 5.1: Hazard Analysis of this report.

Elements at risk: The people, buildings and structures, infrastructure, economic activities, public services, or any other defined values exposed to hazards in a given area.

Refer also to section 5.2: Consequence Analysis of this report.

Vulnerability: The expected degree of loss to a given element or set of elements at risk, resulting from the occurrence of a natural hazard event of a given magnitude.

Refer also to section 5.2: Consequence Analysis of this report.

Risk: Expected losses (i.e. the probability or likelihood of specified negative consequence to life, well-being, property, economic activity, environmental, and other specified values) due to a particular hazard (or group of hazards) for a given area and time period. Risk is the product of hazard, the value of elements at risk, and vulnerability.

The hazard-risk model can be modified to take into account both temporal and spatial variability of the elements at risk (Glade et al, 2005). For example, people may only occupy a hazardous place for certain hours of the day or certain elements of risk may only occupy parts of the hazardous area.

5.3.2 Quantitative risk calculation

Quantitative risk calculation is carried out by expressing hazard frequency and consequences in measured, numerical terms and determining their product. For example, property risk can be calculated (Australian Geomechanics Society, 2000) from:

R_(Prop) = P_(H) x P_(S:H) x V _(Prop:S) x E

where:

R_(Prop) is the risk to property (annual loss of property value)

P_(H) is the annual probability of the hazardous event

P_(S:H) is the probability of spatial impact by the hazard (e.g. of the hazard impacting the property)

V _(Prop:S) is the vulnerability of the property to spatial impact (proportion of the property value lost)

E is the element at risk (e.g. the value or net present value of the property measured in monetary terms)

The term 'probability of spatial impact' [P_(S:H)] is included in the above equation because the areal unit used in assessing hazard and risk is not always identical to the area specifically impacted by the hazard. Spatial probability is the ratio of the area impacted by the hazard to the assessment area, multiplied by the ratio of the area of the element of interest to the assessment area. Similarly some elements at risk are mobile and have only temporary presence in the area impacted by the hazard. The probability of presence can be taken into account by including the term temporal probability [T_(P:S)]. For example a person may occupy a threatened building for only part of the time, or a vehicle may be in the location only for a proportion of the time.

Example of quantitative risk calculation

The problem: calculate the risk to a given highway subject to landsliding

E the element at risk is a stretch of highway that runs along the base of a range of hills from which landslides periodically impact the highway. The highway replacement value is estimated at $10,000,000.

P_(H) is hazard; research has shown that there have been 5 landslide events affecting the highway in the last 100 years. The average return period is thus 1 event in 20 years. The chances of this occurring in any one year are 1/20, i.e. 0.05 probability; in other words a 5% chance of occurrence in any one year.

P_(S:H) is the spatial probability of the contact of landslide with the highway. In other words, in these sorts of events for example 30% of the highway's length is affected by landslides, i.e. 0.3.

V is the vulnerability of the highway when hit by a landslide. In other words, in the places where the landslides impact the road, it is the proportion of the impacted stretch of highway damaged. Complete (100%) damage would be given a value of 1.0. The value in this example is 0.6, i.e. 60% damage to the value.

Risk is expressed as the expected annual loss in the dollar value of the highway. It is calculated such that:

R_(Prop) = P_(H) x P_(S:H) x V _(Prop:S) x E

becomes:

$90.000 = 0.05 x 0.3 x 0.6 x $10,000,000

It is stressed that a quantitative approach such as indicated in this equation provides only a very limited estimate of risk dealing with only one component - essentially direct damage to property in economic terms. There are likely to be many other indirect consequences associated with property damage. For example, in the case of damage to an industrial plant, this may involve loss of profit, loss of clients, loss of employment and earnings, as well as the adverse effects experienced by retailers and suppliers of raw materials associated with that industrial plant (Glade et al, 2005).

Quantitative loss on a regional basis is amenable to treatment in GIS of including algorithms in map area (polygons) to call up other thematic layers for elements at risk, vulnerability and hazard.

Quantitative risk calculation, as described above, provides a useful model for identifying many of the factors that contribute to risk as well as the extent to which they influence the total risk. For risk reduction purposes, scenarios may be run to test the benefit of reducing the influence of any of the individual factors in this model. This can then be augmented by taking into the account the cost of those treatment measures and hence provide a cost/benefit assessment of risk reduction measures.

5.3.3 Qualitative risk calculation

Qualitative estimates of risk are employed when there is a need for estimation of risk but the required numerical data on the factors influencing risk are of poor quality or unavailable. Such situations may occur when risk investigations are in their early stages. In many situations, there is sufficient information to allow a broad verbal ranking of risk which may serve to point to areas where risk reduction is required. In other words, qualitative estimation of risks can be a good starting point as a building block in a long term programme of risk assessment and management. Whereas qualitative measures may provide an accurate ranking of risk amongst a group of hazards, they rarely provide absolute values of sufficient precision for design of specific treatment measures.

