Task 2.4a.1: Selection of case study sites

Three case study locations will be selected, in each of which

• some earthquake history, seismotectonic and subsoil data are available

• there is a wide and representative range of building types and uses

• some basic (default) building inventory data can be acquired

• the urban authorities are actively developing mitigation plans

Likely candidate cities/regions are Istanbul, Lisbon, Catania, Thessaloniki, which are known to fulfil these criteria, but the final decision will be made after reviewing the options in conjunction with authorities in the candidate cities.

Task 2.4a.2: Develop scenario earthquakes

For each case study location, ground motion scenarios will be developed, based on a most-probable 50-year and 500-year event defined by location, magnitude and other parameters. These scenarios consist of time histories of ground velocity and acceleration at selected sites and maps of peak ground motions, Arias intensity, response spectra, power spectra, durations at regional and local scale. Synthetic ground motions can be considered as bedrock input. Earthquake scenarios will be computed with hybrid stochastic-deterministic approaches, which represents an innovative tool to predict ground motions and to characterize their spatial variability.

To allow for local site effects, microzonation studies will be performed to characterize the local site amplification. Where available, effects will be correlated with damage observed during moderate to large magnitude earthquakes. For one case-study site (Thessaloniki) scenarios will be developed from combined theoretical 1D/2D analysis and based on available weak and strong-motion recording, allowing for the influence of topography and basin effects.

Task 2.4a.3: Develop building inventories

To provide the bulk of the inventory data needed use will be made of building stock data already available for the case study locations. Building classifications will be made according to the definitions to be proposed in the work of Sub-Project 5. Building inventories will include all major building uses – residential, commercial and industrial, with special identification of areas of historic buildings. Innovative techniques based on sample surveys will be developed for providing more detailed data for structural assessment than is available from basic inventories; and remote sensing techniques will be used for studying the rate of change of building stock in the case study area. The uncertainties in building stock definition will be examined.

Task 2.4a.4: Develop vulnerability data 

A range of building performance data will be needed to be consistent with the loss-modelling software to be used, including capacity curves, fragility curves in terms of spectral values of ground motion. Development of this data has been the subject of previous EU-funded projects (Risk-UE, SEISMOCARE); further development of vulnerability data will be part of the work of this project (Cluster 2.3) and use will be made of this data with adaptations where needed for the specific building types found.

LNECLoss will implement a displacement‑based approach to compute building seismic response. An innovative procedure will be introduced to take into account second (and/or higher) modes effects on structural responses. Furthermore fragility and limits of damage state curves will be expressed not as global displacement but in terms maximum inter‑story drift allowing the possibility of reproducing local mechanisms of collapse (e.g. soft-storey collapse). LNECloss will also introduce reduction factors of capacity curves to take into account for vertical or plan irregularities.

Task 2.4a.5: Adapt and develop loss modelling software

Loss modelling will be carried out using one of two existing GIS-based software packages which have been developed for this kind of urban analysis, KOERIloss, and LNECloss; each will be developed in specific ways to make it suitable for the specific outcomes envisaged in this project; on at least one of the test sites, both packages will be used to study the consistency of output, and as a measure of uncertainty. A range of model runs will be used for each earthquake scenario, to examine the sensitivity of the results to the key inputs.

KOERILoss has been developed to generate an estimate of the losses under probabilistic earthquake hazard or exposure to a "scenario earthquake". In simplified form, the steps for using the methodology are:

• Select the region to be studied.

• Create Probabilistic Earthquake Hazard or Specify the scenario earthquake with potential fault breaks.

• Collect/compile information for delineating local soil conditions.

• Collect/compile information of inventory of building classes and other lifelines.

• Compute earthquake hazard information in the form of site-specific ground motion.

• Using vulnerability data embedded in the software (or externally provided) assess damage to different classes of buildings and casualties.

