Wikipedia:Reference desk/Archives/Science/2015 May 25

Science desk
< May 24	<< Apr | May | Jun >>	May 26 >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

May 25

Radon and illness

In reading through the article on Radon here, there is a section discussing the effects of Radon within the mining community. I am wondering about health problems coming from Radon in residential areas. More specifically, my parents (who live on the Canadian Shield in Central Ontario Canada)lived in a house for over 30 years. This house used a large portion of natural rock (granite, I assume) for the foundation. So, in the basement (where my father had a small workshop) was exposed rock. My question is what effects, if any, would there be re: radon (or anything else)? Are there any health hazards coming from the exposed rock (no moisture so no mold, or the like)? If so, would painting the rock have made any difference? 50.101.125.48 (talk) 03:01, 25 May 2015 (UTC)

We have some relevant information at Radon#Accumulation_in_houses and Health_effects_of_radon#Health_risks. SemanticMantis (talk) 03:22, 25 May 2015 (UTC)

Great! Thanks. 50.101.125.48 (talk) 04:12, 25 May 2015 (UTC)

As you work through your concerns please consider the following:- What, if any, radon was and is present in the surrounding soil? The granite is, mostly likely, denser than that soil itself and therefore is itself a barrier to radon from that source.

There is an obvious similarity between a basement and a mine in that both are below ground. Radon is a heavy gas and so gravitates to lower levels. I suggest however that greater importance as a source of radon be given to the materials and activities that have occurred & are occurring above the basement, in and about the house, than to the (undisturbed) granite walls of the basement. Consideration also to the ventilation of the basement.

The article observes where radon is found. It does not discuss how fast radon gas that is trapped within rock, where that rock is left totally undisturbed, how long that radon gas will take to migrate (move) to the surface of that rock so that it can then be released into the air. Clearly what is on the surface of the rock is in contact with the air, and may be released. Radon trapped within the rock and in a dense rock such as granite may not migrate at all, therefore it may never be released. For example, the article observes that radon is present in lower concentrations in limestone. Not surprising as limestone is more porous than granite. The important difference between the basement and a mine however it that in a mine the rock is not undisturbed. In a mine rock is being dug, drilled, blasted, fractured, pulverised, moved about, etc. during which trapped gas is released. The same activity is not happening the the walls of a typical basement.

As to painting the walls. Not all paints are effective sealers. Their porousity differs. Then comes the trade off between the chemicals in the paint and radon. Would you feel comfortable if you had used a lead based paint? Lanyon (talk) 05:58, 25 May 2015 (UTC)

The article on radon discusses means of testing and of trying to lower the levels. I think it should be emphasized that radon is an exhalation of the entire earth beneath your feet, or at least so much of it as gas can escape from in a few days - the amount coming from a few little pieces of exposed granite is going to be pretty miniscule; the question is whether fissures exist that channel gas from below, which can occur even if a house is built on an innocent-looking layer of sediment. Note also what the article quotes about the issues of trying to seal out radon without leaving some other means of escape; it's not all that reliable. Really though, this is a professional question requiring personalized advice, and you shouldn't rely on us to guess whether you need a radon test or what method of radon-proofing will be most effective in your specific setting. Wnt (talk) 11:05, 25 May 2015 (UTC)

It is spelt radon.
Sleigh (talk) 11:31, 25 May 2015 (UTC)

There is no doubt that radon is not a good thing to be breathing. It's a heavier-than-air gas so it tends to pool in basements (especially in areas with granite deposits). The amount floating around in the air normally isn't an issue, but a growing concentration of the stuff in a basement area is sufficiently worrying as to warrant taking some action. I believe that the usual remedy is to have an air circulation system in place that pulls air from the lowest point in the basement and pushes it outside the building. That prevents a buildup of radon in the building - which is really the best you can do. SteveBaker (talk) 14:31, 25 May 2015 (UTC)

@SteveBaker: The article actually explains more than this - several methods involve capturing the radon before it enters the building at all, or immediately afterward while it is still trapped beneath plastic. The article doesn't make it sound like it stays in a separate layer for long, though I imagine there would be some temperature stratification affects in some specific cases. Wnt (talk) 15:48, 25 May 2015 (UTC)

Some resources relevant to radon in Canadian homes - Lung Association Ontario, who say that radon is the second leading cause of lung-cancer in Canada, CBC News information, which says 6.9% of Canadians are living in homes with radon levels above Canadian guidelines, and Health Canada Cross-Canada Survey of Radon Concentrations in Homes - Final Report. The Lung Association page has links to further information, and to buy a test kit. DuncanHill (talk) 23:07, 25 May 2015 (UTC)

And Public Health Ontario have a page with information and further links. DuncanHill (talk) 23:10, 25 May 2015 (UTC)

Geology vs physical geography, is geology a sub-field of physical geography?

