Wikipedia:Reference desk/Archives/Science/2012 October 8

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.

October 8

Accuracy of forward looking economic statements

There are many cases were businessmen, economists, and similar people have to predict what the market is going to look like for their product weeks, months, or even years in advance. For example, farmers hope to plant crops that will give them a good profit come harvest time. Or alternatively, a book publisher tries to predict demand when deciding on the size of a print run. In some cases there are robust futures markets to help, in other cases in-depth business studies might be undertaken, but often such predictive judgments seems to boil down to some expert's personal opinion. I'm looking for materials I might read reviewing the successes and failures of such forward looking economic statements. For example, when experts predict that the GDP will increase over the next 12 months, how often are they right? Do the target prices set by stock market analysts have predictive value? Etc. Also, I would like to read about cases where detailed mathematical modeling has been applied to economic problems, and where such models have tended to do well or poorly. Thanks for your help. Dragons flight (talk) 11:29, 8 October 2012 (UTC)

I have no answer to the general question, nor, unfortunately, references. But one of the first scientific meetings I attended (back in 199X as a student) was a local meeting of machine learning researchers. Two or three groups had tried to predict stock prices from past performance, and their analysis seems to agree that the past data has essentially no predictive value. I don't know if that is still the state of the art, and I know even less how this applies to more macroscopic questions. --Stephan Schulz (talk) 12:04, 8 October 2012 (UTC)

I recall reading a review of a book on just this subject... I believe it was this one. The book's conclusion, and the conclusion of many other studies, is that experts have a positively lousy track record for accurately predicting the future broadly. You might check out the book, though. --Mr.98 (talk) 14:59, 8 October 2012 (UTC)

A well-received book with a similar theme is The Black Swan (as well as the author's earlier book Fooled by Randomness). They're not going to have comprehensive data about what you seek, but rather selected examples to further the author's argument (basically that predictions of the future are worthless). Buddy431 (talk) 23:01, 8 October 2012 (UTC)

At one stage in my career, I worked in the planning group of a large governemnt department, making decisions on the provision on capital infrastructure. This involved working out the Net Present Value (see http://en.wikipedia.org/wiki/Net_present_value), and applying Jacobson Provisoning Theory. JPV is a mathematical theory that tells you what size increments to install to meet a requirement that grows over time in order to minimise the cost, for example the provision of hospital beds in a city whose population is growing. You could build in one big project that meets growth requirments for twenty years and get econoimies of scale, or build say every two years another ward. JPV tells you which costs less. Much infrastructure can only be planned on a long term basis, 10 or 20 years.

The point is, both NPV and JPV require a prediction of demand growth, interest rates, and inflation in order to make any sense. We used to get a yearly report from Treasury that predicted these things. Generally, they gave upper and lower limits at a certain level of confidence, for intervals 1, 2, 5, 10, and 20 years out. Experience showed that their predictions within thse limits were very good, with certain exceptions I will describe below. Naturally the limits got broader as the time frame increased. My experience is Australian, but if Australia could do it, other advanced countries can too. These reports were not made public when published, but would be available from National Archives.

Economic predictions released to the media by certain "experts" are far far less reliable, as they don't have the resources, and often use cheap data of dubious value, such as the number of job adverts printed in morning newspapers.

Treasury ecomonic predictions generally made an assumption that the government of the day would remain "competent", and say something like "we expect the politicians to take this reasonable option, or possibly that option, and factored this in when estimating the upper and lower limits." Mostly, Australia has had "competent" governments - the sort where the political leaders took advice from the public service. An example from Australia is the Menzies Govt (conservative) in the 1950's and 1960's. At other times we've had "incompetent" govts, where political leaders assume the public service is not to be trusted. This can lead to reform, but makes new economic predictions very difficult, and existing predictions incorrect. The 1970's Witlam Govt (Labor) is an example. More recently, we've had political leaders that intelligently challenge the advice from the public service. The Keating Govt (Labor) is an example. This is the best form of government in all respects and allows reasonably good economic predictions.

Things like war breaking out, oil crises, can ruin any prediction.

