User talk:Mikenorton/sandbox

Hi, Mike. Yes, looks good. Some suggestions?

• Abbreviating 'kilometres' to 'km' and dropping the parentheses reduces clutter, and avoids the annoying line-break. Something like '30 miles / 48 km' seems quite reasonable.
• Strong suggestion: link to the USGS Quaternary Fault and Fold Database. Both text (most authoritive source available) and location map.
• Replace 'Orientation' with 'Strike'. It is the precise term, and though generally unfamiliar a note (bottom of the table?) quickly fixes that.

Would you like me to throw in some QFFDB links as examples? ~ J. Johnson (JJ) (talk) 21:58, 30 April 2012 (UTC)

I tried out some of the changes that you suggested, although I have some comments

• The QFFDB link is a good idea - actually we really need an article on that - but how do we handle the case of the San Andreas with 10 fault sections, each with their own page on the database, but no overview page for the fault as a whole?

• Orientation was intended to cover both 'dip and strike' - most of them are near vertical, but some are not.

I'm open to any other suggestions - this was just a quick first pass. One of the things that I do for a living is produce fault data sheets in spreadsheet format (well at least, whenever I get a big enough project to make this approach worthwhile, which is every few years or so), including other things like age of activity, and evidence for presence and displacement. We're obviously more space limited here. Mikenorton (talk) 16:29, 9 May 2012 (UTC)

Wow, you're already on top of it! Well, as to multiple links, I think listing all (?) looks workable. 10 links/segments is a bit heavy, but workable. I would put in the fault/segment numbers. And at the top of each entry have a link to the proper map. Also relabel the column "QFFDB", as that is the source.

I think "Strike/dip" would work. Also, could "miles/km" be put under the label for length, and omitted from each row?

I could throw those in for a try out, but don't want to stomp around unless it's okay with you.

Fault data sheets, eh? Sounds interesting. Do you do hypocenter relocations? At some point I want to collect some well-located hypocenters for a swarm of quakes and try a 3D visualization. (When I get a "round toit"?) ~ J. Johnson (JJ) (talk) 21:52, 9 May 2012 (UTC)

As to the sources column I would like to keep the option of sources that are not from the QFFDB, just in case there's any newer data published not included there.

Strike/dip would work fine. Being a Brit of course I'd like to just stick to km, but I felt that someone would be sure to add the miles - I'll try out your suggestion.

Oh and feel free to make your own changes - I've no problem with that.

As to hypocentres, that's really not my thing, I do faults because that's what I'm interested in and people pay me to look at them because they chop up the oil and gas fields and control deposition of the reservoirs and source rocks. My only involvement with earthquakes professionally has been in seismic hazard work for nuclear installations and in using earthquake distributions and focal mechanisms to understand the seismotectonics of an area. Mikenorton (talk) 22:42, 9 May 2012 (UTC)

For faults in the U.S. the QFFDB is the authoritative source, summarizing all other sources. The problem is that it isn't kept up-to-date, so some new information — or worse, some "newer" faults — might not be included. But for a summary listing of faults it is probably good enough, and any notable new information can be included in the pertinent article. Assuming that the fault is notable enough to even have an article. ~ J. Johnson (JJ) (talk) 23:02, 9 May 2012 (UTC)

Tried several tweaks. And had a thought: listing the map names gives some idea of where in the state the faults lie. Well, see how you like it. ~ J. Johnson (JJ) (talk) 23:57, 9 May 2012 (UTC)

When referring to other sources I knew there was something specific - I was thinking of this as a source for future seismic hazard - updated in 2008, but these can be added in the relevant column. The maps are a good idea. I think that we're close to getting a model for how the list could look - after that there's just the construction to do :). Mikenorton (talk) 18:48, 10 May 2012 (UTC)

Actually that's the wrong source - I used that for my fault parameters, the link should be to Page 74 of the full report, Tables 12 & 13, which list the 30-year probability of M ≥ 6.7 events for each fault - of course if all the hazard information comes from those tables we can just add a single reference to the column header. Mikenorton (talk) 19:13, 10 May 2012 (UTC)

I'm off tomorrow for a week's holiday, so will not be active here - feel free to keep on playing! Mikenorton (talk) 22:50, 11 May 2012 (UTC)

Actually that's the wrong source - I used that for my fault parameters, the link should be to Page 74 of the full report, Tables 12 & 13, which list the 30-year probability of M ≥ 6.7 events for each fault - of course if all the hazard information comes from those tables we can just add a single reference to the column header. Mikenorton (talk) 19:13, 10 May 2012 (UTC)

Have fun! A couple of minor comments.

Not "wrong", just not the latest version. I suspect there is an original source at USGS or SCEC, which we should link to.

I had a further thought regarding any notable updates not in the QFFDB — put a link in the comments field. Though I am still inclined to think the article would be the more appropriate place.

I may add some more entries, just get a better feel for the process. And to look for any anomalous cases.

~ J. Johnson (JJ) (talk) 22:57, 11 May 2012 (UTC)

More thoughts for consideration.

1. I am starting to think the QFFDB fault numbers/links (and maps) are important enough to be at the front. (And the map names give some indication of where the segments are.) This would be especially needed if any of the other columns should present per-segment info.
2. Followed by "Notable earthquakes"? After the name of the fault I suspect this is what most readers would find most notable/interesting. And these probably would align "per-segment".
3. The bare numbers under "Extent" look a bit strange. Perhaps they should have "miles" and "km". But not kilometers or the parentheses.
4. The more I think about "seismic hazard" the more I think this needs some careful, deep consideration. I know that "presentation of seismic hazard" (implicitly: to the public) has been a big issue with geologists for a while. I think the documentation discusses this (and there is a another document about it elsewhere), and I think we should give these careful heed lest we do a disservice.
5. More particularly, it appears there are different kinds and even different estimates of seismic hazard. I don't know if any single one is significantly superior to the others, so I could see having multiple estimates. How to present them might be a challenge, as there is usually a bunch of caveats or extended description (e.g., "probability of a M≥6.7 event by 2031"). These could be handled at the top of the table. The actual estimates are presented "per section", which is not quite the same as the QFFDB ("Q-faults") segments, but perhaps close enough. So rather than multiple lines in one column (like "Extent") it might be necessary to have multiple columns. Well, that certainly needs more thought.

