Wikipedia:Reference desk/Archives/Science/2015 May 16

Science desk
< May 15 << Apr | May | Jun >> May 17 >
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 16 edit

What flower is this? edit

You may need to enlarge it [1]. Thank you. 69.22.242.15 (talk) 06:59, 16 May 2015 (UTC)[reply]

I think it might be a Kalanchoe (different variety from the one in our article, rather like this red one), but I'm not an expert on house plants. Dbfirs 07:42, 16 May 2015 (UTC)[reply]
Yup, that's almost certainly it. Not only does it just 'look right', we can see the that the petal number matches, the leaves are succulent, the margins are indented, and the inflorescence is a showy raceme, held well above the phylloclades. That and the OP's picture looks like a plant straight from the florist, where yellow kalanchoe are a popular option. SemanticMantis (talk) 16:11, 16 May 2015 (UTC)[reply]

Journal publication and peer-reviewer anonymity edit

I've read articles on two controversial science topics, unrelated to each other except for a link to creationism, and both of which somehow involve anonymity or the use of pseudonymous: the Sternberg peer review controversy on the publication of a journal article on intelligent design, and Answers in Genesis' Answers Research Journal. I have a number of questions regarding the two topics (just to make it clear: I am perfectly aware that intelligent design is classified as pseudoscience and personally I am against it for various reasons, including but not limited to apparent academic dishonesty by its proponents (quote mining, among others), and I am also aware that the ARJ is a discredited journal which promotes a creationist agenda it is condemned by the mainstream scientific community; I wouldn't want to support such a journal).

1. The ARJ apparently allows authors to submit articles under a pseudonym (though I can't seem to find any statistics on how many authors actually do so). RationalWiki suggests that many universities and organizations prohibit publishing academic papers under a pseudonym. Since I am not a scientist, I am asking: is the practice of publishing journal articles anonymously or under a pseudonym strongly frowned upon by the scientific community?

2. For the Sternberg controversy, Richard Sternberg claims that his paper, published in the Proceedings of the Biological Society of Washington, was peer-reviewed by three biologists, all of whom requested anonymity (although according to the article, doubts have been raised if the three were actual biologists). According to the article, the Proceedings published lists of its peer-reviewers in the years prior to 2004, which is the year the controversial paper was published; no peer-reviewer list was published for 2004. I have two questions regarding this:
A. After 2004, did the Proceedings again publish lists of peer-reviewers for succeeding years, or did their publication of lists of peer-reviewers cease after 2003? The aforementioned article does not answer this question.
B. Is keeping a list of peer-reviewers for a journal confidential a standard practice, or is otherwise commonplace, or is it considered an unusual practice among academic journals? Is it also common for journal peer-reviewers to request that their names not be published?

Narutolovehinata5 tccsdnew 08:51, 16 May 2015 (UTC)[reply]

Post-script: I've read the article on peer review, which discusses "anonymous" or "blind" peer-review; however, the article only discusses anonymity between authors and reviewers (meaning, for example, in a double-blind peer review, the authors do not know the reviewers' identities, and vice-versa), and it does not discuss the publication of lists of peer-reviewers after peer-reviews of articles are completed, although the article does suggest that reviewers, for various reasons, want anonymity at least during the review process itself. Narutolovehinata5 tccsdnew 09:04, 16 May 2015 (UTC)[reply]
Speaking personally, I don't have a strong opinion against anonymous / pseudonymous publications as long as the actual research was described accurately and in enough detail that it could be repeated. If I knew a work was being published under a pseudonym, I'd probably initially look at it more skeptically because one wouldn't have a foundation of trust in a known author; however, if the content was good it wouldn't really bother me. That said, I can't think of any modern scientists who used a pseudonym professionally. It would be extremely unusual, especially since name recognition is such a valuable asset for a scientist.
In my experience, most journals do not publish lists of peer reviewers. It is not uncommon for reviewers to request anonymity, and that anonymity generally continues indefinitely and is not limited to the publication / rejection of the paper. Anonymity is especially common with negative reviews because people worry about reprisals (i.e. you gave me a bad review in the past, so I'll give you a bad review in the future even if you don't deserve it). Dragons flight (talk) 09:22, 16 May 2015 (UTC)[reply]
If I were a "scientist" touting creationism, I'd want to stay anonymous too. ←Baseball Bugs What's up, Doc? carrots→ 09:37, 16 May 2015 (UTC)[reply]
  • A theoretical paper could probably be published under a pseudonym, although it's hard to imagine people wanting to do so in the modern world, as one of the major benefits of publishing a paper is the credit you get for it. There was a time, centuries ago, when many mathematics papers were published pseudonymously. I doubt that any respectable journal would accept an experimental paper under a pseudonym, since there would be no way of verifying that the work was really done as described. Looie496 (talk) 12:26, 16 May 2015 (UTC)[reply]
There is a phrase amongst scientists "Publish or Perish". A scientist's career is so completely dependent on their publication list, I struggle to think of circumstances in which they would want to submit with a pseudonym. Women scientists often struggle to maintain a name constancy on their publication list if they change their name through marriage/divorce. Furthermore, universities (here in the UK) go through an assessment every 5(?) years. The University funding is dependent on this and part of the assessment includes the publications of their staff, so correct attribution is needed.DrChrissy (talk) 13:23, 16 May 2015 (UTC)[reply]
Regarding reviewers, I have encounted journal which publish reviewers names with a thanks, but only at the end of the year so that any one reviewer can not be linked to a paper.