These systems, methods, and data have been coded into user-friendly software that operates through a geographic information system (GIS), MapInfo. Use of a GIS makes possible the convenient manipulation of loss estimation data concerning building stock, population, and economy. The software permits losses and consequences to be portrayed on spreadsheets and maps. KOERILoss Version 1.0 is capable to perform building damage estimation analysis using both intensity and spectral displacement based methodology. It is also able to estimate the direct economic losses and casualties related to building damages. KOERILoss is a user-friendly software that operates through Geo-cells systems. Innovations in KOERIloss in this Workpackage will consist of its extension to the range of earthquake scenario definitions appropriate to the case study areas chosen, its application to estimate expected casualties and financial losses associated with building collapse, and the introduction of a range of new vulnerability data developed in Task 2.4a.4. 

LNEC is developing an automatic seismic loss estimate methodology, integrated on a Geographic Information System. LNECloss is a high-level programming environment, composed of several modules to perform seismic risk analysis and compiled in DLL. Outputs and Inputs of each module are all compatible giving the user the possibility of creating is own application with the level of sophistication adjusted to the data and knowledge available. This innovative feature makes maximum benefits of using Windows compiled routines that can be linked and used in any Windows-based software program (ArcView, EXCEL, MatLab, etc.) 

This work is innovative in several important respects:

• Bedrock Seismic Input - Given a seismic scenario (magnitude and focus location) it computes the Power Spectral Density Function (PSDF) of the strong ground motions at bedrock level of any site at a given epicentral distance. It can use either empirical relations or seismological models for computing the spectral characteristics of those motions. 

• Local Soil Effects - Given a stratified soil profile units it computes the new PSDF for any location at the surface level, taking into account the nonlinear behaviour of the stratified geotechnical site conditions.

• Vulnerability Analysis - Giving the PSDF at surface level, it computes the response of building typologies following a displacement-based methodology based on the capacities curves. Maximum displacements and drifts will be computed using non-linear non-stationary random Gaussian analyses.

• Fragility Analysis - For a particular site, taking into account damage observed in each typology, the number of existing buildings in each typology (inventory) and respective occupancy, it computes number of building in each damage state (fragility analysis) and human causalities.

• Human Losses– Taking into account damages in each typology and the occupancy per typology it computes human casualties and homeless.

• Economic Losses - For the number of building in each damage state and human causalities it performs a cost analysis

Task 2.4a.6: Examination of uncertainty

Prediction of damage or failure of complex systems will be significantly influenced by uncertainty in its various forms. The associated consequences, both direct and indirect, are also uncertain. Particular attention will be paid to uncertainties in:

• hazard and vulnerability prediction (i.e. capacity-demand situation)

• inventory assessment

• consequence modelling in terms of appropriate utility functions

The overall objective is to develop a decision-support tool with uncertainty handling capability (including Bayesian updating) within the context of an integrated risk framework, as shown schematically in the diagram below.

Risk analysis will be undertaken for the following situations in order to support quantified mitigation recommendations:

• Prior analysis, in which risk (expected utility) is evaluated on the basis of statistical information and probabilistic modelling prior to any decision/activity.

• Posterior analysis, in which changes in branching probabilities and/or consequences are considered as a result of risk reduction/mitigation or the collection of additional information.

• Pre-posterior analysis with which optimal decisions with regard to a future range of activities may be identified.

Task 2.4a.7: Definition and evaluation of mitigation actions, and dissemination

A list of possible mitigation actions for each city to be examined will be drawn up in conjunction with the city authorities in each case. Four different types of action will be considered:

• Modification of existing buildings, using standard techniques and some innovative techniques developed in Sub-Project 7

• Demolition and redevelopment of high-risk areas

• Enhanced building control and construction standards in all future construction

• Restriction of future development in high-risk zones

• Enhancement of emergency treatment facilities

Each mitigation action will be defined as a specific programme in a particular location; costs associated will be approximately assessed; details of changed inventories, vulnerabilities and ground-shaking scenarios will be developed for the modified future city, and revised loss estimates will be calculated using the loss estimation software, thus leading to a measure of the costs and benefits of each specific mitigation action.

Based on these outcomes, a set of quantified statements about the benefits of each possible mitigation action, and its expected costs, will be developed for each city. Each will be presented with an assessment of the uncertainty involved.

These conclusions will be presented to the city authorities and related professionals in a workshop in each city, and responses will be recorded. Training will be provided for the city authorities and chosen professionals in use of the software for further studies. Model proposals for developing promising options will be provided to assist the authorities in developing these actions further.