Wikipedia:Reference desk/Archives/Science/2015 July 12
Science desk | ||
---|---|---|
< July 11 | << Jun | July | Aug >> | July 13 > |
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. |
July 12
editWhy an even 100 mg/dL as blood glucose standard?
editIt seems rather a large coincidence that the standard American expected blood sugar level should center on 100 mg/dL. Is this just a rounded number? An average? A statistical artifact (i.e., sig. fig.s?) Can anyone point out where this number originated. (To stave of certain unnecessary clarifications, I am well aware there are various ranges, such as 70-100 for fasting non-diabetics, 90-130 for controlled diabetics, etc. I am only interested in the origin of the 100 mg/dL, and don't need the other mechanics explained.) Thanks. μηδείς (talk) 01:00, 12 July 2015 (UTC)
- I tried really hard to find this, but failed so far. The various medical associations that state a standard blood glucose concentration don't make it obvious where they get that advice. In the literature, I have yet to find a paper or review that established rather than merely mentioned a blood glucose standard. Papers stating a blood glucose standard go back to 1911 at the very least, so I'd imagine the paper(s) that established the 100mg/dL or 70-130mg/dL standard may be very very old. If there is a more recent work to reinforce that this should be standard, I haven't found it. Someguy1221 (talk) 02:20, 12 July 2015 (UTC)
- It is based on observed fasting glucose levels of nondiabetics and is a coincidence. Edison (talk) 13:52, 12 July 2015 (UTC)
- I tend to suspect that's true, but do you have a source, or can you at least say "as I remember from medical school"? I am pretty good at speculation on my own, so I was hoping for something at least historically that explains this number. μηδείς (talk) 17:12, 12 July 2015 (UTC)
- This paper looking at over half a million patients at least reinforces the claim that 70-100mg/dL is relatively low-risk. Individual studies vary in which specific cutoffs they conclude with, and that makes sense due to different study populations and different risks being evaluated. What I do notice, Medeis, and what I suspect belies the round numbers, is binning. All of the papers I've been finding looking at large numbers of patients invariably chose to bin their data before doing any statistical analysis. The choice of binning is of course arbitrary, and the authors always choose round-number cutoffs. Some groups produce non-round values of mg/dL, but only because they were using round molar values. What is consistent between almost all publications is that 100mg/dL is always inside the lowest-risk bin. The rare exception seems to be studies concerning the elderly, some of which have 90mg/dL as the center of the safest bin, and 100mg/dL in an elevated risk bin. But anyway, yeah, BINNING. That is your answer (I suspect, anyway. I still haven't found the study(s) explicitly stated as having established the standard). Someguy1221 (talk) 23:59, 13 July 2015 (UTC)
- I saw my endocrinologist today and she is newly finished interning, so not too many years out of school. She gave me a blank stare, until I finally compared the number to 98.6F as the "normal" body temperature. Then she said that there are a choice of ranges, and that 100 was a number based on averages that was close enough to normal that it is used for convenience. She didn't say data binning, but I understand the concept. Thanks for the assistance. μηδείς (talk) 00:40, 14 July 2015 (UTC)
- This paper looking at over half a million patients at least reinforces the claim that 70-100mg/dL is relatively low-risk. Individual studies vary in which specific cutoffs they conclude with, and that makes sense due to different study populations and different risks being evaluated. What I do notice, Medeis, and what I suspect belies the round numbers, is binning. All of the papers I've been finding looking at large numbers of patients invariably chose to bin their data before doing any statistical analysis. The choice of binning is of course arbitrary, and the authors always choose round-number cutoffs. Some groups produce non-round values of mg/dL, but only because they were using round molar values. What is consistent between almost all publications is that 100mg/dL is always inside the lowest-risk bin. The rare exception seems to be studies concerning the elderly, some of which have 90mg/dL as the center of the safest bin, and 100mg/dL in an elevated risk bin. But anyway, yeah, BINNING. That is your answer (I suspect, anyway. I still haven't found the study(s) explicitly stated as having established the standard). Someguy1221 (talk) 23:59, 13 July 2015 (UTC)
- I tend to suspect that's true, but do you have a source, or can you at least say "as I remember from medical school"? I am pretty good at speculation on my own, so I was hoping for something at least historically that explains this number. μηδείς (talk) 17:12, 12 July 2015 (UTC)
- It is based on observed fasting glucose levels of nondiabetics and is a coincidence. Edison (talk) 13:52, 12 July 2015 (UTC)
Ferromagnetism
editWhy are magnetic field lines nearly always normal to the surface of a ferromagnet at every point? Yes, this is a homework question, and yes, I tried to figure it out but I can't. Yashowardhani (talk) 03:12, 12 July 2015 (UTC)
- If there exists any component of the magnetic field that is tangential to the surface, that implies that there also exists a current flowing in the surface (either an actual flow of charge, or a displacement current, appropriately chosen to correspond to the magnetic field). You can verify this using the right hand rule; or, if you are a little bit more mathematically inclined, you can apply Maxwell's equations and see for yourself. This can occur in the real world (it is not "unphysical"); but it is not a typical characteristic of what we call a ferromagnet. It's much more plausible to see that kind of behavior in an active device that is expending energy to cause current to flow (like an antenna driven by an electric circuit).
