Detailed results for this study are hereafter compared by: 1) building location, 2) general building stock, and 3) PESH resolution.
Lastly, buildings are analyzed with PESH values at the centroid of the tract rather than the centroid of the parcel.
The assumption made by the model developers is that PESH variation across a tract is not significant, therefore grouping general building stock at the tract centroid is also not significant to loss estimates.
For the comparison of low- versus high-resolution PESH maps, the entire building inventory was studied rather than just a sample as in the location analysis.
Table 4 indicates a total loss across all building types for the low-resolution PESH study of US$ 2.
Table 5 breaks out the loss percentages (structural loss $ / building value $) by building profile, along with the difference in loss percentage by comparing a low-resolution PESH analysis with one at a high resolution.
Overall, all building types experienced a drop in loss percentage from low to high-resolution PESH input.
The Parcel-PESH methodology (along with the resulting high-resolution PESH maps) described above can be used to prepare and analyze smaller data sets in AEBM.
Further work to examine parameters in a PESH generalization from raster to vector (with the resulting impact on parcel-level loss estimation studies) may develop an appropriate compromise between hazard maps in low-resolution vector or high-resolution raster.