快乐海洋论坛

 找回密码
 注册
查看: 8131|回复: 18

为什么洋壳的平均厚度是一个不随扩张速率变化的固定值

    [复制链接]
JixinWang 发表于 2012-3-3 23:27:06 | 显示全部楼层 |阅读模式
本帖最后由 JixinWang 于 2012-3-15 00:46 编辑

地球动力学课上杨挺老师讲述关于洋中脊的一些知识,最后罗列的问题之一——"为什么洋壳的平均厚度是一个不随扩张速率变化的固定值"。(按道理说,平均厚度就是一个相对不变的值呀,这个问题成立么?以下讨论建立在假设它成立的基础上)

Geodynamics (second edition by Turcotte and Schubert,chapter 10, page 410.) said "the uniformity of MORBs' isotopic signatures indicates that the upper mantle reservoir from which they are extracted is a well-mixed geochemical reservoir, not pristine, but is systematically depleted in incompatible elements" and "when the partial melting of the mantle occurs beneath a mid-ocean ridge the incompatible elements are systematically fractionated into the melt. Thus the basaltic oceanic crust is enriched in the incompatible elements".简要言之,MORB is depleted in the incompatible elements, but basaltic oceanic crust is enriched.

原先我的认识是洋壳表层的玄武岩是由ridge轴处喷发的岩浆形成的,而脊轴处喷发的岩浆也形成MORB,就好似等价于MORB与洋壳玄武岩等同。岩浆的喷发量与扩张速率有关,所以个人觉得“为什么洋壳的平均厚度是一个不随扩张速率变化的固定值”类似于“为何MORB是depleted而洋壳是enriched”这个问题。
可是为什么MORB亏损而玄武岩洋壳富集不相容元素呢?由Geodynamics的叙述可知,玄武岩洋壳的形成还由于脊轴处上地幔的部分熔融。所以,喷发的岩浆和部分熔融物质共同协调产生厚度较均一的洋壳。除非这部分熔融的物质不属于脊轴处喷发的岩浆之列,不然我就理解不能了。那为什么部分熔融的上地幔物质不经过喷发就能成为玄武岩洋壳的一部分呢?


今日20120314经杨老师解释书中的原话,虽然MORB指示的地幔源区是亏损的,但它依然比其源区富集呀;洋壳也是这种相对意义上的富集。
我的发问从题目错到了题干,还加入了自己的碎碎念,很不好意思;但意外得到了老师们珍贵的回复和拓展知识的信息,很是感谢。
jlinwhoi 发表于 2012-3-4 01:23:19 | 显示全部楼层
本帖最后由 jlinwhoi 于 2012-3-4 05:13 编辑

洋壳的平均厚度并非总是一个不随扩张速率变化的固定值:

(1)只有当洋中脊的全扩张速度大于2厘米/年时,洋壳的平均厚度(5-7公里)才不怎么随扩张速率变化(图1)。这是因为扩张速率越快,需要的岩浆越多,但快速中脊下的地幔岩浆产生率也高,正好平衡了。



(2)但在超慢速洋中脊(全扩展速度小于2厘米/年),岩石圈变厚,使上地幔岩浆产生率剧减,因此洋壳变得很薄(图1)。

来源:Dick, H.J.B., J. Lin and H. Schouten, An ultraslow-spreading class of ocean ridge, Nature, 426, 405-412, 2003 (下载:http://www.whoi.edu/science/GG/p ... out_Nature_2003.pdf)。

(3)热点对海洋地壳厚度也有有极大影响。北大西洋中35%地区的地壳厚度超过7公里。这种“厚地壳”的体积占北大西洋地壳总体积的49%(图2)。



来源:Wang, T., J. Lin, B.E. Tucholke, and Y.J. Chen, Crustal thickness anomalies in the North Atlantic Ocean basin from gravity analysis, Geochemistry, Geophysics, Geosystems (G3), Q0AE02, doi:10.1029/2010GC003402, 2011. (下载:http://www.whoi.edu/fileserver.do?id=102244&pt=2&p=68128).

