Wikipedia:Reference desk/Archives/Science/2019 December 11

= December 11 =

River tributaries by discharge volume
I'm hoping to diagram a few river basins around the world. Ideally I'd like to include any tributary that meets a certain threshold—say 5% of the basin's total discharge. I'm currently using the Murray River as a trial. The article on the Murray–Darling Basin is helpful in understanding the sequence of the tributaries and their connections but does not contain discharge volumes. I've been looking for volumes just by tracing branches and opening lots of browser tabs. Some individual articles include discharge volumes in the infobox, but many others (most?) do not. One may reasonably suppose that larger rivers are likely to have more complete information available, but I suspect that assumption fails from time to time. Presently, I've traced 426 m³ of the basin's 767 m³ to tributaries discharging more than 30 m³, suggesting that the other 341 m³ is contributed by feeders below the threshold, but I don't have a sense if that's a good assumption, and I don't have an obvious way to check my work in case I missed a branch somewhere. Does anyone know of source that would make this kind of information-gathering easier? Thanks! —jameslucas ▄▄▄ ▄ ▄▄▄ ▄▄▄ ▄  22:39, 11 December 2019 (UTC)
 * (P.S. I'd be open to using basin area instead of discharge volume if that made an appreciable difference in data-gathering ease. —jameslucas ▄▄▄ ▄ ▄▄▄ ▄▄▄ ▄  23:28, 11 December 2019 (UTC))


 * Discharge varies over time, so unless every value you use for every one of the tributaries in the basin is some kind of average measure, or if they are all instantaneous values that were coincidentally collected at the exact same instant, there is no good reason to assume that the numbers should add up nicely. On top of this, stream- and river- discharge constitutes surface water flow - but over a large basin, an enormous percentage of the water volume is not surface-water.  In many parts of the world, most water flow is underground.  So when you look at stream gauge or discharge data, you should treat the values as a sort of reference-number that gives you an idea about the stream; but if you want to really do quantitative analytical study of the amount of water, you need a much more sophisticated methodology.  The key takeaway, though, is pretty simple: water enters the rivers via tributaries, but water can also enter and exit the river in the vertical direction at every single portion of the basin, not only at the confluence of the surface streams.  New water gets added or subtracted upwards and downwards - evaporation, precipitation, and groundwater charge- and discharge- into and out of the subsurface.
 * Also be sure to read hydrology and hydrogeology and groundwater flow.
 * Here's a pretty decent book cited by several of our articles: Groundwater Hydrology: Conceptual and Computational Models.
 * Nimur (talk) 23:47, 11 December 2019 (UTC)


 * Thanks, ! The notion of water entering and leaving from underground makes sense but I hadn't considered it. Ultimately, my project is more artistic/symbolic than scientific, so I probably won't be delving into that level of detail until I have a bit more time to spare, but I'm definitely appreciative of the resources you recommended. For what it's worth, I was looking at average discharge and wasn't concerned about great precision. Since I asked my question, though, I started looking at the Yangtze, and I realized that English resources (here or elsewhere) for Chinese rivers are so underdeveloped that I can't see being able to take this strategy global anytime soon. I'll see if I can make it work with basin areas instead. Cheers! —jameslucas ▄▄▄ ▄ ▄▄▄ ▄▄▄ ▄  01:08, 12 December 2019 (UTC)