It's more difficult now than it used to be. Today's politicians don't have the intellect of earlier leaders, and tend to go more on ideology than on facts or reasonable inference. Also, economies used to be quite regional, but economies are now strongly linked globally. For example, Australia and the USA used to be largely industrial - with large numbers of folk employed in factory production of all kinds. This makes things controllable. Nowadays, factory production is not the great employer - the service industries are. This makes things unstable. Education standards are falling. As less well educated thinkers enter the workforce, things will get even more unstable.

Wickwack124.178.43.210 (talk) 00:25, 9 October 2012 (UTC)

This (article starting on p.38) is from 2001, but I'd be surprised if much has changed in the forecasting game. Zoonoses (talk) 05:55, 10 October 2012 (UTC)

See Efficient-market hypothesis. This theory, which has had a lot of holes poked in it but still is considered to have a large element of truth in it, says that as soon as any information relevant to the appropriate price of a particular stock, or to the level of the stock market as a whole, becomes available, it is instantly incorporated into the stock price; any trades that take place after the info becomes available will take place at the newly appropriate price. So no one can beat the market unless they can obtain or predict new information before it becomes generally available to stock traders. In the absense of insider trading, this is tough, though lots of investment advisors are in the business of trying to do it. But studies show that none of them consistently beat the market -- at any given time, someone is beating the market, but later on it's someone else who's beating the market. So the answer is no, stock market forecasts and forecasts of individual stock prices are no better than throwing darts at a dart board, and the above is the theoretical reason for that. The classic book on this is Burton Malkiel's A Random Walk Down Wall Street. Of course an infinite amount has been argued this way and that way about these ideas over the years, such as questioning why the market as a whole didn't see the recent financial crisis coming (in which case the bubble would never have been allowed to grow in the first place, and therefore there would have been no bubble to burst).

As for non-financial data predictions, see Economic forecasting and the links therein, and Economic model#Tests of macroeconomic predictions. Duoduoduo (talk) 18:29, 10 October 2012 (UTC)

Pyramids at Giza

A claim often repeated on the internet is that the Great Pyramid at Giza could not be reproduced with today's technology. It is aligned to true north with a greater degree of accuracy than any modern building and we could not match or better it. The quality and accuracy of the stonework is beyond anything that could be done today, even with today's modern tools and equipment.

I just cannot believe that this is true. The sort of website this appears on is usually to do with the paranormal and that fact that the pyramids must have been built by Alien's etc - yet it even appears on some more "reputable" sites.

Surely this is pure bunk? — Preceding unsigned comment added by 195.188.208.251 (talk) 11:53, 8 October 2012 (UTC)

Another claim is that Giza is located at the precise geographical centre of the earth's landmass and the chances of the pyramid being located there are billions to one. I have researched and debunked this particular gem - how do you find the centre of the surface of the sphere???? — Preceding unsigned comment added by 195.188.208.251 (talk) 11:52, 8 October 2012 (UTC)

That last one is more appropriate to the math desk, I guess. However, the claim is not "the center of the surface of the sphere", but "the center of the landmass". Land is distributed unevenly over the surface of the earth, so there is an asymmetry that can be used to define a "center" (of course, for most definitions there would be two, one on either side of the earth). Given that there are plenty degrees of freedom for defining "land" (what about floodplains? Do mountains count more than plains? What about the dead sea?) and "center", it's probably easy to come up with a formula that places the pyramids (well, one point in their vicinity) in the center. --Stephan Schulz (talk) 12:31, 8 October 2012 (UTC)

I found [1] which indicates that the center of mass of the upper layer of the Earth is displaced 29.0 km toward 15.4 N, 150 E ... which looks like the middle of Mariana Trench. Am I reading this backwards? Admittedly I haven't figured out quite what I'm really reading about here as of just yet... they even talk about a "magnetic center". Wnt (talk) 14:46, 8 October 2012 (UTC)

Presumably, they mean "land area", as in areas where the land and is above sea level. The center of that should be above the equator. If you neglect the Americas, it should be further East. So, perhaps, including the Americas, that might bring it back into Africa. However, the Earth, being round, brings up the question as to whether a pocket of land near the opposite side draws the center of mass towards it, on whichever side is closest, weighted 100% for that area, or whether the area weighting is reduced to account for it "pulling from both sides". You can also find the CG point in 3D, putting it near the Earth's center, then project it back up to the surface, normal to the nearest surface. This seems like the best approach, to me. StuRat (talk) 19:40, 8 October 2012 (UTC)