本帖子中包含更多资源

您需要 登录 才可以下载或查看,没有帐号?注册

x
滚滚长江都是水 发表于 2012-3-4 12:10:03 | 显示全部楼层
I should track more papers, but I am presently lazy to do so. So I throw out the question here: what is the buoyant and passive flow model? Is that related to density of magma?
滚滚长江都是水 发表于 2012-3-4 12:21:10 | 显示全部楼层
MORB is depleted but crust is enriched.
I am not a geochemist, but I can try to explain.
Crust is formed from partial melting at ridges. During the melting, incompatible elements like to stay in the melt (liquid) and later enrich the crust when the melt is cooled down to generate oceanic crust. Therefore, MORB, the leftover of the partial melting is depleted.
jlinwhoi 发表于 2012-3-4 17:36:47 | 显示全部楼层
本帖最后由 jlinwhoi 于 2012-3-4 17:38 编辑
what is the buoyant and passive flow model? Is that related to density of magma?


"Passive" refers to mantle upwelling driven by divergence of plates at mid-ocean ridges.  

Examples of buoyancy in "buoyant flow" models include density changes associated with (1) melt retention in the mantle, i.e., melt pockets trapped in mantle matrix; (2) mantle depletion due to melting, i.e., changes in mantle density after melt extraction; and (3) mantle thermal buoyancy.
 楼主| JixinWang 发表于 2012-3-4 18:59:57 | 显示全部楼层
回复 4# 滚滚长江都是水
Though Geochemistry is my specialty, my now available knowledge is not systematic and so poor.
Your explanation gives me the awareness that the mantle partial melt cools down to create ocean crust first, then later, the liquid leftover becomes MORB. After rethinking, here is my question: Since incompatible elements try to stay in melt during partial melting, they should choose to stay in the liquid leftover during melt cooling. But they don't, why?
So hard to meet someone to talk about it, thank you!
 楼主| JixinWang 发表于 2012-3-4 19:15:02 | 显示全部楼层
回复 2# jlinwhoi

I got to know the three papers and infos recently, and have read parts of them though not throughoutly. I know the crust thickness varies a lot in slow-spreading ridges. But I just got trapped by the question shown on PPT. After seeing your explanation, I tried to find the PTT to recall how it present the question. However, my mobile HDD crashed. I will get the PPT source later on.
Thanks for your adequate explanations!
 楼主| JixinWang 发表于 2012-3-4 19:33:17 | 显示全部楼层
回复 3# 滚滚长江都是水

and the paper visualize the passive flow and buoyant flow. The former corresponds to the sheet-like 2D upwelling beneath fast-spreading ridges, while buoyant flow corresponds to the 3D mantle upwelling beneath MAR.

   

Personally, I think these two may correlate with the passive rifting and active rifting to some extent.

本帖子中包含更多资源

您需要 登录 才可以下载或查看,没有帐号?注册

x
jlinwhoi 发表于 2012-3-4 22:43:10 | 显示全部楼层
本帖最后由 jlinwhoi 于 2012-3-4 23:30 编辑
Since incompatible elements try to stay in melt during partial melting, they should choose to stay in the liquid leftover during melt cooling. But they don't, why?


Here are the key assumptions about melting beneath mid-ocean ridges:

Step 1: Melt generation:

- While mantle upwells, tiny pockets of melts first appear at the triple junctions between grains of the olivine minerals in the mantle.

- As the degree of partial melting increases, the volume of each melt pocket also increases. Eventually, adjacent melt pockets might link, forming a network of melts inside the mantle matrix.

Step 2: Melt migration:

- Since melts are lighter than the surrounding mantle minerals, the melts will ascend under their own buoyancy towards the surface.  Eventually some (or most) of the melts will migrate towards melt a melt aggregation zone beneath the ridge axis.  

- Why and how melts migrate towards the ridge axis are still debatable. Proposed mechanisms include the following:

a) Melts flow along the base of a rigid lid towards the ridge axis (Sparks and Parmentier, 1993);

b) Melts are forced to move towards the ridge axis under the poro-elastic pressure gradients inside a deforming mantle, corresponding to the mantle corner flow (Spiegelman & McKenzie 1987);

c) Melts flow along channels in the mantle caused by anisotropy of the deforming mantle olivine crystals (Phipps Morgan, 1987).

Step 3: Formation of oceanic crust:

Melt aggregation could be quite different between fast- and slow-spreading ridges.  There is a steady-state, thin magma lens beneath much of the fast-spreading ridges, from which the melts will reside.  Eventually light melts will erupt to form layer 2a (basalts) and layer 2b (dikes), while the heavy residuals will sink to form layer 3 (gabbros).

Melt aggregation beneath slow-spreading ridges is much more complex, intermittent, and discontinuous (e.g., Cannat, 1993, 1996).