- Here's a lecture note from MIT's OpenCourseWare on boundary conditions in electromagnetics for your review.
- Nimur (talk) 09:04, 12 July 2015 (UTC)
- Wow thats what you call a model answer. I dont know if there are any prizes for good answers, but if there were i would give you one!! --86.176.9.176 (talk) 21:02, 12 July 2015 (UTC)
- Thanks for the article, although it didn't quite solve my query. From what I've read, ferromagnetic substances get strongly magnetised when placed in an external magnetic field. Magnetic field lines tend to pass "through" them, but that seems to mean that the field lines are only nearly perpendicular to the ends of the ferromagnetic domains, and as the domains are aligned in the same direction, to the ends of the ferromagnet, and not the entire surface. My textbook only says that "Proof of this fact is based on boundary conditions of magnetic fields at the interference two media. When one of the media has μ >> 1, the field lines meet this medium nearly normally." Can you suggest somewhere I can find this proof?
- Sorry for the late reply. Yashowardhani (talk) 15:00, 19 July 2015 (UTC)
- And sorry I sound so confused. I really need to clarify this. I know WP isn't a place for elementary science discussions, but it was my last resort. Yashowardhani (talk) 15:11, 19 July 2015 (UTC)
Difference in flight time between SEA-HKG and YVR-HKG
editGoogling "yvr hkg" gets me "...13h 10m duration", while googling "sea hkg" says "14h 5m duration". That's almost an hour of difference. Why the huge difference? YVR and SEA is only about 200km apart, plus that 200km is (close to) orthogonal to the flight path. My other car is a cadr (talk) 06:00, 12 July 2015 (UTC)
- Looking only at the lowest-cost flights, which appear first, the shorter flights are all serviced by B777s and the longer one by A330s. I conclude that the B777s are flying roughly 7% faster than the A330s, and both are flying at their best fuel economy speed. On the other hand, given that our articles show cruise speeds of 560 mph and 541 mph, not a 7% difference, I could easily be concluding incorrectly and there may be other factors involved that I'm not aware of. ―Mandruss ☎ 06:22, 12 July 2015 (UTC)
- To correct myself, the 200km is actually parallel to the great circle route flight path, so that definitely added some flight time (about 10 to 15 minutes). My other car is a cadr (talk) 06:32, 12 July 2015 (UTC)
- Yes, according to gcmap.com the distances are 10287 and 10460 km, so Seattle is 1.68% farther. If the 777 cruises at 560 mph vs. 541 mph for the A330, that's a 3.51% difference. That leaves only 1.62% unaccounted for, or about 12 minutes. I could see the difference in taxiing time between one and another airport being that large, though I don't know about those specific airports. --174.88.135.232 (talk) 09:45, 12 July 2015 (UTC)
- There also appears to be some rounding going on, to the nearest 5 minutes. That could randomly add or subtract a few minutes. StuRat (talk) 21:15, 12 July 2015 (UTC)
You're looking at the scheduled flight durations, but it turns out that the average true flight durations, comparing YVR-HKG and SEA-HKG are much closer. The website flightstats.com shows ratings of 0 to 5 stars in three categories for each flight: on-time performance, delay performance, and overall performance (which is a combination of the other two). Here's what I found on that web site, in the "overall performance" category:
- Delta 281 SEA-HKG 4.7 stars [1]
- Air Canada 7 YVR-HKG 1 star [2]
- Cathay Pacific 889 YVR-HKG 0.1 stars [3]
- Cathay Pacific 837 YVR-HKG 0.3 stars [4]
The Delta SEA-HKG flight is usually on time, while the YVR-HKG flights are frequently late. I can think of three possible reasons for this:
- (1) There is a systemic problem at YVR airport such that transpacific flights are often late to take off, resulting in late arrivals, compared to SEA airport; or
- (2) The schedulers at Air Canada and Cathay Pacific are, either in general or for all transpacific routes, more optimistic than the schedulers at Delta; or
- (3) For this specific route, the schedulers at Air Canada and Cathay Pacific are more optimistic than those at Delta.