Where do you slice open the earth and how do you spread it out on a flat surface to define the center of any subset of it? --Jayron32 19:44, 8 October 2012 (UTC)Edit: Stu made a clarifying point here which renders my question somewhat silly. When he originally made his statements, it was unclear what he was defining. It's clear now, and my question is now moot. --Jayron32 19:48, 8 October 2012 (UTC)

(ec with Jayron's addition above) I don't think you need to. Aside from the method I listed above, another would be to try every point on the surface of Earth, and for that point find the weighted great circle distance to each bit of land. I could easily write such a program, if I had a file with, say, each square kilometer's longitude and latitude numbers, along with the percentage of land above water (smaller resolutions would make the program really slow). You would want to first look at large intervals, let's say a point every 30 degrees of longitude and latitude, then look at a smaller interval, say a point every 10 degrees of longitude or latitude, around the previous best result, etc., until you get an answer of sufficient accuracy. StuRat (talk) 19:53, 8 October 2012 (UTC)

Our article Great Pyramid of Giza gives numbers for the precision of its construction. The sides of the base are over 230 meters long and have an average error of 58mm. The base is horizontal and flat to within 15mm. The finished base has an average alignment error (from true north) at the corners of 12 seconds of arc. Those numbers are very impressive for such an huge and ancient construction, but I can't see why we wouldn't be capable of reaching the same accuracy today. I can't find good references for the tolerances of large modern buildings, probably because the accuracy in placement and construction simply isn't noteworthy. 209.131.76.183 (talk) 14:55, 8 October 2012 (UTC)

For what it's worth, most modern construction uses a tolerance of about 1/4" in 10 ft (less than 1 cm in 3 m), non-cumulative, for everyday work, and greater accuracy can be attained at extra cost, bearing in mind that settlement, thermal expansion and loading can and will have effects (though the latter two are probably inapplicable for a pyramid). I don't see the challenge of layout to the observed standards as beyond the means of an early civilization which chooses to devote the appropriate resources. Acroterion (talk) 15:06, 8 October 2012 (UTC)

We can compare that to the Channel Tunnel. The length of each bore is 50.45 km (per that article). The Institution of Civil Engineers' book The Channel Tunnel (ISBN 9780727719225) p54 says the two bores for the service tunnels (the first to meet) were off by "4mm vertically and 15mm horizontally". -- Finlay McWalterჷTalk 15:10, 8 October 2012 (UTC)

• The "center of the land mass" question has come up on the Ref desks several times before, see for example Wikipedia:Reference desk/Archives/Science/2011 January 27#Geographic center of the world. Looie496 (talk) 20:19, 8 October 2012 (UTC)

And here is a very relevant article that was linked in the discussion: Geographical centre of Earth. It mentions Giza and some of the historical claims about the location. 209.131.76.183 (talk) 11:43, 9 October 2012 (UTC)

• The precision of construction is remarkable, but not unearthly. I should note that Thuban was, in 2787 BC, quite a reliable pole star, less than 2.5 arc-minutes away from true north. But by 200 years later it would have been a full degree from the pole, so presumably the people laying the lines of stones would have had to watch its circling quite carefully to come within 4 arc-minutes. I wonder if, at the time, they derided their efforts as a sloppy job, wistfully remembering the stories of the constant pole-star of a bygone age? I assume they had a choice of many stars to watch coming up from the east to use to align the sides going in that direction. The levelling is another question - if it rained on the base of the pyramid, how accurately could you level it based on which way it flows? (Though the Sahara was apparently 'as dry as it is today' by 3400 BC, so maybe they had to flood it by hand?) Wnt (talk) 16:54, 10 October 2012 (UTC)

Aviation-related question on the humanities desk.

I'm taking the liberty to cross-post, because there may be someone knowledgeable here, who does not usually read the humanities desk. The question is Bob Hoover anectote - Urban legend? If someone has information relevant to the question, please answer at the humanities desk. Thank you. --NorwegianBlue ^talk 12:23, 8 October 2012 (UTC)

de-salting urine

In the lab where I work, we sometimes do thin-layer chromatography on urines. Part of the prep is to "de-salt" the urines by adding propan-2-ol to a small amount of the urine, and then freezing these preparations for at least one hour before using them. What I would like to know is:

• what does it mean to "de-salt" urine? What's happening?
• How does propan-2-ol achieve this? What are its properties that make it work?
• Why is the freezing necessary?

