Wasatch SLC and addt'l epistemic uncertainty on GMMs

[pmpowers-usgs]

@mpetersen-usgs and @mmoschetti-usgs

The additional epistemic uncertainty that is applied to the crustal GMMs (NGAW2) used throughout the WUS is not applied to the Wasatch SLC segment models, both clustered and unclustered, but perhaps it should?

[pmpowers-usgs]

The nshmp-haz source model is configured by default to apply addf'l epistemic uncertainty to all fault source in the WUS. Uncertainty is not yet supported in cluster models. We'll therefore update the unclustered wasatch input to include uncertainty mag-distance matrix.