I would conjecture that the true explanation is (3), and there is a simple economic rationale: Delta has the only daily nonstop flight from Seattle to Hong Kong and therefore it has no incentive to be optimistic in the scheduled flying time (i.e. it would not attract any additional business by shaving 10 or 20 minutes off its posted travel time on those flights). By contrast, Air Canada and Cathay Pacific are in competition on the YVR-HKG route, and for some travellers, a difference of 10 or 20 minutes in the posted travel time might be the deciding factor in which airline they choose. Mathew5000 (talk) 08:04, 17 July 2015 (UTC)
Coneflower (Echinacea purpurea) identification
editCan anyone tell me which cultivar of Coneflower (Echinacea purpurea) is the flower in this photo? Thanks --Captain-tucker (talk) 10:10, 12 July 2015 (UTC)
- Coral Reef looks like a good match. Richard Avery (talk) 06:37, 13 July 2015 (UTC)
- Thank you! --Captain-tucker (talk) 08:51, 13 July 2015 (UTC)
- Coral Reef looks like a good match. Richard Avery (talk) 06:37, 13 July 2015 (UTC)
Protecting a hard drive
editHow might one protect a hard disk drive and a floppy disk from the Solar storm of 1859? Assuming modern materials available. Would putting them inside an idle microwave (acting as a faraday cage) protect them? — Preceding unsigned comment added by 2001:DA8:D800:87:12BF:48FF:FEE2:10BA (talk) 12:59, 12 July 2015 (UTC)
- The storm in question was notable for producing induced high electric voltages on telegraph lines and the article also mentions it was detected on a magnetometer. That does not mean that the magnetism was so strong that it would have magnetized or demagnetized ferromagnetic media, or caused all steel objects in a room to fly up to the ceiling like in a cartoon. I have observed hard drives and floppies surviving being in the same room with electromagnets, transformers, inductors, motors, and powerful permanent magnets. A box made of copper or aluminum screen wire could function as a Faraday cage and shield against electric fields or radio waves, but would not screen against the field of a relatively slow changing but strong magnetic field. A steel or iron box, on the other hand, would shield against a magnet pretty well. A relatively thin steel shell isolates against a fairly strong magnet, from observation with permanent magnets, coffee cans, and compasses. The ,microwave is designed to shield against very high frequency electromagnetic waves, and not to shield against strong but slowly changing magnetic fields. The steel walls of a microwave might be a good shield against a steady magnetic field, but I have doubts about the screen in the window. Edison (talk) 13:36, 12 July 2015 (UTC)
I know this might sound silly, but would a modern car be considered a faraday cage? If lightning struck the car and your cell phone was inside (and all the windows up door closed etc) the cell phone would function just fine. So incase of another Solar storm should we throw all of our computers in our cars? (I know this is facetious but good intentions I assure you) Void burn (talk) 21:01, 14 July 2015 (UTC)
How can the butterfly effect be detected using the scientific method?
editThrowing dice we can slightly alter the initial conditions and see whether this produces a quite different outcome.
However, if someone claims that the same happens to the weather, how could someone prove or dismiss the claim? You can observe change in infinitesimal small starting conditions (butterfly beating its wings, but also many other factors), and you can observe different outcomes in weather prediction.
How can you prove that the small initial variation influences the outcome? --YX-1000A (talk) 15:42, 12 July 2015 (UTC)
- Here is the answer. We know for a fact that computer systems that predicts the weather suffers from the butterfly effect where tiny initial differences in the intial conditions will lead to huge differences in weather prediction outcomes. However this DOES NOT PROVE that the butterfly effect exists for real weather, it only shows that it exists for computer simulation of the weather.