I have a decent understanding of chemistry and so replies hopefully don't need to be dumbed down too much! Thank you. Sternenjägerin (talk) 18:00, 8 October 2012 (UTC)

Desalting is used in DNA extraction techniques. Here and here are very short explanations of why and how it is done. Does that help? --Jayron32 18:10, 8 October 2012 (UTC)

Found an even better one. This page explains why salt, freezing, and isopropanol are all used in preparation of a DNA sample. I assume most of these techniques are used for Gel electrophoresis rather than TLC, but there is likely a similar rationale. --Jayron32 18:14, 8 October 2012 (UTC)

• It would be easier to answer this if I knew what compound(s) you were trying to isolate. There are many different reasons that could come up. From [2] I see that one reason is that during liquid-liquid extraction, stable emulsions can occur, which "might be overcome by either an increase of the phase ratio (organic/aqueous), or saturation with neutral salts, or freezing of the aqueous phase, or the application of supported LLE (adsorption of the aqueous phase on diatomaceous earth before extraction with water immiscible organic solvent)." Extractions of this type are a bit of an art, devised by trial and error, usually by frustrated postdocs or grad students who will do whatever comes to mind trying to get a procedure to work this week rather than next, which will then be followed religiously (by which I mean, by rote and without critical thought) by what may be literal generations of future researchers. (If you think I'm kidding, you should search "rubidium chloride" sometime) [3]

• If the above source is the application you're thinking of, then its role there is very different from in DNA extraction. For DNA, the idea is that when you have a less polar solvent, the highly charged phosphates that make up the DNA backbone will find some nice positive ion someplace to snuggle up with, and bring it out of solution in big white lumps. The excess salt is simply washed out with 70% ethanol later on, because the DNA is huge and won't dissolve in ethanol, especially once you've compressed it up against the wall of a centrifuge tube at high g-forces, but the salt can. But for resolving the emulsion, you're trying to make it so that the bits of aqueous and of organic in a mixture will fuse together. (Actually, emulsions aren't a topic that I understand very well; but you can see from this much that it must be different from what you do to DNA) Wnt (talk) 05:55, 9 October 2012 (UTC)

• Thanks for the answers so far. I had a look at the links provided by Jayron and they didn't really help, because DNA extraction is a very different thing. The TLC I'm doing is for sugars in urine; glucose, sucrose, fructose and such. I'm aware that this is quite a common application, but most of what I've found on the web assumes a basic level of understanding, or just says, "this is what works" without making an explanation as to why. I know that desalting helps give clearer streaks, less 'clumpy' interference I suppose, but wondered if there were also other reasons, perhaps dropping out of solution some compounds that would coelute. I also wondered if propan-2-ol does some deproteinising as well (based on the sometimes rather large pellets after centrifuging). Sternenjägerin (talk) 07:06, 9 October 2012 (UTC)

Without more details, it's hard to guess what's going on. First hit I got for "desalting urine TLC" mentions drying the urine first (evaporating off the water). There's a huge difference between a salt-sugar-isopropanol mixture with water and one without. Salt will separate a water-isopropanol solution, salt solubility in water and in isopropanol decreases at lower temperatures, freezing water will drive out salt, freezing a water-isopropanol mixture is fractional freezing; Sugars dissolve in isopropanol, salt to a lesser extend. All these may play a role, but without more information, I can only guess. Freezing for at least one hour doesn't make much sense imo, not much can happen when everything is frozen. Slow freezing, partial freezing and cooling down a liquid on the other hand have many uses. If we're talking about a waterless mixture, you may drive out the remaining salt by lowering the temperature and with it the solubility, a better alternative to using cold isopropanol in the first place because that would make the dissolution of the sugars unacceptably slow and you'd get water condensation from the air. Ssscienccce (talk) 13:22, 10 October 2012 (UTC)

If you can link any of those things you found on the Web that detail the basic procedure it will be a lot easier to figure out what they're talking about. Wnt (talk) 16:36, 10 October 2012 (UTC)

cells

if a cell under goes cell division every minute how many cells will you have after five minute? — Preceding unsigned comment added by 69.121.183.37 (talk) 19:43, 8 October 2012 (UTC)

