Interactive Tools

Probabilistic Seismic Hazard

Haiti Seismic Hazard Explorer

Enter longitude, latitude and return period, or click directly on any map. The selected location is shown simultaneously on PGA, PSA(0.2 s) and PSA(1.0 s) maps.

Model: 2010 Haiti gridded hazard curves · firm rock · Vs30 = 760 m/s. Annual frequency of exceedance = 1 / return period.
Preparing the embedded hazard grids…
PGA
g
PSA (0.2 s)
g
PSA (1.0 s)
g
PGA
PSA (0.2 s)
PSA (1.0 s)
Hazard values are derived from the supplied PGA, 0.2 s SA and 1.0 s SA gridded hazard curves. The colored surface is clipped to the supplied Haiti ADM0 country boundary (WGS84 / EPSG:4326). Spatial values use bilinear interpolation; return-period values use logarithmic interpolation between densely precomputed points from the original log-log hazard curves.
Behind the Hazard Map

How was Haiti's seismic hazard calculated?

The values shown in the interactive hazard map are based on a probabilistic seismic hazard assessment developed for Haiti after the 2010 earthquake. Rather than predicting the next earthquake, the method combines possible earthquake sources, their occurrence rates, and the shaking they may generate to estimate how frequently different levels of ground motion may be exceeded.

From earthquakes to a hazard curve

For a given location, many possible earthquakes are considered. Each earthquake source contributes some probability of producing a particular level of shaking. These contributions are combined into a hazard curve relating ground-motion intensity to its annual frequency of exceedance.

Conceptual hazard curve
Ground motion Annual frequency of exceedance 1 / Return Period

The calculation in four steps

This follows the probabilistic seismic hazard methodology used by Frankel and colleagues for the Haiti hazard maps.

1

Define earthquake sources

Major crustal faults, subduction zones and distributed background seismicity are represented.

2

Estimate earthquake rates

GPS-derived fault slip rates, earthquake history and recurrence models are used to estimate occurrence rates.

3

Estimate shaking

Ground-motion models estimate PGA and spectral acceleration produced at a site by each possible earthquake.

4

Combine the contributions

Annual exceedance frequencies from all sources are summed to construct the final hazard curve at each location.

Where can the earthquakes come from?

The model does not rely on a single fault. It combines several types of earthquake sources so that both known major faults and earthquakes occurring away from those faults contribute to the calculated hazard.

Crustal faults
Enriquillo–Plantain Garden, Septentrional and Matheux–Neiba fault zones.
Subduction zones
Northern Hispaniola subduction zone and the Muertos Trough system.
•••
Spatially smoothed seismicity
Represents hazard from earthquakes not explicitly assigned to the major modeled faults, including events at different depth ranges.

Major crustal faults in the model

The fault recurrence rates were derived primarily from slip rates, rather than simply counting recent earthquakes.

Enriquillo–Plantain Garden
7
mm/year slip rate used
Characteristic / maximum magnitude used: approximately M7.7.
Septentrional
12
mm/year slip rate used
Characteristic / maximum magnitude used: approximately M7.8.
Matheux–Neiba
1
mm/year slip rate used
Approximately M7.7 maximum magnitude. Its slip rate was identified by the authors as poorly constrained.

How is shaking estimated?

Once a possible earthquake is defined, ground-motion prediction equations estimate the shaking expected at different distances. The authors used different models for different tectonic environments rather than applying one equation everywhere.

Crustal earthquakes
Boore & Atkinson (2008), Campbell & Bozorgnia (2008), and Chiou & Youngs (2008), with equal weights.
Subduction interface
Youngs et al. (1997), Atkinson & Boore (2003), and Zhao et al. (2006).
Deep earthquakes
Youngs et al. (1997) and Atkinson & Boore (2003) were used for deep intraslab events.

What site condition does this interactive map represent?

The hazard-curve files used by this interactive tool correspond to the paper's uniform firm-rock site condition. The study also developed a separate set of maps incorporating estimated spatial variations in site condition, but those are not the curves used here.

760
Vs30 (m/s) · firm rock

What does the return period mean?

The model expresses hazard through an annual frequency of exceedance. A return period entered in the interactive tool is converted to this annual frequency before the corresponding ground-motion value is read from the hazard curve.

Annual frequency = 1 / TR
475 yr approximately 10% probability of exceedance in 50 years
975 yr approximately 5% probability of exceedance in 50 years
2,475 yr approximately 2% probability of exceedance in 50 years

What does the interactive map return?

The supplied gridded hazard-curve dataset contains three intensity measures. For the selected coordinates and return period, the tool interpolates the corresponding value for each one.

PGA
Peak Ground Acceleration
PSA (0.2 s)
5%-damped spectral acceleration at a 0.2-second period (5 Hz).
PSA (1.0 s)
5%-damped spectral acceleration at a 1.0-second period (1 Hz).
The calculation assumes a Poissonian, or time-independent, earthquake recurrence model. The time since the previous large earthquake is therefore not used directly.
Not every active fault in Haiti and Hispaniola was explicitly represented. Spatially smoothed seismicity was used in part to account for earthquakes on unmodeled sources.
The Matheux–Neiba slip rate was considered poorly constrained, producing substantial uncertainty in its hazard contribution.
The firm-rock maps assume Vs30 = 760 m/s and therefore do not represent local soil amplification.
The paper also notes that topographic amplification was not incorporated into the maps.
The authors describe these maps as an initial hazard assessment intended to improve as additional geological, geophysical, paleoseismic and site-response data become available.
Scientific basis: Frankel, A., Harmsen, S., Mueller, C., Calais, E., & Haase, J. (2011). Seismic Hazard Maps for Haiti. Earthquake Spectra, 27(S1), S23–S41.