What is the difference between geology and physical geography? Geology is the study of rocks, while physical geography encompasses atmospheric sciences, hydrology. So could we say that geology is a sub-field of physical geography? --IEditEncyclopedia (talk) 11:41, 25 May 2015 (UTC)

Physical geography focuses on the large scale. So for example the crystal structure of quartz would be a topic in geology but wouldn't ordinarily be thought of as a topic in physical geography. Looie496 (talk) 12:23, 25 May 2015 (UTC)

I think that it's better to say that there is a small degree of overlap between physical geography and geology. Geology covers a much wider field of study than physical geography. (Note as a geologist, I am naturally biased in my view) Mikenorton (talk) 14:05, 25 May 2015 (UTC)

There is some overlap, but the best way I could explain it is that geology is the "what happens" and physical geography is the "where it happens". Physical geography is all about mapping geologic processes and their evidence to specific places on earth. Geology is more about describing and explaining those processes themselves. Though, of course, each feeds the other. --Jayron32 01:33, 27 May 2015 (UTC)

By what mechanism do some diseases cause loss of appetite?

Clearly several categories of disease cause a temporary loss of appetite (influenza, for example).

• I wonder why (and how) they do that.
• Would it be possible to isolate the mechanism of that symptom into a super-effective weight-loss treatment without ungodly side-effects?
• Is anyone working on such a thing? (Seems like they should be - because something that would turn someone off of eating on demand would be a far better solution than things like baryatric surgery...and sales of such a treatment would be off the charts).

Anorexia (symptom) (not anorexia nervosa) describes some drugs that produce this effect - but it doesn't seem like any of them are safe or without horrendous side-effects. SteveBaker (talk) 14:23, 25 May 2015 (UTC)

I believe it's an attempt by the body to clear the digestive system out, when it may contain potentially harmful microbes. Vomiting and diarrhea also serve to clear it out. See appetite suppressant for drugs. One risk of such drugs, besides the side-effects, is malnutrition, since, if the person had an unhealthy diet, but managed to get enough vital nutrients due to quantity, then just lowering the quantity without improving the quality might lead to a lack of those nutrients. StuRat (talk) 15:22, 25 May 2015 (UTC)

Complete WHAG here, but is it the body's method of diverting resources away from digesting food (which takes a lot of energy) and into attacking invaders? --TammyMoet (talk) 16:00, 25 May 2015 (UTC)

I know WAG = Wild-Assed Guess, but what's the H ? StuRat (talk) 03:42, 26 May 2015 (UTC)

Wild Half-Assed Guess (technical distinction?).--Shantavira|^{feed me} 15:37, 26 May 2015 (UTC)

• Good questions! Your first question, ie the question of mechanism is definitely being studied. For example here is a request for proposals from NIH from 20 years back. If one would trace who received the grants, and their subsequent publications that would be useful. Haven't done so yet, but here are a couple of review papers on the proposed mechanism for anorexia in hemodialysis patients and in cancer patients, which suggest that though there are commonalities there is unlikely to be one universal mechanism. Side note We should be wary of just so explanations of what the body is "trying" to do. For example, one could have argued that since the body needs energy to fight invaders and digestion is a net energy gain, ones appetite increases when one is ill. Of course this "explanation" is contradicted by observed facts, but on its face is as plausible as any similar argument. Abecedare (talk) 16:46, 25 May 2015 (UTC)

• See also cachexia which mentions Tumor Necrosis Factor and Interferon. The later was viewed as a wonder drug until the fact it caused cachexia was understood. μηδείς (talk) 23:12, 25 May 2015 (UTC)

Cutting patterns out of fabric: humans against computers

Apparently, given a roll of fabric for producing garments, humans can find the best arrangement of patterns to cut the most pieces out of it, resulting in less wasted fabric. Obviously, the patterns are not regular geometrical figures.

Why can humans, if this is indeed the case, perform better than a computer?--Yppieyei (talk) 17:24, 25 May 2015 (UTC)

Who says they can? ←Baseball Bugs ^{What's up, Doc?} carrots→ 19:39, 25 May 2015 (UTC)

I too question the premise of the OP's question, at least if they are considering state-of-the-art algorithms. There is in fact a rich literature on th subject known as irregular shape packing problem in mathematics and the nesting problem in the textile industry. See this paper for a review. Abecedare (talk) 20:08, 25 May 2015 (UTC)

I think a computer could produce a better result just by exhaustively trying a gazillion possibilities. However, the amount of time it might take would be pretty big...but in some applications you can afford to wait. It is remarkable how good humans are at doing this - particularly if they have a lot of practice in doing it.

In case you think the combinatorial problem is too large - for the case of cloth, it's common for the major parts to have to be aligned in specific ways to the warp and weft of the cloth so that the garment falls properly. Sometimes there is a demand that the pattern on the cloth is aligned in some specific way too. That actually simplifies the problem spectacularly for the computer because there are just less options to try and a 'brute force' search becomes more plausible.