- The only way to prove with ABSOLUTE CERTAINTY that the real weather suffers from the butterfly effect is to use a time machine capable of traveling to the past. We need a Tardis
- However all is not lost. You can built a physical device called Double pendulum. This device has a chaotic behaviour in its motion. For certain, initial conditions, it will never have the same trajectory twice because it suffers from the butterfly effect. Because it is a real physical device, it proves scientifically that the butterfly effect is real for a physical system. This increase the confidence that another real physical system, the weather also has the butterfly effect. 220.239.43.253 (talk) 15:56, 12 July 2015 (UTC)
- No, your experiment with the pendulum won't increase the confidence that the butterfly effect is real for physical systems, unless they are very simple physical systems. That's similar to a die. You can demonstrate that the result of throwing one die is strongly influenced by the initial conditions. If you were to throw 1,000 dice, the result would be more or less the same no matter what you do.--YX-1000A (talk) 16:19, 12 July 2015 (UTC)
- In science, we rarely prove anything. Instead, we provide convincing evidence and make statements with great confidence. Your question contains two key phrases: the scientific method, and the word "prove" (in the context of applied mathematics). These are two very different things! In science, we use evidence to test hypotheses, refining a theory until it matches observation, demonstrating that a result is consistent. In mathematics, we apply formal rules to axiomatic principles to prove that a result is consistent.
- In the case of proving mathematical instability, we start with a definition of stability that is relevant to this context (for example, we expect a smoothly-varying partial derivative for any equation that governs natural behavior). If a sophisticated equation is used to model a weather prediction, we can perform a sensitivity analysis directly on the equation (or use a numerical method to approximate this analysis). When you read the historical publications that forged the basis of what we today loosely call "chaos theory" or "the butterfly effect," you can see that most of the emphasis is on mathematical analysis, and not empircal or observational science. In pure mathematics, we can prove that a particular equation is unstable, for some definition of stability!
- In the case of demonstrating the "the butterfly effect," the demonstration is applied to a mathematical model of weather rather than to "the weather" itself. The scientific task is to determine if this equation actually describes the real world. We collect data, using controlled experiment, to test whether the equation has predictive power. In particular, because we know that the equation has poor predictive power, we must be very careful in the formulation of a testable hypothesis. For example, the equation might be unpredictable in the time domain, but it can have powerful and testable consequences in the frequency domain; or in some other transform domain. This is one reason why phase space plots are so useful when we study chaotic systems.
- Whether the mathematical model applies to the real world is a separate question, and is not actually a matter of proof! All we can do is provide a compelling argument, guided by observation, to strongly conclude that the mathematical model matches the data.
- Nimur (talk) 15:59, 12 July 2015 (UTC)
- So, can I conclude that: we have a set of equations that model weather, and these equation are both very sensitive to initial conditions, and, the best match to empirical data that we have. And, there is no empirical evidence against the possibility of developing a complete different set of equations in the future that match empirical data even better (that means, better predictions, at lest under some aspect) but are completely insensitive to small changes in input data.
- But, coming back to the empirical method. Would this be a valid empirical experiment?
- what would happen if I try to predict the weather one year long using a given model, and, the next year I fire a furnace on my garden every day, and try to predict weather all year long again with the same model. If my predictions don't vary, would that mean that small changes are to be ignored? What if I do alternatively year after year, would we come to a point where we admit that little stuff is irrelevant (or simply compensate by other little stuff)?
- --YX-1000A (talk) 16:27, 12 July 2015 (UTC)
- Mathematics can provide a model of reality but mathematical model of reality is NOT REALITY. So whatever the mathematical model proves is not prove that it is the EXACT SAME as reality. For example: Newtonian Physics is a mathematical model of reality but it is not reality. So whatever mathematical model physicists has today is most likely NOT REALITY either. 220.239.43.253 (talk) 16:39, 12 July 2015 (UTC)
- That was not the question here. The question was what part of reality to include as input to the model, and what to ignore as irrelevant. --YX-1000A (talk) 20:56, 12 July 2015 (UTC)
- Sure - it's a model for reality, and sometimes it has a practical purpose with predictive power. Consider the 100-year flood model. Civil engineers, city planners, home buyers, and insurance agencies use the hundred year flood model to determine risk, evaluate building safety, and weigh costs against risks. The hundred year flood is a model of weather (and other scientific elements of hydrology): it uses mathematical analysis and statistics, guided by observational data, to estimate a probability of flood. But it doesn't even try to predict when that flood will happen! The model is useful: we know which terrain is suitable to build on. We can estimate with great confidence that any specific hilltop might be covered by water at least once between now and 2115. But the model does not try to tell you what year, month, or day the rain storm will happen! We have no mathematical model telling us when it will rain over the next hundred years. Any such model is subject to terrible over-parameterization and unstable, chaotic behavior. But we can still make other useful long-term predictions about the weather.