So if the practice of publishing papers under a pseudonym is extremely uncommon, especially in modern times, then why does the Answers Research Journal (at least in theory) allow it? Is it because it is a journal which engages in pseudoscience? And is it considered academically dishonest to publish work under a pseudonym? Would publishing papers under a pseudonym raise questions on the author's qualifications and credentials? And is the practice common or at least allowed in other "controversial" journals? Finally, do other, more prestigious or widely accepted journals (i.e. not pseudoscientific and not discredited) have a stand on papers published under a pseudonym or anonymously? Narutolovehinata5 tccsdnew 13:38, 16 May 2015 (UTC)[reply]

And is it considered academically dishonest to publish work under a pseudonym? Depends upon the circumstances and intent. See for example this paper on Anonymity in Science published by a well-known blogger under his pseudonym; no one IMO would consider this dishonest. On the other hand see this paper on a scientist publishing using a pseudonym on arXiv in order to criticize other scientists' work (fwiw, both the pseudonymous paper and the paper exposing it were withdrawn; the former for violating arxiv policies and the latter for "legal reasons"). Abecedare (talk) 15:15, 16 May 2015 (UTC)[reply]
Pure math of course isn't science, but for publishing reasons it's close enough. Bourbaki is probably one of the most esteemed pseudonymous authors, though pseudonymity is indeed very uncommon in mainstream science journals today. There are also a few occasions where a person will add their dog or cat to the author list, but that's a little different. SemanticMantis (talk) 16:15, 16 May 2015 (UTC)[reply]
Are you aware of other such "authors" than Galadriel Mirkwood? --NorwegianBlue talk 17:52, 20 May 2015 (UTC)[reply]

About the inhalation of objects edit

Why when a person Inhale object is more likely to become trapped in the right main bronchus? What is the explanation for that? 192.117.186.253 (talk) 11:56, 16 May 2015 (UTC)[reply]

Who says it is? ←Baseball Bugs What's up, Doc? carrots→ 13:18, 16 May 2015 (UTC)[reply]
This lot do. "[B]ronchial foreign bodies tend to be lodged in the right main bronchus because of its lesser angle of convergence compared with the left bronchus". Tevildo (talk) 14:02, 16 May 2015 (UTC)[reply]
Our bronchus article addresses this specifically, in the section on "aspiration". The right mainstem bronchus is a straighter path than the left. -- Scray (talk) 00:56, 18 May 2015 (UTC)[reply]

Making metal glow blue edit

Now in my random research about Sting I found Weta $8k version: https://www.wetanz.com/sting-fine-art-limited-edition-the-master-swordsmith-s-collection/ and $130 Sting that can glow blue version: http://www.amazon.com/Hobbit-Sting-Sword-Illuminating-Baggins/dp/B009S8EXSK

I found it kinda sad that $130 version can glow while $8k version cannot. But of course, one is plastic while the other one is metal. Now, is it possible to make metal things glow in blue then? 139.193.24.157 (talk) 16:39, 16 May 2015 (UTC)[reply]