Where the computer can really do well is when there is already a reasonable solution - at that point, the computer can use Gradient descent methods to improve the solution. At that point, I'd be surprised if a human could outperform the computer in that final stage.

But it's undoubtedly an exceedingly difficult problem for a computer - and humans seem to just 'get it' in a ridiculously tiny amount of time.

This is actually a problem I have to deal with on a regular basis - my wife makes model buildings from plywood using a laser cutter - essentially the same problem. Very often I have to pack a dozen parts into the smallest amount of wood. What is utterly amazing is that I can work for 10 or 15 minutes on the problem - thinking that I have a good result - but then come back a week later and immediately spot something that makes a massive improvement. It's clearly not a conscious process for us. SteveBaker (talk) 22:44, 25 May 2015 (UTC)

I know these days it appears the question is being flippantly dismissed, but the premise of this question was almost certainly true at some point in time, how long ago it became "not true" arguable, but at least for a while computers were excellent at "brute force" calculations, but quite poor at Pattern recognition, easily out performed by even a novice brain at simple tasks like, picking out a face in a photo. I agree there are probably few tasks like this left that a human would perform "better" than a computer. (although captcha still works under the assumption a human will solve it MUCH easier than a computer!) This is still a very active area of research in Computer Science and there has been great progress especially since the age of the internet. Vespine (talk) 06:07, 26 May 2015 (UTC)

As I understand it, one issue is that these sorts of irregular shape packing problems - especially when you start adding in constraints like aligning the bias in a certain direction - start to reach into the NP-hard/NP-complete categories. That is, it's easy to verify that you have a better solution, but it's difficult to come up with that better solutions. The classic example in this category is the traveling salesman problem, which is finding the shortest path which visits all the cities in a list, or the boolean satisfiability problems, where you attempt to find a set of inputs which satisfies a set of logical conditions. These are hard to solve because the naive implementation of using a brute force method only works in the most simple cases. For other types of problems (e.g. polynomial time problems) the problem has a certain internal structure which makes solving it simpler. For NP-hard/NP-complete type problems, there really isn't any of that internal structure to simplify the problem - or there is, but it doesn't bring down the complexity enough to be solvable. Traditionally, humans have done better at these sorts of problems because they tend to use heuristic approaches rather than algorithmic ones. These are "short-cuts" that are used in problem solving which help simplify the problem at the approach of not being able to guarantee an optimal solution. Early computer development focused on algorithmic approaches because a computer is intrinsically an algorithmic machine. It's taken a lot of effort over the years to figure out how best to encode decent heuristic approaches as an algorithm that computers can follow. This is complicated by the fact that humans are bad at describing the heuristics they're using (e.g. even if SteveBaker was able to find a much better arrangement, he probably couldn't verbalize *how* he could look at the problem and realize it might give a better result). Computers are improving at this though, so the original claim may not strictly be the case anymore. However, part of this comes from the increasing speed of computers - they're trying many more combinations much quicker than humans do. We still don't know all the heuristic tricks expert humans use in solving these sorts of problems, so if you limited humans and computers to the same number of trials (rather than the same amount of time) the expert human would still likely beat out the computer. (Disclaimer - I don't know for certain that the cloth fitting problem is in either NP-hard or NP-complete - I'm just saying it sounds much like problems that are.) -- 160.129.138.186 (talk) 16:25, 26 May 2015 (UTC)

The reason that humans perform well in these tasks is the same reason that humans see faces in mundane objects and patterns in the stars. Humans see what is not there. So, given a blank cloth, humans see patterns that show how to fit things together. Computers see what is there. They have to meticulously move things around, constantly trying to improve. That leads to a local minima problem, which leads to algorithms to break the local minima problem, which lead to complicated solutions that take forever to run. So, the obvious solution is to program a computer to see what is not there in the same way that humans do. There is a lot of work in that field (such as facial recognition), but it isn't at the standard of a normal human brain yet. It will be soon. At this point in time, the algorithm designs are still ahead of the available computing power. When the computing power exceeds the algorithm designs, we will be able to do a lot more. 209.149.115.52 (talk) 17:34, 26 May 2015 (UTC)

• A few additional notes: As 160.129 suspects (common flavors of) the irregular packing problems are indeed NP-hard/NP-complete, which has been know for over 3 decades. And they are combinatorial optimization problems, so gradient descent type methods are not very useful (except sometimes in a inner sub-routine of a broader algorithm). As 209.149 and others have suggested efficient algorithms in the area all incorporate heuristics to help with the search, and 100s of papers have been published on the subject (sample Google scholar search results to get an idea of the approaches). For an interesting alternate application of this research see this book on optimization of cargo-loading for space-flights. Abecedare (talk) 18:29, 26 May 2015 (UTC)