- To understand "chaos theory," or complex systems in general, you need to develop a great mathematical modeling toolkit, complete with some calculus, statistics, numerical methods, and a study of many example problems in physics. A great error in the popular-science version of "chaos theory" is to present the undisicplined version of the problem-statement: a bug flies around, and suddenly the entire universe is completely unpredictable! This is a ludicrous representation of the problem, and it's an even more ludicrous representation of the way that real mathematicians and physicists think about it. Nimur (talk) 17:43, 12 July 2015 (UTC)
- One important thing is to emphasize that outputs become inputs. An analogy that we use (in our freshman intro class with no math or science prerequisites...) is a complicated shot in pool (billiards). You make a shot and the movement of the first ball becomes the input for the second ball, which becomes the input for the third ball, and so on. A tiny error in where the cue strikes the first ball can eventually decide whether you sink the ball. Short Brigade Harvester Boris (talk) 21:27, 12 July 2015 (UTC)
- Right, that's propagation of uncertainty, with emphasis on propagation. Anything that propagates - like a recurrence relation or a differential equation - is susceptible to the problem. You can consider an idealized billiard table as a series of discrete collisions, whose inputs and output parameters are connected together by a discrete recurrence equation. The parameters - position and velocity - for all n billiard balls can be represented as a giant vector or matrix; the collisions can be written as a matrix operator; and the quality of "chaos" can be explored by observing the divergence of the Jacobian of the iterated collision operator. Or, you can intuitively interpret the problem and just watch how the collisions cause the balls to fly off in apparently-random directions! Nimur (talk) 22:22, 12 July 2015 (UTC)
- One important thing is to emphasize that outputs become inputs. An analogy that we use (in our freshman intro class with no math or science prerequisites...) is a complicated shot in pool (billiards). You make a shot and the movement of the first ball becomes the input for the second ball, which becomes the input for the third ball, and so on. A tiny error in where the cue strikes the first ball can eventually decide whether you sink the ball. Short Brigade Harvester Boris (talk) 21:27, 12 July 2015 (UTC)
- Something like that is already used to improve weather predictions. The initial conditions of the simulations are varied and if the outcome doesn't change much then they can predict the weather further in advance. If varying the initial conditions slightly leads the simulation to quite different outcomes then they can only predict reasonably accurately a short time in advance. The predictions are quite accurate when qualified by the length of time and which aspects of the weather they estimate they will be accurate. Dmcq (talk) 17:10, 12 July 2015 (UTC)
- It's clear that meteorologists will ignore the bugs flying around that Nimur points at above. But what is, for meteorology, "too small to bother?" Meteorologists will ignore forest fairies, until empirical evidence appears that indicates they can affect the weather if annoyed. Would they also ignore man-made behavior like energy consumption in the last days when predicting the chances of rain in the next days? What, who and how they decide what data get plugged into their predictive weather models as parameters? --YX-1000A (talk) 20:56, 12 July 2015 (UTC)
- Well, a meteorologist is the exact professional who produces weather prognostications, and they make such a judgement call. Anybody who uses a weather report - farmers, aviators, sailors, surfers, beach-goers... all decide what data to trust, and how much confidence to place in the predictions. At the extremes, when there is significant risk to an incorrect weather prediction (as may be the case for some aviation weather applications), the bar is set much higher for confidence in predictions.
- If you want to read about how modern weather prediction is performed by the United States National Weather Service, you can visit their Model Analyses and Guidance website (which is a little bit technical: the Product Description Document is a good place to start getting oriented). There are many products (i.e., many types of prediction models). For many, human activity is not a factor - I doubt that there's any human activity index input to the mesoscale precipitation forecast, for example. Other predictions - like the air quality index - certainly account for predictions of human activity. Such forecasts are used, for example, to produce Air Quality predictions and smog alerts. Here is an overview of specific human-factors data and measurements used in the Air Quality Model.
- If you're totally lost, a theoretical introduction to meteorology is called for. A few great introductory books that are available at no cost include Aviation Weather, (which has some fun cartoons scattered throughout).
- Most people let a TV-meteorologist reduce their weather forecasts to a much simpler data product: "partly sunny" or "chance of rain." At such a coarse granularity, the predictive power is quite weak and the forecast-model is completely hidden from your view, so you don't know what went into it. Usually, the only people who really deal with the quantitative details are people who care a lot about weather: pilots, sailors, farmers, meteorologists, and applied physics enthusiasts.
