Quantification of geodetic strain rate uncertainties and implications for seismic hazard estimates

Summary Geodetic velocity data provide first-order constraints on crustal surface strain rates, which in turn are linked to seismic hazard. Estimating the 2D tensor everywhere requires knowledge of field everywhere, while geodetic such as Global Navigation Satellite System (GNSS) only have spatially scattered measurements Earth. To use these estimate some type interpolation is required. In this study, we review methodologies for rate estimation and compare a suite methods, including new implementation based geostatistical method kriging, variation between methods with uncertainty one method. We calculate rates southern California using GNSS five different understand sources variability inferred rates. Uncertainty related noise station spacing (aleatoric uncertainty) minimal where dense maximum far from observations. Differences epistemic uncertainty, usually highest areas high due differences how gradients handled by methods. Parameter choices, unsurprisingly, strong influence field, propose traditional L-curve approach quantifying inherent trade-off fit models that reflective tectonic Doing so, find total representative be roughly 40 per cent, much lower value than 100 cent was found previous studies (Hearn et al., 2010). Using multiple tune parameters provides better understanding range acceptable given field. Finally, present an open-source Python package (Materna 2021) calculating Strain_2D, allows same model grid used can extended other community, interface comparing models, statistics, estimating dataset.