Based on the 340 potential faults determined under the methodology described above, a simulation was made of each one of them using the Japanese non-linear shallow water tsunami TIME model. Three nested grids were used, starting with the outer grid with a cell size of 27 arc seconds, followed by the intermediate grid with cell size of 9 arc seconds, and the inner grid with a resolution of 3 arc seconds. Figure 6 shows the 3 grids (outer grid plus the intermediate and inner grids in white).  It is in the inner grid that the model was run in its non-linear mode, allowing for the computation of coastal flooding. The induced sea bottom deformation was determined for each one of the potential faults using the Mansinha and Smylie (1971) method. For nearshore bathymetry use was made of recently acquired SHOALS bathymetry. For deeper shelf waters use was made of NOS data, and for deep water we used the Sandwell and Smith (ETOPO-2) bathymetry. For topography use was made of the USGS Digital Elevation Model data for the island, with a resolution of 30 meters. The tsunami model and the data have been shown to provide good estimates of the observed runup due to the 1918 Puerto Rico tsunami (Mercado and McCann, 1998). 

Adopting a standard hazard mitigation approach of preparing for the worst, the maximum flooding, irrespective of time of occurrence, at each computational cell for each scenario, was saved. Next, the maximum of the maximums (MOM) was obtained at each computational cell by going through each one of the 340 runs and picking out the maximum among all the runs. The maps prepared are based on the MOM. As in the simulation of the 1918 tsunami, a constant Manning coefficient of 0.025 was utilized.

Results of Task I:

Three tsunami flood maps versions were prepared and are available on Tsunami Flood Maps: 
 

1. Version I will show the scanned version of the USGS maps on top of which is overlaid the tsunami flood with sea surface elevation contours in meters above Mean Sea Level (MSL). 
2. Version II of the maps consists of the same scanned version of the USGS quadrangle maps on top of which we have overlaid the (manually digitized) inland limit of the flood. That is, it contains no information whatsoever about the elevation of the flood but only the inland limit of the flood. 
3. Version III of the maps consists of the IKONOS satellite photos of the island on top of which the same inland flood limits have been overlaid. 

(NOTE: It should be understood that due to size limitations, the maps included in this report and on the WEB page are of lower resolution than what is available in the accompanying CD.)