~ J. Johnson (JJ) (talk) 22:57, 14 May 2012 (UTC)

Big doings! I've prepared a reformatted and expanded version here. (Copy it as you see fit.) The expansion covers the big seven California class A faults. I split the San Andreas fault (like the WGCEP did), and it looks fine. I don't know how many of the class B faults anyone would want to cover, but I would see those in a separate table. Still some bits to fill in. And the last two items I raised above still need consideration.

The QFFDB links are a piece of cake: I collect and tabulate four items for each segment, tack them onto a Perl script I wrote, and it generates the formatted links. Once the links are in it's just a click and a scroll to get the other data. The maps need to be checked: sometimes the report doesn't list all of them. Also: I just had a wild idea, and put the maps in their own column, and removed Comments. The table is a lot more compact now, and I think looking pretty good.

Well, I've been having fun. Hope you had at least as much fun on holiday.

~ J. Johnson (JJ) (talk) 22:38, 18 May 2012 (UTC)

More and more perfect. Probably pointless to ask for suggestions, eh?  :-) Looking ahead: I am envisioning an article (California faults) containing this table, followed by a similar table for the class B faults (some hundred or two them). These two tables, in addition to being a list of the California faults, could in effect be an index to such as have Wikipedia articles. I have also thought of including an anchor for each entry so that the articles could link straight to the table, but I am not (yet?) convinced of the usefulness of that. ~ J. Johnson (JJ) (talk) 21:41, 19 May 2012 (UTC)

Break

Decided to make a break in this, so that it's easier to navigate.

I took a look at your new version and I think it looks good, with an expansion to include the B category faults ( and maybe the Cs), I think that this is ready to be inflicted on an unsuspecting world. Mikenorton (talk) 19:22, 24 May 2012 (UTC)

Well, close to ready. I was looking at the QFFDB and was reminded that there are a lot more class A faults (a hundred or so?), so really need review (and reflect in the text) why UCERF included only these. I wouldn't include the C faults ("Geologic evidence is insufficient..."). As to the B faults: there's a gosh darn lot of them. I don't think we can, or should, include all of them. We should have some criterion, however rough or loose, as to which to include.

So would California faults be a good place to stuff these tables? ~ J. Johnson (JJ) (talk) 22:56, 24 May 2012 (UTC)

Oooh, not as ready as I thought. I got to wondering why UCERF listed only six "A" faults, and soon discovered that UCERF's Type A faults are not the QFFDB's class A faults. UCERF is categorizing sources on whether they have enough information to model them for a forecast (and the "Type C sources" are zones without specific fault attribution). My question/comments just above were, well, ill-founded. And there are only 20 type B faults (Table 13, p. 33), so actually quite manageable.

So this fault table is not the one I originally envisioned, of a listing of the major California faults, but of faults for which there is an earthquake rupture forecast. Which is fine, but I need to retool some of the text (working on it), and even the article where this will go; California faults no longer seems appropriate. Two possibilities: California earthquake probabilities, and California earthquake forecast. What do you think? ~ J. Johnson (JJ) (talk) 23:45, 5 June 2012 (UTC)

Hmm, this has the potential to generate a very wordy title. The report says "ruptures on known active faults (Type-A and Type-B sources)", which I take to mean that A & B combined will list all known seismically active faults in California. Seismically active faults of California is therefore a possibility. Mikenorton (talk) 20:21, 6 June 2012 (UTC)

No! That's the same mistake I was making. The QFFDB class A and class B faults pretty much cover, as you said, "all known seismically active faults in California." But! UCERF's type A and type B categorization refers to faults with sufficient data to prepare a rupture probability. Which amounts to only six "A" faults about 20 "B" faults. I think you just missed the changes I just put in (see User:J. Johnson/Sandbox3) which explains this. Take another look. BTW, I am starting to warm up to California earthquake forecast. How does that sound to you? ~ J. Johnson (JJ) (talk) 22:55, 6 June 2012 (UTC)

Added an image, and I think this sucker is ready to fly. The "Type B table" (hmm, that sounds better than what I have) isn't ready, but that can follow later. I haven't wikilinked to Earthquake prediction, because that article looks to be seriously off-the-rails. (Another Project!!!). ~ J. Johnson (JJ) (talk) 22:12, 11 June 2012 (UTC)

Sorry to take so long yet again to reply, been really busy the last few days. It looks ready enough to me, so go for it. Earthquake prediction isn't really what the 'forecast' is about, an earthquake forecasting article (that's currently a redirect to the prediction article) might be good idea, but that's for another day. Mikenorton (talk) 20:11, 13 June 2012 (UTC)

Yeah, can't do everything. Okay, I'll sleep on it tonight, and if I still respect it in the morning I'll throw it in. ~ J. Johnson (JJ) (talk) 22:05, 13 June 2012 (UTC)

Just to show that I can reply in less than two days - sounds good. Mikenorton (talk) 22:07, 13 June 2012 (UTC)

Done! (And the link turns blue.) Or close enough. I haven't added links in all the various faults, but that is the kind of detail that other editors can handle. ~ J. Johnson (JJ) (talk) 20:27, 14 June 2012 (UTC)