 * I'm noting something I can't make sense of: The Lachlan River is a tributary of the Murrumbidgee River, but according to their articles' infoboxes, the rivers' basin sizes are nearly identical (84,700 km² and 84,917 km², respectively). Am I correct in thinking that this makes no sense since the basin of any river should include the basin of its tributaries and the basin of a tributary can't realistically be 400× larger than the river it feeds into? —jameslucas ▄▄▄ ▄ ▄▄▄ ▄▄▄ ▄  02:36, 12 December 2019 (UTC)
 * Yes, those numbers seem suspicious. I also note that neither article cites a source for those data.  It is possible that the numbers are somehow being misinterpreted from a valid authentic original source, but they are being repeated on Wikipedia without attribution so it's very hard to follow up or verify.  It is also possible that those numbers are completely fictional - in other words, they may be the result of intentional vandalism to add incorrect numbers to our articles.  Without a cited source, it's really hard to say with certainty.
 * This type of data-retrieval question exactly the sort of thing that a Wikipedia article is very bad at. It's difficult for our editors to verify who added the numerical values, and whether those values are accurate and correct.
 * If you want to find such numerical data, you almost surely want to find a reliable peer-reviewed source that is not authored and curated by a community of anonymous volunteer editors (... like us). Our Wikipedia articles can give you a good introduction that is largely correct - and our editors can and will remove "obviously-wrong" information - but when it comes to quoting a specific numeric value, the process of article-editing that we use around here is just not well-suited to keeping such data consistent, correct, and current.  We manage to make Wikipedia function very well as an encyclopedia - but we do not have the capacity to make Wikipedia into a very good data-almanac.  In fact, it's a terrible very very bad data-almanac that frequently contains wrong, stale, or maliciously-vandalized numerical data.
 * The good news is, there are lots of better alternatives!
 * For example, Geoscience Australia (a department of the Australian national government) publishes a catalog of data products. They also publish national-scale maps of river basins and catchments; and if you use their search-engine, you can find loads of more detailed specific information about specific basins and catchment areas.  They also have a downloadable GeoFabric database from which those maps were produced.  An enthusiastic and skilled individual can read the how-to tutorials to do things like calculate the basin drainage area using the database - but it is going to require a little bit of effort (and a basic familiarity with industry--standard tools, including ArcGIS software).
 * Overall, that data might be a little "less-conveniently-packaged" than our Wikipedia articles, but it almost surely is of much higher quality: it is curated by experts and it is more likely to be consistent, correct, and current.
 * Waxing philosophically - the reason that those sources are more reliable is not simply because the numbers were provided by experts; nor because the numbers are provided by a government agency. Those sources are more reliable because every line-item of data is accompanied by detailed thorough documentation that expresses where the numbers came from, and how to interpret them, and what pitfalls you need to be aware of when you use those numbers.  There is, so to speak, a deeper chain of accountability that explains the provenance of each element of the data.  Without the context provided by the documentation, those data would be otherwise-arbitrary numerical values of dubious merit.
 * Real scientists don't just copy-paste their numbers from a convenient list - they always check the work! Now, because you're seeking the numbers for "artistic and aesthetic" purposes, we might be able to find some kind of summary that somebody else has already done - like the map and list of data tables that I linked to earlier - but I'm not sure it's detailed enough for what you wanted.
 * Nimur (talk) 04:23, 12 December 2019 (UTC)


 * A look at this article highlights the difficulties of describing flows in anything like fixed numbers. The Darling ran dry this year. HiLo48 (talk) 23:05, 14 December 2019 (UTC)


 * Thank you both. The disappointing conclusion is that—within my current time constraints–I may need to pick rivers based on availability of data instead of based on other criteria. I thought that if I was willing to accept a low-ish bar for consistency and accuracy, I'd be able to take advantage of Wikipedia's global scope and relatively consistent formatting (as opposed to getting more accurate but less systematic data from many different governmental sources), but it seems like even a low bar is too high in most cases. The Yangtze, which had been high on my list is almost certainly going to have to wait until this becomes a more serious endeavor. I appreciate the feedback and advice.. —jameslucas ▄▄▄ ▄ ▄▄▄ ▄▄▄ ▄  18:33, 17 December 2019 (UTC)


 * Following up on this, I've discovered that Wikidata provides numeric values for watershed areas and/or average discharge for a much greater number of rivers than does Wikipedia. Verifying this data is comparatively difficult, but for my (mostly artistic) purposes, it's been an unparalleled resource. —jameslucas ▄▄▄ ▄ ▄▄▄ ▄▄▄ ▄  02:21, 23 January 2020 (UTC)