- Nimur (talk) 22:41, 12 July 2015 (UTC)
- It's clear that meteorologists will ignore the bugs flying around that Nimur points at above. But what is, for meteorology, "too small to bother?" Meteorologists will ignore forest fairies, until empirical evidence appears that indicates they can affect the weather if annoyed. Would they also ignore man-made behavior like energy consumption in the last days when predicting the chances of rain in the next days? What, who and how they decide what data get plugged into their predictive weather models as parameters? --YX-1000A (talk) 20:56, 12 July 2015 (UTC)
- Mathematics can provide a model of reality but mathematical model of reality is NOT REALITY. So whatever the mathematical model proves is not prove that it is the EXACT SAME as reality. For example: Newtonian Physics is a mathematical model of reality but it is not reality. So whatever mathematical model physicists has today is most likely NOT REALITY either. 220.239.43.253 (talk) 16:39, 12 July 2015 (UTC)
- Models for numerical weather prediction aren't made from fits to data, but are based to the greatest extent possible on fundamental physical laws. The core of a NWP model is Newton's second law of motion, the first law of thermodynamics, and conservation of mass. These equations when coupled do indeed exhibit chaotic behavior. If someone can come up with a an alternative physical framework that does not use these equations I'd be most interested to know. Short Brigade Harvester Boris (talk) 21:15, 12 July 2015 (UTC)
- @YX-1000A: Some key info I don't think anyone has posted: the so-called "butterfly effect" can be formalized and even quantified in terms of the Lyapunov exponent of a dynamical system. If you are curious about how such a thing could be measured from real-world data or experiments, see e.g. these peer-reviewed publications (should be accessible) [5] [6] [7]. The first,DETERMINING LYAPUNOV EXPONENTS FROM A TIME SERIES is published in the highly reputable Physica, and should serve as a highly authoritative starting point for understanding how scientists detect and quantify "butterfly effects" in real-world systems. In regards to the previous discussion - these analyses would not be derived from a model, but derived from data. No need to get in to full-fledged weather prediction - even the simple logistic model will display deterministic chaos (remember, biology was the other key historical science that motivated chaos theory as a discipline). If you're interested in that angle, you'll get plenty of hits searching google scholar for /lyapunov exponent population/, with many applications of the above methods for very specific real systems - [8] SemanticMantis (talk) 13:47, 13 July 2015 (UTC)
Is fog harvesting impossible too far inland?
editAll the projects for harvesting water from fog that I've seen (admittedly on a very cursory inspection) seem to take place fairly close to the sea ([9], [10], [11], [12]). Is fog absent too far inland? How come none of those projects take place say in the middle of the Sahara or in the Gobi desert? Is there a difference between absence of humidity and absence of fog? For example the Atacama desert, one of the driest places on earth (I suspect that is defined as minimal rainfall, not average humidity of the air?), seems nevertheless to have enough fog to make a fog harvesting project worthwhile at least in theory. (Although the WP article says in practice it was a failure.) Thank you for any information clarifying this interesting topic. Contact Basemetal here 20:00, 12 July 2015 (UTC)
- There might be a language issue here. "Fog" normally refers to visible water droplets in the air. The air always has some humidity, even in a desert, but it doesn't always form those visible water droplets. One advantage of a desert is that temperatures drop very low at night, due to a lack of cloud cover to keep the heat in. Thus, even though humidity and thus the dew point are low, the temperature can still drop below it, and you can get dew. So, that dew can be collected, say using a large tarp over an inverted cone dripping into a bottle. This only produces a rather small amount of water, so perhaps enough for survival, but not enough for household needs, like bathing, cooking, washing clothes, etc. (You could theoretically make hundreds of such rigs to collect enough water for all that, but this would be rather impractical, as it would all be damaged in the first sandstorm.) StuRat (talk) 21:07, 12 July 2015 (UTC)
- Are you saying the references I gave do not know the difference between fog and dew? Contact Basemetal here 21:33, 12 July 2015 (UTC)
- No, only that, while the process is similar, it's likely to be done on a smaller scale in the desert, as a survival method rather than a permanent source of water, and have a different name. It's like the difference between an oven and a camp stove. StuRat (talk) 21:37, 12 July 2015 (UTC)
- BTW, one of those sources said that water collected that way doesn't need to be filtered or processed in any way. I'm not so sure about that. Bacteria and dust will be collected, too, and boiling it or otherwise disinfecting it seems like the safe thing to do. StuRat (talk) 21:44, 12 July 2015 (UTC)
