User customisation in JBA's new UK Flood Model

We’ve recently released the latest version of our UK Flood Model, which now combines JBA’s market-leading 5m UK Flood Maps with a revolutionary new approach to catastrophe model development.

This means the model is now not only the highest resolution on the market, but also the most flexible, with almost every parameter of the model modifiable by the user.

But what exactly does this mean?

Modelling Technology

JBA’s modelling technology redefines the traditional approach to catastrophe model development. For the first time ever, users can now build their own model for anywhere in the world. Our modelling technology removes traditional pre-built parameters and instead brings the model into being on-the-fly at run-time, specific to locations from a portfolio, enabling users to produce analysis tailored to their individual requirements and specific exposure. Analyses are optimised by only querying the relevant hazard and damage data for the user’s exposure.

This means:

• The ability to change almost every model parameter to generate analysis for specific requirements and views of risk
• The ability to ask and investigate in-depth ‘what if?’ questions
• An understanding of exactly what goes into the model for more openness and transparency
• The ability to configure the model to more applications, from assessing the long-term view of risk to a residential portfolio to competitive reinsurance pricing
  

It’s important to note that while users can modify model parameters, JBA will always provide a recommended view of risk and default parameters.

Hazard datA

This new technology enables users to modify and customise all hazard data within the model:

• Flood maps: the default view uses JBA’s 2020 5m UK Flood Maps for river, surface water and coastal flood. Users can modify which maps are used, whether to include new extrapolated maps, or to substitute the flood maps to a different JBA map vintage or custom flood maps.
• Defences: the default view uses JBA’s defended areas data, with the option to analyse risk as undefended; defended with overtopping; or defended with overtopping and probabilistic defence failure. Users can modify the standard of protection of existing defences, or substitute the data with custom defended area datasets.
• Event Set: the default view includes the JBA UK Flood Event Set, climate change event sets for scenarios RCP 4.5 and RCP 8.5 and time horizon 2046-2050, and a catalogue of historic and scenario events. Users can modify the event set to choose which one to run or substitute the data with a custom event set, with JBA’s expert consultancy team on-hand to help with customising data.
  

Vulnerability

Users can now modify and customise the vulnerability data and settings used within the UK Flood Model:

• Vulnerability functions: the model includes functions for residential, commercial, industrial, motor, engineering and agricultural risk types for buildings, contents and business interruption coverages. Users can substitute their own custom vulnerability functions.
• Built environment data: the model includes a spatial dataset used to assign vulnerabilities to locations with unknown risk types. Users can modify these to substitute custom built environment data.
• Flood depth threshold: the default view uses a flood depth of a typical front step height (i.e. depths shallower than this are excluded from the analysis). However, this can be modified to higher thresholds – for example 0.5m – to account for properties with known flood resilience measures in place.
• Demand surge: users have the option to incorporate demand surge by scaling the demand surge factor up in analysis.
  

Model uncertainty

The model moves away from the traditional, rigid method of estimating uncertainty by sampling from predefined distributions and instead using the uncertainty in the actual data to inform the uncertainty in the loss.

This method perturbs hazard depth and damage ratios by scaling values, runs an ensemble of the perturbations, and calculates uncertainty from the ensemble output.

Users can modify the number of runs to suit their own sensitivity analysis needs.

Location analysis and disaggregation

The UK Flood Model comes with multiple location analysis options to meet different needs – exposure locations can be imported at lat/long or aggregated to various levels, from postcode to country level.

Building footprints can be used to cover different lines of business, with specific building footprints for commercial and industrial sites and caravan parks.

Within these, users can modify a number of parameters, including:

• Buffer radius (for example increasing the buffer for larger industrial and commercial sites or removing the buffer to query flood maps at point locators)
• Preferred depth summary metric, including mean, median, min and max
• Number of disaggregated sample points if exposure is aggregated
  

And many more.

How can the UK Flood Model help me?

This flexibility and model customisation means users can truly own their view of risk for the first time, with a deeper understanding of the different parameters impacting loss and the ability to produce tailored analysis specific to individual requirements. This all helps with:

• Competitive pricing
• Exposure accumulation management and geographical hotspot identification
• Portfolio planning, diversification and optimisation
• Capital management
• Solvency and regulatory requirements
• ‘What if?’ investigation
  

How can I access the model?

While user customisation options are vast, the model remains easy to use and easy to access. The model is available via intuitive, user-friendly interfaces, including JBA’s new User Interface, Nasdaq Risk Modelling for Catastrophes, Oasis LMF, and ELEMENTS from Aon's Impact Forecasting team.

These complement JBA’s standard consultancy offering, as well as in-house installations.

If you'd like more information about the model or would like to discuss a test and evaluation free trial, get in touch with the team.