No.--Aspro (talk) 16:48, 16 May 2015 (UTC)[reply]
Actually, you can make metal glow blue. You better be viewing it through several inches of lead glass though... WegianWarrior (talk) 17:01, 16 May 2015 (UTC)[reply]
By blasting relativistic electrons through the metal. The lead glass is for protection both against the electrons and against the invisible ionizing radiation in the Cherenkov spectrum. Robert McClenon (talk) 17:29, 16 May 2015 (UTC)[reply]
Vanadium has a bit of a bluish tint, and there may be other metallic alloys which are bluish, such as blue gold. Of course, shining a blue light on it would help, and/or it could have a coating that reflects blue light. Best yet, how about coating it with blue glow-in-the-dark paint: [2]. StuRat (talk) 17:17, 16 May 2015 (UTC)[reply]
Vanadium does not 'glow' it reflects. WegianWarrior is talking about Cherenkov radiation which stimulates (scintillates) the medium 'around' the object. ZnCl2 & ZnBr2 is better than lead glass so the latter is not used anymore.--Aspro (talk) 17:33, 16 May 2015 (UTC)[reply]
What is meant by the medium around the object? If one blasts relativistic electrons through metal at a speed exceeding the local speed of light in the metal, the metallic object is the medium. Cherenkov radiation is normally observed in a liquid, but it could be observed in a solid. Robert McClenon (talk) 22:03, 16 May 2015 (UTC)[reply]
"Glow" is a bit vague. In a dark room a light reflecting off a single object will appear to make it glow, even though the light does not originate at the object. StuRat (talk) 21:22, 16 May 2015 (UTC)[reply]
Cadmium and osmium are a lot bluer, but are unfortunately poisonous. Although you might well want that in your sword, but then only osmium is suitable, because a cadmium sword is kind of like attacking people with plaster (cadmium and gypsum are both around Mohs 2.0).
(Oh, and to anyone who thinks of that: no, you can't simply heat a metal to blue-hot to answer this question, because it'll have melted way before that – even close-to-melting tungsten doesn't even reach white hot, as you can tell from incandescent light bulbs. It would certainly be a magnificent sword at killing things, but the trouble is that the intense heat produced may not be too concerned with distinguishing between you and your opponent when it comes to killing through vaporization.) Double sharp (talk) 16:23, 17 May 2015 (UTC)[reply]
There's a sort of logic issue here in that a "metallic" object is expected to reflect the light that falls on it perfectly, which rules out glowing blue. You can of course have metal compounds that appear clear (I'm thinking of rhenium carbide, which AFAIR is transparent, very hard, very strong, very expensive, melts at something on the order of magnitude of the temperature of the surface of the sun, and so in all ways but the glow seemingly well suited for a legendary magic blade). But a clear metal isn't quite a metal, though there's some distinction between what is conductive and what is transparent, despite the intro physics explanation I recall. Anyway, if you have a transparent substance of any sort you can dope it with something that will glow blue, such as by fluorescence or as a light emitting diode. (I'm not sure you can make an entire metal blade into a sort of LED with clever design and a hidden source of electric potential, but I'm not sure you can't. The omnipotence of ignorance... :) ) Wnt (talk) 18:14, 16 May 2015 (UTC)[reply]
For many years, Apple computers have been made out of aluminum, and some of those machines have LED indicator lights that glow through the aluminum case. Here is a news article from many years ago: Granted Patent: Invisible, Light-Transmissive Display System, and here is a patent granted: Invisible, light-transmissive display system, which may protect some of the methods and designs used by Apple to make a metal surface that glows. In one embodiment of this invention, holes that are too microscopic to be resolved by an unaided human eye may perforate the metal surface, allowing light transmission through the metal: an illuminant on the reverse (interior) side of the metal can cause the outer surface to appear to glow.
I woud posit two statements (speaking as a private individual, and not as a legal representative of any entity):
  • First, a device so constructed after the fashion of Apple's aluminum computers would probably be unsuitable for use as a sword. Aluminum is a strong metal, but when so perforated, I would doubt its structural integrity meets the needs of a sword, which may be used to strike opponents. (I have struck many opponents with an aluminum computer, and while I have never damaged the perforated aluminum indicator illuminants, I have caused considerable cosmetic damage to the computer and to other objects).
  • Secondly, the use of an illuminant in a sword very probably still constitutes a "method" for "display" - especially if the purposes of this glow are to represent information displayed to the user for the purposes of alerting the user to the presence of Orcs. Unauthorized use of this described method for display would very probably infringe on intellectual property rights held by Apple and others.
Nimur (talk) 16:16, 17 May 2015 (UTC)[reply]
So that's why the Orcs keep attacking! They simply care deeply about the enforcement of intellectual property, as one would expect they would. Wnt (talk) 18:38, 17 May 2015 (UTC)[reply]

A bluish shade of yellow... edit

Factoid I heard, with regards to demonstrating the limitations of the human brain...

