Are GeoNet quakeml horizontal uncertainties reported in km or m?

[KennyGraham1]

Are the horizontal uncertainties in the GeoNet quakeml reported in km or m. I could the see the depth uncertainties are reported in m.

E.g. for resource_id: smi:nz.org.geonet/2016p913840

horizontal_uncertainty: 13.5701713 depth_errors.uncertainty: 12746.92009

[salichon]

Hi @fiikennyg you can refer to https://quake.ethz.ch/quakeml/docs/REC?action=AttachFile&do=view&target=QuakeML-BED-20130214b.pdf

• horizontalUncertainty Circular confidence region, given by single value of horizontal uncertainty. Unit: m

• depth Unit is meter: depth value162800/value uncertainty12500/uncertainty /depth

• lat /lon have their uncertainties in km

FYI @francovm Cheers jerome

[calum-chamberlain]

It seems like the uncertainties given in the confidenceEllipsoid are in km for GeoNet events, but should be in m for quakeml standard. From the event @fiikennyg stated, the semi_major_axis_length is 24.306108. According to QuakeML standards, this should be in m (page 18 of that pdf), but this seems unlikely given that the depth uncertainty (in m) is apparently 12746.92009. It looks like for one of these, the units are wrong, because it doesn't make sense that the longest semi-major axis should be ~1,000 times smaller than the depth uncertainty. Regarding the lat/lon uncertainties being in km. I'm surprised by this - the quakeML spec states on page 11 that SI units are used, with the exception of angles, where degrees are used. Looking further on at p.32, it looks like the uncertainties in lat-lon might be expected to be in degrees. I would have expected uncertainties of a quantity to have the same units as that quantity.