A significant amount of melts probably never make to the ridge axis and thus are frozen in the mantle.  Also increasing lines of evidence suggest that melt aggregation zones might exist at a much wider region beneath ridges than previously thought.
jlinwhoi 发表于 2012-3-4 23:20:02 | 显示全部楼层
本帖最后由 jlinwhoi 于 2012-3-4 23:32 编辑

Notes on density changes in the depleted residual mantle:

Melt extraction reduces the density of melt-depleted mantle since the residual mantle has a lower Fe/Mg ratio (atomic weight ratio of 55.847/24.305) (e.g., Oxburgh and Parmentier, 1977). Thus the melt-depleted mantle is actually lighter than the original mantle.

Notes on chemistry of melts:

The key factor to keep in mind is the partition coefficients (C) of various elements between the melt and residual mantle during partial melting.  An example of how to integrate the above physical and chemical processes into a self-consistent model of the East Pacific Rise is given by Gregg et al. (JGR, 2009) http://www.whoi.edu/science/GG/p ... _Grove_JGR_2009.pdf
 楼主| JixinWang 发表于 2012-3-5 00:15:12 | 显示全部楼层
回复 10# jlinwhoi

Sir, I am shocked by your response. It always surprises me that you have worked on what I exactly want to know and decided to do. Thanks a million for sparing time answering it. I shall keep on reading and thinking more.
 楼主| JixinWang 发表于 2012-3-5 09:20:03 | 显示全部楼层

RE: 为什么洋壳的平均厚度是一个不随扩张速率变化的固定值

本帖最后由 JixinWang 于 2012-3-5 12:44 编辑

(NOTE: Jixin - Sorry that I mistakenly wrote over your posting of Prof. John Chen's plots, please re-post John's figures here.  Jian)

Ok.Here comes John's figures:

and his conclusion: the average crustal thickness along a ridge segment (apart from regions of hot spot-ridge upwelling interactions) appears to be roughly 6 km regardless of spreading rate.

本帖子中包含更多资源

您需要 登录 才可以下载或查看,没有帐号?注册

x
jlinwhoi 发表于 2012-3-5 11:03:14 | 显示全部楼层
本帖最后由 jlinwhoi 于 2012-3-5 11:06 编辑
Compared with ... more data were compiled and the average thickness seems thinner around ultraslow-spreading ridges where some crust thickness reaches zero.


That's correct.  When Prof. John Chen wrote the GRL paper in 1992, there was little work on ultraslow spreading ridges.

In 2000 and 2001, we had two US NSF-funded cruises to an oblique super-segment of the ultraslow Southwest Indian Ridge, were we found very thin or no crust.  At the similar time, German and US scientists had a couple of cruises to the ultraslow Gakkel Ridge in the Arctic ocean, where the crust was also found to be thin in many places.  Our 2003 Nature paper includes these new data from ultraslow ridges.
hbsong 发表于 2012-3-6 11:47:10 | 显示全部楼层
林教授最近主要用重力方法计算了全球大洋洋壳厚度,可能是最新的成果吧。
 楼主| JixinWang 发表于 2012-3-6 23:16:53 | 显示全部楼层
回复 14# hbsong

嗯,北大西洋的。
hbsong 发表于 2012-3-7 07:26:03 | 显示全部楼层
回复 15# JixinWang


    他有全球的,可能还没发表。
 楼主| JixinWang 发表于 2012-3-7 12:29:56 | 显示全部楼层
回复 16# hbsong

那得多大的工作量。。对于结果,我竟然怀着期待和不安的心情。
hbsong 发表于 2012-3-7 13:32:36 | 显示全部楼层
回复 17# JixinWang


    林教授很勤奋的,不象我,懒得很。
jlinwhoi 发表于 2012-3-7 16:21:36 | 显示全部楼层
回复 14# hbsong

是的,我们刚完成了用重力方法计算全球大洋洋壳厚度的工作,正在写文章。本帖子里显示的是2011年已经发表的北大西洋局部图。
您需要登录后才可以回帖 登录 | 注册

本版积分规则

QQ|小黑屋|手机版|Archiver|快乐海洋论坛 ( 浙ICP备11010086号 )

GMT+8, 2019-7-21 07:52

Powered by Discuz! X3.4 Licensed

© 2001-2017 Comsenz Inc.

快速回复 返回顶部 返回列表