Imagine a colour that's blueish-yellow (not green, or a gradient between blue and yellow). Can't do it? That's because the human brain didn't evolve to perceive this colour and thus it's literally physically impossible to imagine what it looks like, even though it exists.

Is this true? I know we could get into discussions as to whether something exists at all if there is no-one to perceive its existence, but aside from that... --Kurt Shaped Box (talk) 19:47, 16 May 2015 (UTC)[reply]

Isn't that because the concept can be described but doesn't actually exist in the electromagnetic spectrum? Different people perceive colours in different ways, but we are socialised to use the same words for the same combinations of frequencies of light. Dbfirs 19:55, 16 May 2015 (UTC)[reply]
You can have either combination of blue and yellow light frequencies, or just the green frequency. It's true that we lack the capacity to distinguish between these two cases. Not having the cone cells to tell the difference in turn means our brains aren't designed to tell the difference, either. StuRat (talk) 20:51, 16 May 2015 (UTC)[reply]
... but yellow and blue don't make green additively -- only when subtracting colours using pigments. Our cones can tell the difference because we have green receptors. When we look at yellow, the red receptors are also stimulated. Dbfirs 21:05, 16 May 2015 (UTC)[reply]
Yes, I should have made my comment more generic: "There are combinations of two frequencies of light which we find indistinguishable from a single frequency". Red plus green, versus yellow alone, would be a better example. StuRat (talk) 21:18, 16 May 2015 (UTC)[reply]
I wonder if the reason for the strange effects with yellow and blue is that they make light grey additively, but each eye is seeing a strong separate colour, so the additive outputs in the brain don't give a strong sense of colour but the separate signals do. Dbfirs 21:47, 16 May 2015 (UTC)[reply]
I'm pretty sure there's nowhere in the visual system where inputs from the two eyes are combined in linear LMS/XYZ/RGB space, so the brain has no idea that the combination "should" be grey. The same odd effect happens with non-opponent colors, such as yellow and red. -- BenRG (talk) 22:33, 16 May 2015 (UTC)[reply]
 