- I am still baffled. Why is none of those projects in the middle of the Sahara? Don't they need water there? It can't be humidity level per se since the Atacama desert is extremely dry, yet there was such a project in the Atacama desert. What makes such a project viable in the driest place on earth but not in the middle of the Sahara? Any ideas? Contact Basemetal here 22:13, 12 July 2015 (UTC)
- Noting your use of the word was, I suspect that the Atacama desert project wasn't viable after all. You would need a lot of those collectors, each collecting only a few ounces of water in that low humidity. Then there's the issue of how to collect that all into a sealed reservoir. You'd either need lots of pipes, or to collect it manually by dumping little jars of water into larger ones and delivering those to the reservoir. Then, as I mentioned, the collectors would need frequent maintenance after sandstorms or just normal wind. After doing this you still wouldn't have nearly enough water for agriculture, so how exactly would the people there make a living ? Perhaps if they were miners and there was a rich mineral wealth there, that might make some sense, but they would still need to build all the other infrastructure needed, like roads. I suspect that in such a scenario, the area would become a ghost town as soon as the mineral wealth was gone. StuRat (talk) 02:05, 13 July 2015 (UTC)
- [citation needed]. Looking at [13], it seems viable enough to have plants growing. And the people behind it say that when they have enough, they will change the face of Arrakis. :) Seriously - but I'm only guessing here - I would distinguish between two measures of the "dryness" of a desert. One is the amount of rainfall; the other is the relative humidity. Atacama gets a lot of fog, which means it must have pretty high humidity levels on frequent occasion; it is merely in a double rain shadow that prevents drops of water from falling. So if you could tip the balance a bit, you might get somewhere. It's possible that a large continental desert like the Sahara, receiving winds from all directions but generally dried out by air descending from the high atmosphere and warming up, might have a lower average humidity even while getting more millimeters of rain in the gauge. But that's a guess. Wnt (talk) 20:20, 15 July 2015 (UTC)
- The fog indicates high relative humidity, but that mostly means low temperatures, so the little humidity in the air forms fog. They claim it can be scaled up, but I'm rather skeptical. After all, each fog fence removes some of the small amount of water from the air, so a million fog fences won't get you a million times as much water. I'm reminded of the merry-go-round water pumps in Africa, which were supposed to pump water as children played on them. Sounded good at the time, but the kids got tired of them (not much fun because they don't spin long, due to the drag of running the pumps) and there was no way to fix them once they broke. I suspect something similar happens here, where, as long as some charity is footing the bill to maintain the fog fences, then the locals will be happy to get the free water. But, once they have to pay the costs, it will no longer be worth it. StuRat (talk) 21:09, 15 July 2015 (UTC)
- I would be quite amazed if humans can dehumidify the outdoor air with a simple construction project, even if there is one in every town. (If they can do that, maybe they should build these things in wet, humid areas to make them more comfortable...) I read Fog collection which narrates the tale in Chungungo sort of how you describe, but on a search I found [14] which says the collectors went out of use in 2000 after the government built a desalination plant. (AFAICT the desalination plant provided water at government expense, whereas the fog catchers remained a local project, so there was no local financial motivation to keep them going, but I'm inferring based on a few vague sentences) Wnt (talk) 23:18, 15 July 2015 (UTC)
- "One in every town" ? Wouldn't they need thousands in every town to provide for the town's water needs ? According to your link, 12 fog collectors produced 2000 liters per day. That's under 17 liters per fog collector per day, and presumably only that much in ideal locations. Certainly not enough water for a town, and not enough for agriculture, either. Your link also says they had "funding from the Global Environment Facility", so not local funding. StuRat (talk) 00:25, 16 July 2015 (UTC)
- The articles shown don't explain what the difference is between one of these "fog collector" machines and a dehumidifier. Both extract water from the air. Both work well in high humidity and do nearly nothing in low humidity. 199.15.144.250 (talk) 12:40, 13 July 2015 (UTC)
- A dehumidifier typically uses electricity to remove most of the humidity from a small room, while a fog collector uses no electricity and collects only a small portion of the humidity from an outdoors area. StuRat (talk) 12:58, 13 July 2015 (UTC)