Secondary perils#

Getting raster values at sites#

Digital elevation data and its derivatives are often given as rasters. However, in the case of probabilistic analysis of secondary perils (particularly for risk analysis) the analyst may need to deal with sites that are not distributed according to a raster grid.

Raster values may be extracted at sites using a GIS program to perform a spatial join, but following inconvenient historical precedent, this operation often produces new data files instead of simply appending the raster values to the point data file.

Therefore we have implemented a simple function, sample_raster_at_points, to get the raster values. This function requires the filename of the raster, and the longitudes and latitudes of the sites, and returns a Numpy array with the raster values at each point. This function can be easily incorporated into a Python script or workflow in this manner.

Additional function for $V_{s30}$ estimates, vs30_from_slope_wald_allen_2007 is implemented in the engine. It requires that the slope is calculated as the gradient $\frac{dy}{dx}$ rather than an angular unit, and the study area is categorized as tectonically active or stable.

We also provide a more general, wrapper function, slope_angle_to_gradient. This function can calculate gradient from the slope in degrees (a more common formulation), and will be able to use different formulas or relations between slope and $V_{s30}$ if and when those are implemented.

Site model#

Besides usual input required for hazard assessment due to ground-shaking, several additional parameters are required to supplement the site_model.csv. Before building up the site model, one should refer to the underlying science behind these analyses and familiarise themselves with additional parameter requirements, as different models may require different parameters. Below we present an example of a site model including several parameters that are characterising soil density and wetness.

site_model.csv