Yellow and blue can make green additively. this yellow and this blue mixed equally make this green. I could make a more saturated green if I wasn't limited to the sRGB gamut.
"Our cones can tell the difference because we have green receptors. When we look at yellow, the red receptors are also stimulated" is wrong. The L and M cones are both very sensitive to green, and both retain significant sensitivity well into the red and blue, as shown in the diagram to the right. There are no red and green cones. -- BenRG (talk) 22:26, 16 May 2015 (UTC)[reply]
Yes, fair criticism, though you had to choose those shades carefully to get the effect. The brain does combine inputs from the cones, but not in a simple way, and viewing squares is different from viewing through filters. I was using what is probably an unhelpful shortcut in describing them as red and green cones. Red stimulates L cones much more than it does M cones, and M cones have their peak sensitivity at green (534–555 nm), but I agree that there's a lot more going on than simple colour addition. Dbfirs 08:12, 17 May 2015 (UTC)[reply]
μηδείς (talk) 19:57, 16 May 2015 (UTC)[reply]
Thanks for that link, Medeis. When I try to see "blueish-yellow", the colour keeps alternating between blue and yellow with just an occasional glimpse of what looks like a muddy mix on a badly-aligned printer. I guess other people see something different. A simple experiment that I thought up fifty years ago (though I'm sure I wasn't the first to think of it) was to look through a good red filter at one eye and a good green filter at the other. As expected, I saw in yellow, proving that it is the brain that combines and interprets colours. Dbfirs 20:49, 16 May 2015 (UTC)[reply]
I have also been unable to see anything special, but I think it has to do with computer settings and personal genetics and luck. I have also not been able to resolve those 3D images from dots that were popular in the 90's. Oddly, I got a perfect score on the expensive copyrighted colorblindness tests, getting all 30 or 40 whatever plates correct, which my Bio 101 proff said must have been cheating. She relented from failing me when all my lab mates swore I got them right. I have noticed over the years that things through my right eye seem bluer, compared to my left eye, which seem oranger. μηδείς (talk) 21:17, 16 May 2015 (UTC)[reply]
Opponent process is relevant. (Impossible color also links it.) -- BenRG (talk) 22:26, 16 May 2015 (UTC)[reply]
You might find THIS ARTICLE (which I wrote, over on Medium.com) relevant - although I'm not referring to the exact same situation, the principles are the same. SteveBaker (talk) 05:41, 17 May 2015 (UTC)[reply]
Steve, you've shown over and over on this desk, for years, that you don't understand color vision. You make the same mistakes every time, even though people correct you every time. Now you're writing articles about your misconceptions for other web sites? "It's interesting that we've given yellow its own name - separate from red and green — but that's just a cultural thing. In a physiological sense, yellow is just "reddish-green" or maybe "greenish-red"."—this is nonsense. Read opponent process. Yellow is one of the four psychological primary hues. It's not a "cultural thing". Look at the peaks of the cone response curves in the diagram above. Notice that none of them are anywhere near red. "Over on the right hand side — about where the color violet is, our red-sensitive cone cells are actually slightly increasing in sensitivity."—no, they are not. Look at the diagram above, or at the experimental data. I don't know where you got your plot but it's wrong. -- BenRG (talk) 06:53, 17 May 2015 (UTC)[reply]
@BenRG: @SteveBaker: The core of this controversy is the responsivity curve for the L-type (long wavelength) cone cell. I do see a high-wavelength peak in the human cone action spectra listed at [3]. Our Wikipedia figure File:Cones SMJ2 E.svg contains the disturbing word "simplified" in the explanation; the first reference [4] is not trivial to go through. These are experimental measurements, made in a certain way, and could be made in other ways. Please don't make this a personal conflict because it is a genuine, interesting scientific issue and it won't be easy to get to the bottom of it, but the very essence of violet awaits us as a reward. Wnt (talk) 12:31, 19 May 2015 (UTC)[reply]
Even without tracking down the source of this plot you should be suspicious of it, because the L and M cone curves are cut off at 400nm despite being still at >30% of peak sensitivity. They've still got a large signal, so why not continue the curve? Despite the lack of error bars, the end of the curve must actually be borderline unplottable. There's also the problem that the S cone is shown as being highly sensitive well into the region labeled UV. You shouldn't look at this plot and think, "yeah, that looks just as plausible as the other one".
The origin of the plot seems to be Fig. 2 of J. Physiol. (1980), 298, pp. 501-511 (free download). They measured the cones in vitro. Fig. 3 shows their data adjusted for frequency- and angle-dependent absorption in the eye, and that seems to be consistent with the diagram earlier in this thread. It's that adjusted curve that's relevant for ordinary color vision.
In any case, I don't know why you think this is the core of the "controversy". I also pointed out that Steve thinks that yellow is not a primary color; that's also wrong, and for totally unrelated reasons. The core of the "controversy" is that Steve thinks RGB and psychological color are the same thing. I didn't understand the difference either a decade ago, but I've since learned more about color vision. Steve never has, despite people (not just me) trying to explain it to him in previous Ref Desk threads. I was blunt in my reply this time because he hasn't responded to polite correction in the past, though I doubt it will help. -- BenRG (talk) 21:17, 19 May 2015 (UTC)[reply]
BenRG: It's very simple. I have also studied this issue in mind-numbing detail - and in past jobs, I've actually performed experiments. I have concluded that you are incorrect. However, I'm not going to engage with you in some major bust-up of an argument. I will continue to tell it as it is - you can do what the heck you like - I don't care...I truly don't...I'm simply not going to engage in an argument with you as you seem to desire - so please don't expect me to respond to this, or future demands. SteveBaker (talk) 00:48, 20 May 2015 (UTC)[reply]
If you've performed experiments that overthrow the opponent-color theory or show that Stockman and Sharpe's L cone sensitivity curve is wrong at short wavelengths, you should publish in a peer-reviewed journal. Wikipedia's policy is to stick to published results.
Whether you write a paper or just keep promoting your new theory in web forums, people are going to tell you you're wrong. That's a thing that happens when you contradict established science. You need to be prepared to defend your position. Scientists have to do that. -- BenRG (talk) 08:05, 20 May 2015 (UTC)[reply]
@BenRG: I'm a bit confused by your response in that the Wikipedia figure is one lacking a tail on the L-type absorption curve, though of course it could be both a bad figure and be right on this point. Anyway, I looked up Stockman and Sharpe, finding [5]. Now normally I dig into biology papers readily, but yikes, this is math. But in Figure 11a, they show a curve for the L-type receptor with an apparent rise at the highest wavelengths. True, it is a really tiny difference, but then again all the L- and M- type receptor differences are tiny; it's positively amazing anyone is not red-green color blind when you see what we have to work with. They say "As Fig. 11a makes clear, MSP is of little use in defining cone spectral sensitivities except close to the photopigment [lambda]max. The large discrepancies between MSP and other estimates of cone spectral sensitivities arise because of the small signal to noise ratio of the MSP measurements." However, I don't see noisy numbers on that chart in our region of interest; it misses only the weakest parts of the spectrum for each.
One reason why I need more convincing to believe that it is experimental error is that apparently there may be some kind of chlorophyll accessory pigment system at work. For the truly bizarre, a recent loony news report I ran across [6] is based on more serious work [7] and indeed can be tracked to a publication [8] in which exogenously administered chlorophyll ends up specifically accumulating in photoreceptor outer segments and triggering retinal activity! So something like diet, or the exact method of measurement, might turn out to be of some significance.
If you want the download on why I need you and Steve to work together, the, ah, logic works like: the philosopher's stone is what converts people to the pursuit of high philosophy, and base matter into gold... which includes the conversion of the light that bounces from gold into a gold color; you can tell even in films like "The Hobbit" that making an image that looks "golden" is extremely difficult, and when I look close there is always some little bit of purple involved, just as atomized gold is purple, and purple hematoxylin solution accumulates liquid gold looking sediment on top; which brings us to the need to understand what is purple. And according to the "takes gold to make gold" principle of alchemy, we need the two of you to make peace in order that together you can find a way to understand the tail of the L-receptor so that we can understand what makes the color purple, convert that into philosophical terms, and use it as the blueprint to create a perfect society of peace, freedom, and enlightenment. :) Wnt (talk) 22:51, 20 May 2015 (UTC)[reply]
@Wnt: I don't know what you mean when you say the Wikipedia image lacks a tail, if you're talking about File:Cone-fundamentals-with-srgb-spectrum.svg. All three curves are plotted from 390 to 710 nm.
Fig. 11a of Stockman and Sharpe has the same origin and meaning as the figures you linked last time. In a normal human eye, that tiny upward slope is swamped by the large downward slope of the lens absorption curve. Without the lens, people can see ultraviolet light (see Aphakia).
As far as I know, the reason the short-wavelength end of the spectrum has some psychological red in it is simply that the S cones contribute to the R-G opponent calculation on the red side, as shown in Fig. 2 of Hurvich and Jameson (1957). There's nothing mysterious about it, unless you think L = red. -- BenRG (talk) 08:45, 21 May 2015 (UTC)[reply]