This sounds like part of a homework problem on exponential growth. Is there anything in particular about the problem that you are having trouble understanding? 209.131.76.183 (talk) 19:51, 8 October 2012 (UTC)

You formula is (C)2ⁱ, where "C" is the number of cells initially, and "i" is the number of time increments, with doubling occurring at each time increment. However, note that doubling at that rate can't happen for long, as soon the resources the cells need will be used up and they may also poison (or eat) each other. StuRat (talk) 22:15, 8 October 2012 (UTC)

Doubling at that rate can't happen at all. Looie496 (talk) 00:57, 9 October 2012 (UTC)

What if the petri dish or whatever they're in is being supplied with nutrients and toxins are being removed somehow? ←Baseball Bugs ^{What's up, Doc?} carrots→ 03:51, 9 October 2012 (UTC)

You could keep it going a bit longer that way, but soon the petri dish would overflow, as would the barrels, vats, and tanks you moved them to next. StuRat (talk) 03:58, 9 October 2012 (UTC)

OK, if I'm reading you correctly, they would collectively take up too much room. ←Baseball Bugs ^{What's up, Doc?} carrots→ 02:33, 10 October 2012 (UTC)

Correct. If we double the population with each cycle, we end up with over a thousand times the initial population in 10 cycles, a million times in 20 cycles, a billion times in 30 cycles, a trillion times in 40 cycles, etc. StuRat (talk) 21:22, 12 October 2012 (UTC)

Depends on how long a minute is. You can still calculate 2^x based on a given x. The question seems to be at about a fourth or at the most fifth grade level. μηδείς (talk) 02:48, 9 October 2012 (UTC)

Bugs & Stu, I am fairly certain Looie meant doubling can't happen at the rate of a minute because mitosis takes on the order of an hour to repeat; one source says (without specifying if they mean time to occur or to repeat) of 8 minutes in a fly embryo and 20 minutes in bacteria. (Hence my cryptic remark as to depends how long a minute is.) Also, the early stages of cell division in human embryos don't include growth, so the many-celled embryo when it implants is the same size as the egg when it is fertilized. μηδείς (talk) 01:48, 12 October 2012 (UTC)

Unknown flower on Japanese woodcut

This Iwasaki Tsunemasa woodcut is from the Kew site but without ID. Anyone know what it is? It looks like some sort of Peony, but I'd like to be sure. Perhaps someone who reads Japanese...... Thanks Paul venter (talk) 20:38, 8 October 2012 (UTC)

To save others the trouble, neither TinEye nor Google reverse-image-search yields much of interest for this image, bar its Kew page and some mirrors and Pinterest-pins of it. It seems that the name Honzo Zufu simply means "plants illustrated" (the name of the work from which this illustration comes), and Kew (et al.) simply calls it "blue flowers". -- Finlay McWalterჷTalk 22:23, 8 October 2012 (UTC)

That sounds cool, how does one do a reverse image search? (BTW, a google search for blue flower is also pointless, and yields mostly falsified images.) μηδείς (talk) 01:47, 9 October 2012 (UTC)

http://www.tineye.com/ --Jayron32 02:14, 9 October 2012 (UTC)

That's quite cool. I would consider giving you a proper star, if it wouldn't get me excommunicated. I am wondering if perhaps we (1) alread do have a page with a list of resources like http://www.tineye.com/ and http://books.google.com/ngrams/ or (2) if wwe should create one? μηδείς (talk) 02:46, 9 October 2012 (UTC)

There's Wikipedia:WikiProject Resource Exchange/Shared Resources which is a bit unweildy, but it has a bunch of resources. There's also Wikipedia:List of free online resources and really anything from Category:WikiProject reference libraries. You could add to something from there, or you could also create a short list in your own userspace of your favorite stuff. Or even in the Wikipedia space, if it doesn't substantially duplicate the purpose of some other resources. --Jayron32 03:31, 9 October 2012 (UTC)

Google also allows a reverse image search, go to Google Images [4] and click on the camera icon. Nil Einne (talk) 06:12, 9 October 2012 (UTC)

Thanks to both, I figured there had to be something like that. μηδείς (talk) 18:06, 9 October 2012 (UTC)

Given this hasn't been answered I posted it to the lang desk. μηδείς (talk) 03:44, 10 October 2012 (UTC)

