Talk:2020 Zagreb earthquake

Epicentre coordinates?
The detailed Croatian source links to https://www.youtube.com/watch?v=FI4Vqm_FvGI where one can infer from the shown Openstreetmap of the area that the coordinates are roughly 45.86°N, 16.036°W. This differs non-trivially from 45.907°N, 15.97°W cited from the USGS page. Indeed the difference is about 7 km, which is comparable to the distance to the city center. The epicentre of the second, largest aftershock is indeed a bit more to the northwest. Do we have these clearly more precise coordinates somewhere where we can cite them from? --Joy &#91;shallot&#93; (talk) 13:15, 1 May 2020 (UTC)


 * Note also that they have it in image form at http://www.pmf.unizg.hr/geof/seizmoloska_sluzba/o_zagrebackom_potresu_2020?@=1lq1i#news_97581, where the latitudes and longitudes are indicated more clearly, yet still roughly. --Joy &#91;shallot&#93; (talk) 13:20, 1 May 2020 (UTC)


 * The epicentral location of an earthquake like this is calculated by many agencies. If you click on the "authoritative data" link of the International Seismological Centre in the External links section, you will see 13 different calculations for the epicentre from organisations based in Germany (3), Spain (2), US, Czech Republic, France, Austria, Italy, Russia, Australia and Denmark. The ANSS gives another location again, so with the Croatian one, that gives us 15 to choose from. There is no way at this stage that we can do better than to use the ANSS values as we have no idea where the true location lies. In time, once seismologists have had the opportunity to review all the data, the epicentre will probably be revised, although there may still be alternatives. With earthquakes there are also normally many versions of the magnitude, depth and other parameters. Some of these can be handled as ranges, but that's a bit trickier with the location of the epicentre. Mikenorton (talk) 12:12, 2 May 2020 (UTC)


 * I would support the Croatian Seismological Service version (45.86°N, 16.036°W). This figure is not preliminary and was released weeks after the earthquake. CSS has stated it's based on an extensive review of their data. CSS's preliminary figure was something like 45.8806°N, 16.016°W. I don't see why we should particularly favour ANSS, whose observation networks are based on a different continent, for fine-grained epicentre location.  Daß &thinsp;  Wölf  16:53, 2 May 2020 (UTC)


 * If you can provide a citation for that, I don't see any objection to replacing the ANSS values, if it's not just a preliminary value (and I've struck out part of my earlier comment that was a bit OTT). We generally use the ANSS coordinates, but sometimes local agencies have better crustal seismic velocity models, which may improve their location and depth estimates. I still reckon that the location is likely to be revised when the scientific papers are written. Mikenorton (talk) 17:07, 2 May 2020 (UTC)


 * Sorry for the late response, I've been busy last few days. Here's the seismological service's FAQ which states 45.85°N, 16.03°W. Their webpage is unfortunately messy and missing a lot of statements they made in media interviews. I haven't been able to find an online confirmation by the service's seismologists that these are reviewed figures which were calibrated after a previous rougher estimate, I believe I heard this in one of the interviews they gave on TV. I'll try to dig those up when I have some free time. I agree that this won't be the final figure, but I think it's likely to be a better guess than others.  Daß &thinsp;  Wölf  16:48, 7 May 2020 (UTC)


 * Thanks for the link - my estimate from the map that's linked to from that page also matches your suggested coordinates, but reading the lat long off a map is definitely a last resort. Mikenorton (talk) 17:08, 7 May 2020 (UTC)


 * While we're at it, I don't think it is realistic to assess all these sources as being of equal value with regard to deciding which coordinates to list in the article top-right corner, or even as the current best guess as the epicentre. They're scattered all over the place, over an area of 50-100km, so it's reasonable not to prioritize those that don't correspond to the information we already have from the ground about where the earthquake was the most potent. The localized data on the entire cluster of aftershocks from the local source seems more relevant than any single data point about the first shock. For the general readership of the encyclopedia, even a precise location of the epicentre as the main coordinates of the article can be dubious if it turns out to differ significantly from the general area of the earthquake impact, as an article whose general notability is based on affecting not some mountain but a city should be pointing readers to the latter. --Joy &#91;shallot&#93; (talk) 14:09, 3 May 2020 (UTC)
 * That would be OR on our part, to judge which fits best and the distribution of aftershocks is often not a good guide to the epicentral location. Ground conditions also have a significant control on observed intensity. However, as I said just above, I would be happy to see the epicentral coordinates from the Croatian Seismological Service used, as long as we have a source for them. I have searched for them myself, but so far to no avail. Mikenorton (talk) 15:46, 3 May 2020 (UTC)


 * Yes, I don't think we need to judge which epicenteral reading is best, which is why I'm bringing this up, to avoid WP:OR and instead reach a basic editorial discretion in pointing people where to look on a map. If it's hard to get any source to pinpoint something like this, until more scholarly secondary sources about the event are written, we should use what we have - and choosing the literal middle of the map at https://www.pmf.unizg.hr/_news/icons/5dc9b9633453a32061e693765dad41621526_icon.jpg for the article coord seems more sensible than using the epicentre coord. --Joy &#91;shallot&#93; (talk) 16:47, 8 May 2020 (UTC)


 * We don't know the uncertainty in the epicentre location as given by the Croatian Seismological Service (CSS) (we know the ANSS estimate of this it's 3.7 km) or whether the CSS deployed a temporary dense seismograph network immediately following the earthquake (something that is often done), which could mean that the aftershocks are better located than the mainshock, but that's just speculation. I don't think that we can just take a stab at it from a map - that would not constitute a reliable source that we can cite in support. We may end up having to use the map that you linked to in the second part of this thread, but let's wait and see. Mikenorton (talk) 19:55, 8 May 2020 (UTC)


 * Again, that's besides the point. The coordinates of the article do not have to correspond to any particular sourced reading of an epicentre. There is no actual contradiction between WP:V and pointing the Zagreb earthquake on a map to a location in Zagreb. Heck, we have millions of toponyms with coordinates in Wikipedia that have coordinates based on a rough reading of various maps and gazetteers. --Joy &#91;shallot&#93; (talk) 08:20, 10 May 2020 (UTC)


 * You may have missed my comment above digging up the coordinates on the Croatian Seismological Service's website. Would you support including these coordinates?  Daß &thinsp;  Wölf  06:51, 20 May 2020 (UTC)


 * From the map I get coordinates of 45.861, 16.037, which are not hugely different to the 45.85, 16.03 that Das Wölf gave a source for a few sections above. In the absence of anything any more definitive I'm going to change the infobox coordinates to those. It's not a perfect solution, but at least it's based on a solid source. Mikenorton (talk) 10:28, 20 May 2020 (UTC)


 * Thanks guys. This does seem like a reasonable solution for the time being. --Joy &#91;shallot&#93; (talk) 20:39, 20 May 2020 (UTC)