What's the opposite of "mellow" in the context of soil? edit

Mellow soil is soil that is crumbly and easily worked. But is there a catch-all term for the opposite of this - soil that is either too compacted or too loose and silty? As far as I can tell, "unmellow" and "non-mellow" are hardly ever used, but it seems like there ought to be something... -Elmer Clark (talk) 23:27, 16 May 2015 (UTC)[reply]

The opposite of the word "mellow" in general could be "harsh". ←Baseball Bugs What's up, Doc? carrots→ 00:13, 17 May 2015 (UTC)[reply]
For soil that is difficult to work, we have hardscrabble. StuRat (talk) 03:45, 17 May 2015 (UTC)[reply]
Thanks to both of you, but I'm more looking for a term that you might find in a textbook or academic paper. -Elmer Clark (talk) 05:20, 17 May 2015 (UTC)[reply]
Then you wouldn't use "mellow" in that circumstance. The technical term is "friable". I think the term you're after is "compacted". --TammyMoet (talk) 13:18, 17 May 2015 (UTC)[reply]
Google does find some uses of "mellow soil" in technical/academic contexts. This is from a soil survey , by the United States Department of Agriculture. The examples I have found do all seem rather old - early 20th century. DuncanHill (talk) 21:32, 19 May 2015 (UTC)[reply]