No luck at the language desk, thanks to Kagetora. It does look somewhat like a peony, but according to our article they are red, yellow or white. Also, the little yellow heads make it look like a member of the Compositae, but peonies are in the Saxifragales. Peonies do have yellow anthers, but they are elongate, not spherical. μηδείς (talk) 06:28, 10 October 2012 (UTC)

It might still be enlightening to know what the Japanese inscription says. The images at look pretty similar. The blue colour is just artistic licence. Peonies used to be Ranunculaceae and the image does shout that out. Paul venter (talk) 20:46, 10 October 2012 (UTC)

I don't think you can suggest the blue color is just artistic license, given the book is a medicinal, and the text according to Kagetora's comment at the lang desk mentions the blue color. Also, the leaves as depicted are sessile and simple, while the peony's are petiolated and compound. μηδείς (talk) 02:27, 11 October 2012 (UTC)

KägeTorä interprets it as 'purple' rather than 'blue' - so perhaps the blue colour was more of a printer's gremlin than a colour that reflected reality - herbals get that a lot. Paul venter (talk) 06:13, 11 October 2012 (UTC)

(Yes, my previous post said "blue color' before KageTora had offered his reading of murasaki as purple.)

Japanese splits up color space slightly differently from English. The term aoi covers a range including colors we consider blue and green. That murasaki is usually translated as purple doesn't necessarily mean it doesn't also cover indigo, which is the word I would use to describe this flower's color if I wanted to be more specific than blue. (Indeed, ROYGBIV has the two colors indigo and violet that overlap in the purple range, while Russian, for example, doesn't even have a single basic word that covers the blue range in the English sense. It has голубой, (goluboi) light blue, and синий, (sinii) (darker) blue, which are considered separate basic colors in the way we consider green and yellow separate basic colors.) μηδείς (talk) 17:01, 11 October 2012 (UTC)

That is all truly interesting - if only we knew what the flower was! Paul venter (talk) 17:19, 11 October 2012 (UTC)

My expert on NE N. Am. plants agrees with my opinion that it looks like a member of the compositae but has no better guess. The big problem is that we can't see into the center diagnostic part of the flower to tell if the yellow things are sexual parts (anthers) as they would be if it were a peony, or actually little separate flower heads in a big composite like they would be in a daisy or a mum. The leaves, which are depicted as simple, dentate, opposite and sessile, rule out a peony.μηδείς (talk) 01:20, 12 October 2012 (UTC)

I asked at a forum in Japan and got this database link. See this image. It's a different version, but the description is same. 千葉 is Chiba. 鋸 is a saw or this might be Mount Nokogiri (Chiba). 紫 is purple. The sentence could be "Found purple one at Mount Nokogiri, Chiba" or "Found saw-purple one in Chiba". 紫色淡紫色ノ物 means "Purple and pale purple colored one". The other flowers on the pages before and after are poppies. Someone in the forum said it was P. somniferum var paeoniflorum. The Japanese name of the flower is 牡丹芥子/peony poppy. I'll call u-tokyo and asked for details. If they are kind and answer my questions, I'll post the information here. Oda Mari (talk) 05:30, 12 October 2012 (UTC)

I had found a database at National Diet Library before I asked at the forum, but could not find the image as there were too many images. As I knew which volume to look up, I found this one, the same version of ours. I noticed the description was different. It says 鋸歯/serrate leaf. I think this description is the correct one. This page says it's piony poppy. The flower is definitely P. somniferum var paeoniflorum. According to the description of the book, it's not wood block print, but a manuscript (1844). As for u-tokyo, I have to e-mail for inquiries, but I don't think the inquiry is needed anymore. Oda Mari (talk) 08:17, 12 October 2012 (UTC)

Lovely bit of detective work! Poppies before and after our plate surely clinch the ID. Thank you for the effort and interest! Do wish I could read Japanese! Paul venter (talk) 09:08, 12 October 2012 (UTC)

Ah, yes, brilliant, it is quite obvious from the leaves and petals it's a poppy. The little yellow spheres inside the flower are quite inaccurate depictions of the ring of elongate yellow anthers. But still very interesting. μηδείς (talk) 17:46, 12 October 2012 (UTC)

Artistic licence, perhaps, and the ability to read Japanese........Paul venter (talk) 21:25, 12 October 2012 (UTC)