Several methods are available for establishing a qualitative measure of risk, with the choice of method depending on the quality of information available. This section provides an overview of the general approach to qualitative risk calculation, and describes the FEMA ranking method, which incorporates considerations found in most methods.

General approach

A qualitative calculation of risk can be achieved by combining qualitative evaluations of likelihood (see Table 2, section 5.1.1 for definitions of likelihood) and qualitative evaluations of consequences. The integrated matrix below (Table 5) demonstrates this approach.

Table 5. Qualitative risk calculation matrix (SAA/SNZ, 1999)

Likelihood	Consequences
	Insignificant 1	Minor 2	Moderate 3	Major 4	Catastrophic 5
a: almost certain	H	H	E	E	E
b: likely	M	H	H	E	E
c: moderate	L	M	H	E	E
d: unlikely	L	L	M	H	E
e: rare	L	L	M	H	H

Key: L: Low M: Moderate H: High E: Extreme

An example of qualitative calculation of consequences is provided in Table 6. While the consequences described in this table refer to structures, the descriptors can be adapted for other elements at risk. (A similar assessment table is provided in SAA/NZS (1999) but it is oriented towards public and private institutional operations.)

Table 6. Qualitative measures of consequences to property (based on: Australian Geomechanics Society, 2000)

Level	Descriptor	Description
1	Insignificant	Little damage
2	Minor	Limited damage to part of structure, or part of site requiring some reinstatement / stabilisation works
3	Medium	Moderate damage to some structure, or significant part of the site requiring large stabilisation works
4	Major	Extensive damage to most of the structure, or extending beyond site boundaries requiring significant stabilisation works
5	Catastrophic	Structure completely destroyed or large scale damage requiring major works for stabilisation

 

FEMA ranking

FEMA ranking follows a system developed by the Federal Emergency Management Agency (FEMA) of the United States of America. It has been previously used in New Zealand in the 'Preliminary Hazards Analysis' carried out by the Taranaki Regional Council (Taranaki Regional Council, 1991), and by the Marlborough District Council (Crozier and Paterson, 1999).

Under this system, a number of parameters are assessed (such as history, vulnerability, maximum threat, and probability) using a semi-quantitative method. A weighted score is achieved for each parameter, and then these are totalled to give an overall score for the hazard. Assigning a score to each hazard provides opportunity to rank and compare the hazards and risks.

The FEMA parameters, scoring criteria and weightings used in the Marlborough study are given below in Table 7. The class descriptors used correspond to the criteria levels of the original FEMA system, whereas criteria and scoring have been made more precise in the Marlborough study. One new parameter, 'Trend in occurrence', has also been introduced.

An example of the application of this system follows in Table 8. This shows a FEMA calculation for surface faulting.

Table 7. The FEMA ranking system adapted for Marlborough (Crozier and Paterson, 1999)

History: The occurrence of a potentially damaging event.

Parameter weighting: x 2

CRITERIA	CLASS	SCORE
0-1 time in the past 100 years	Low	2
2-3 times in past 100 years	Med	5
4 or more times in past 100 years	High	10

Vulnerability: People and property: interpreted as the % damage to those impacted by the event under consideration.

Parameter weighting: x 5

CRITERIA	CLASS	SCORE
<1%	Low	2
1-10%	Med	5
>10%	High	10

Maximum threat: % of Marlborough district / community impacted.

Parameter weighting: x 10

CRITERIA	CLASS	SCORE
<1%	Low	1
1-4.9%	Low	2
5-25%	Med	5
>25%	High	10

Note: with some hazards the reference area relates to a part of the District, e.g. flooding where the percentage of floodplain is important.

Probability: Chances per year of an event expressed per 1000.

Parameter weighting: x 7

CRITERIA	CLASS	SCORE
< 1	Low	1
1-4.9	Med	3
5-9.9	Med	7
10-19.9	Med	8
20-100	Med	9
>100	High	10

Trend in occurrence: Changes for physical reasons over next 50 years.

Parameter weighting: x 2

CRITERIA	SCORE
Likely to increase	10
Possible increase	5
Stay the same	0
Possible decrease	-5
likely to decrease	-10

Table 8. Example of a FEMA ranking calculation - surface faulting (Crozier and Paterson, 1999)

Parameter	Class	Raw Score		FEMA Weighting		Weighted Score
History	0-1/100 yrs	2	X	2	=	4
Vulnerability	>10%	10	X	5	=	50
Maximum Threat	<1%	1	X	10	=	10
Probability	1-4.9/1000	3	X	7	=	21
Trend	No Change	0	X	2	=	0
				TOTAL	=	85