Hazard module (catastrophe model)

The hazard module is the part of a catastrophe model that turns each catalog event into a physical intensity at each insured site: wind gust, ground shaking, water depth. It knows nothing about buildings or policy terms. In 2011, version 11 of the RMS North Atlantic hurricane model revised, among other things, how fast a hurricane weakens once it moves over land. Modeled winds far inland went up, and so did the PMLs (probable maximum losses) of portfolios that had thought they were out of reach. The module works in two steps. First the source: for a hurricane, a track, a central pressure, a radius of maximum winds and a forward speed; for an earthquake, a fault, a magnitude and a depth. Then propagation to the site. Wind gets adjusted for surface roughness, so a forest slows it down while an open bay lets it through. Ground motion goes through prediction equations and is amplified on soft soils, which modelers summarize as Vs30, the shear-wave velocity in the top thirty meters. Secondary hazards plug in where the vendor has built them: storm surge, liquefaction, fire following earthquake. The output is an intensity footprint per event, which the vulnerability module converts into a damage ratio. At renewal meetings this is the component that gets argued over most. When two models disagree on a PML, the gap usually sits here, in the frequency of high intensities, more than in how fragile the roofs are assumed to be. A practitioner's caveat: hazard is calibrated on a short record. The North Atlantic has reliable measurements for only a few decades, and a tail drawn from so few years remains an assumption.