PSYCHOLOGY AND SCIENCE BREAKTHROUGH QUESTION

this is half psychology and half science breakthrough question. the psychology question follows.

is it possible to build time viewing machine to very distant past by finding a shorter path light takes around massive objects and following short cut catch up with it and view very distant path through powrful megascopes?

secondly isn't your opinion of the science validity of this question greatly improved by reference to this stupid meme - http://www.quickmeme.com/meme/3r8eie/ - which is far less useful, as it does not describe how someone might get the distant past image back to earth to view. whereas, finding a tortuous light path that ends up near us is much more likely.

followup question:

- is anything shiny enough and millions of lightyears away to reflect even a single photon from earth back to earth, today? if so, then is even a single photon hitting earth on average of once per year that shows earth from the dinosaur era? — Preceding unsigned comment added by 80.98.245.172 (talk) 21:59, 8 October 2012 (UTC)

Warning: Many consider using all uppercase letters here to be a capital offense. StuRat (talk) 22:06, 8 October 2012 (UTC)

A photon would be not so much reflected by a mirror as deflected around a massive object, like a black hole. It's possible a few "dinosaur photons" from long ago might return to Earth, however, the problem is the signal to noise ratio. That is, how would we know which photons are from dinosaur Earth, versus the huge number of photons from other sources ? And, even if we could identify them all, I doubt that there would be enough to see anything from them. StuRat (talk) 22:10, 8 October 2012 (UTC)

(ec)First, please don't post in all caps; it is the internet equivalent of shouting and is generally harder to read (I have altered the text to lowercase). For the first question, yes, in principle, a telescope can act as a "time viewing machine", and if you throw out engineering limitations, a telescope plus a distant mirror could let you observe Earth millions of years ago. However, you can't see anything prior to the point when the mirror was constructed, so unless we find giant space mirrors floating around, we can't actually see anything on Earth before now (though we could allow our descendants to see us now, or as soon as we build said mirror). A gravitational lens could, in principle, act as such a mirror, but there are no suitable candidates (plus all those engineering limitations re-emerge as soon as you want to talk about actually doing something). For the last, there is no super-shiny space mirror out there, so far as we know, but it wouldn't be unreasonable to expect photons to return after some arbitrary number of usual reflections, etc -- though anything meaningful will be long-since lost in the noise. — Lomn 22:16, 8 October 2012 (UTC)

Quibble: You could view things reflected from Earth, at twice the age of the distant mirror, all engineering considerations aside. StuRat (talk) 22:24, 8 October 2012 (UTC)

Since the Apollo moon landing program, scientists have been able to see backward about 2.6 seconds at least so far as determining that a laser beam was transmitted to bounce off the [[Lunar Laser Ranging experiment] installed then on the Moon. In principle, if not in practice, if a flat mirror many kilometers in diameter were built on the Moon, couldn't we see images of what happened 2.6 seconds ago? A large enough mirror on the Moon might allow a Hubble quality scope in near Earth orbit to see an object 400 meters across on the Earth 2.6 seconds ago, judging by an article on aiming Hubble at the Moon, and doubling the distance.. I expect the cost of such a mirror to provide even 400 meter resolution of events 2.6 seconds ago would be staggering, but (again only in principle) couldn't a mirror farther away allow a 1 hour or 1 day (or any particular period) retrospective view of Earth? How big would the mirror have to be for the system to function? Would making the mirror concave like a long focal length telescope mirror help? A quick calc implies that with Hubble, if the mirror were 1000 times the Moon's distance, it could see 2600 seconds (43 minutes) into the past, but the resolution would only be 400 km. A bigger telescope could improve the resolution. Just sending the telescope to the remote point to look back at Earth would likely be cheaper than putting a giant plane or convex mirror far from Earth. A more practical way to view our past (fuzzily) is when a distant spacecraft looks back at Earth , like Pale Blue Dot when Voyager 1 took a picture (if a pixel is a picture) and sent back an image of how the Earth looked 11 or so hours earlier. If we built a humongous and impractical mirror in space to reflect back light to a nearby star, it would in principle let them see themselves as they were, if they built an impractically large telescope. Such reciprocity would get around the need to wait eons for our own space telescope to get lightyears away from Earth. Edison (talk) 00:28, 9 October 2012 (UTC)