# Deaggregation Explanation

At more than 60 cities in the Central and Eastern U.S. (CEUS) and more than 50 cities in the Western U.S. (WUS), the seismic hazard corresponding to a two per cent probability of exceedance in 50 years is deaggregated by magnitude (Mw, or moment magnitude) and by epicentral distance (CEUS) or hypocentral distance (WUS). Hazard with respect to magnitude is binned into intervals of width 0.5 Mw. Hazard with respect to epicentral distance is binned into intervals of 25 km width. The hazard probabililities are deaggregated for the following ground motion parameters: PGA, 1.0, 0.3 and 0.2 second PSA.

Four matrices of per cent contribution to hazard are available at this web site. The matrices are organized with magnitude intervals corresponding to columns and distance intervals corresponding to rows. The first row of numbers gives the upper endpoint of the magnitude interval. For example, the number 6 means that seismic sources with magnitudes in the interval 5.5 < Mw <= 6 are included in hazard calculations for that column. The first column of numbers gives the upper endpoint of the epicentral distance interval. For example, the number 150. means that source-to-station distances in the interval 125 < d <= 150 km are included in the hazard calculations for that row. Missing rows, or gaps in the matrix, correspond to distance ranges for which the greatest per cent contribution to hazard is less than 0.0005, yielding a row of zeros to the level of precision given in the below data.

For the CEUS, the lowest magnitude considered for hazard calculations is MbLg 5.0. This magnitude corresponds to Mw = 4.7 using the Johnston (1996) relationship between the two magnitudes. Thus, for CEUS cities, the interval width for the first column of contribution to hazard is about 0.3 Mw units, rather than 0.5 units, the usual interval width. For the WUS, the lowest magnitude considered for hazard calculations is Mw = 5.0.

An example graph of deaggregated seismic hazard for 1 Hz spectral acceleration for Washington, D.C., for 2% probability of exceedance in 50 years.

A second example of deaggregated seismic hazard for Washington, D. C., indicates sources at their geographic locations.

We display seismic hazard in a three dimensional map view at each of the cities for which deaggregated hazard matrices have been prepared. If you click on the "second example" above, you will see a hazard map for Washington, D.C. The map view allows you to observe the geographic locations at which significant sources of seismic hazard are concentrated. The hazard, i.e., the annual rate of ground motion exceedances, is computed for all sources (all magnitudes) in cells surrounding the city. The color of the bar over each location indicates the average magnitude of all potential seismic sources at that location. The height of the bar is proportional to the sum of hazards from all sources at the location. The maximum height (annual rate of exceedances) is given in the figure caption as "hmax." If you divide hmax by 4.04*10^-4 and multiply this by 100%, you will have the percent which that cell contributes to the total seismic hazard for your site. For many cities in California and elsewhere, a single cell may contribute over 50% of the hazard. Such a high percentage represents a clearly dominating earthquake! In contrast, there are many sites, for example, in the northeastern U.S., where no single source or even single region clearly dominates the seismic hazard.

For hazard associated with Quaternary faults, including Cascadia subduction, the hazard bar is plotted at the location on the fault nearest the city.

The cell radius is either 10 km or 25 km, depending on how concentrated seismic hazard is. In the western U.S., we generally display the hazard in 10-km annulus increments, while in the central and eastern U.S, we generally display the hazard in 25-km annulus increments. The "cells" are equal-area. This means that the cells closest to the site will span a large range of azimuths (180 degrees for 10-km increment; 120 degrees for 25-k increment), and the cells on the outer perimeter will span only a few degrees. Because the deaggregation cells are equal area (or equal volume), the maps display hazard per unit area at all locations. This strongly contrasts with the matrix representation, which sums the hazard in circular annuli about the site.

We display the hazard associated with relatively high-frequency ground shaking (5 Hz or 0.2-second period) and that which is associated with intermediate frequency ground shaking (1 Hz = 1-second period). Hazard associated with high-frequency ground shaking always emphasizes nearer sources than intermediate or long-period ground-shaking hazard. Because of this, some of the mapped regions that are displayed for the 1-sec hazard are larger than those for the 0.2-sec hazard.

The primary purpose of the mapped deaggregation of hazard is to display the sources that are considered in calculations by the National Seismic Hazard Mapping Project. We have selected many important urban areas around the U.S. These maps may also assist in the determination of earthquake parameters (magnitude and distance) that seismic engineers use in their work in earthquake-resistent design and retrofitting.