Monday, August 17, 2015

粗粒化 redshifts The coarse grained data live on smaller screens obtained by moving the 1rst screen further into the interior of the space.


4.1 The Poisson equation for general matter distributions.
Consider a microscopic state, which after coarse graining corresponds to a given mass
distribution in space. All microscopic states that lead to the same mass distribution
belong to the same macroscopic state. The entropy for each of these state is de ned
as the number of microscopic states that
ow to the same macroscopic state.


Figure 4: The holographic screens are located at equipotential surfaces. The information on
the screens is coarse grained in the direction of decreasing values of the Newton potential .
The maximum coarse graining happens at black hole horizons, when =2c2 = 􀀀1.


The coarse grained data live on smaller screens obtained by moving the rst screen
further into the interior of the space.


The coarse grained data live on smaller screens obtained by moving the rst screen
further into the interior of the space. The information that is removed by coarse
graining is replaced by the emerged part of space between the two screens. In this way
one gets a nested or foliated description of space by having surfaces contained within
surfaces. In other words, just like in AdS/CFT, there is one emerging direction in
space that corresponds to a "coarse graining" variable, something like the cut-o scale
of the system on the screens.
4 Emergent Gravity for General Matter Distributions.

Space emerges at a macroscopic level only after coarse graining. Hence, there will be
a nite entropy associated with each matter con guration. This entropy measures the
amount of microscopic information that is invisible to the macroscopic observer. In
general, this amount will depend on the distribution of the matter. The information is
being processed by the microscopic dynamics, which looks random from a macroscopic
point of view. But to determine the force we don't need the details of the information,
nor the exact dynamics, only the amount of information given by the entropy, and the
energy that is associated with it. If the entropy changes as a function of the location
of the matter distribution, it will lead to an entropic force.
Therefore, space can not just emerge by itself. It has to be endowed by a book
keeping device that keeps track of the amount of information for a given energy distribution.
It turns out, that in a non relativistic situation this device is provided by
Newton's potential . And the resulting entropic force is called gravity.

By combining the above equations one of course again recovers F = ma as the entropic
force. But, by introducing the number of bits n associated with the particle, we
succeeded in making the identi cations more natural in terms of their scalar versus
vector character. In fact, we have eliminated ~ from the equations, which in view of
our earlier comment is a good thing.
Thus we conclude that acceleration is related to an entropy gradient. This will be
one of our main principles: inertia is a consequence of the fact that a particle in rest
will stay in rest because there are no entropy gradients. Given this fact it is natural to
introduce the Newton potential and write the acceleration as a gradient
a = 􀀀r :
This allows us to express the change in entropy in the concise way
S
n
= 􀀀kB

2c2 (3.16)
We thus reach the important conclusion that the Newton potential keeps track of
the depletion of the entropy per bit. It is therefore natural to identify it with a coarse
graining variable, like the (renormalization group) scale in AdS/CFT. Indeed, in the
next section we propose a holographic scenario for the emergence of space in which the
Newton potential precisely plays that role. This allows us to generalize our discussion
to other mass distributions and arbitrary positions in a natural and rather beautiful
way, and give additional support for the presented arguments.


4 Emergent Gravity for General Matter Distributions.
Space emerges at a macroscopic level only after coarse graining. Hence, there will be
a nite entropy associated with each matter con guration. This entropy measures the
amount of microscopic information that is invisible to the macroscopic observer. In
general, this amount will depend on the distribution of the matter. The information is
being processed by the microscopic dynamics, which looks random from a macroscopic
point of view. But to determine the force we don't need the details of the information,
nor the exact dynamics, only the amount of information given by the entropy, and the
energy that is associated with it. If the entropy changes as a function of the location
of the matter distribution, it will lead to an entropic force.
Therefore, space can not just emerge by itself. It has to be endowed by a book
keeping device that keeps track of the amount of information for a given energy distribution.
It turns out, that in a non relativistic situation this device is provided by
Newton's potential . And the resulting entropic force is called gravity.
We start from microscopic information. It is assumed to be stored on holographic
screens. Note that information has a natural inclusion property: by forgetting certain
bits, by coarse graining, one reduces the amount of information. This coarse graining
can be achieved through averaging, a block spin transformation, integrating out, or
some other renormalization group procedure. At each step one obtains a further coarse
grained version of the original microscopic data. The gravitational or closed string side
of these dualities is by many still believed to be independently de ned. But in our
view these are macroscopic theories, which by chance we already knew about before we
understood they were the dual of a microscopic theory without gravity. We can't resist
making the analogy with a situation in which we would have developed a theory for
elasticity using stress tensors in a continuous medium half a century before knowing
about atoms. We probably would have been equally resistant in accepting the obvious.
Gravity and closed strings are not much di erent, but we just have not yet got used
to the idea.


引力是“熵力”吗?(by 李淼)


格式:ppt
时间:2014-08-26 06:00
浏览:1
页数:38
收藏:0
积分:100 举报

引力是“熵力”吗?(by 李淼)

    引力是“熵力”吗李淼 中国科学院理论物理研究所 2010.01.122011年9月19日星期一本报告介绍Erik Verlinde最近的工作: On the Origin of Gravity and the Laws of Newton arXiv:1001.0785v1[hep-th]2011年9月19日星期一很久以来,一直有人怀疑万有引力不是基 本的,是一种宏观现象。 例如,Ted Jacobson在Thermodynamics of Spacetime: The Einstein Equation of State arXiv:gr-qc/9504004v2用类似黑洞热力学的办法推导了爱因斯坦 方程2011年9月19日星期一Verlinde在他的工作中指出,不仅引力本 身,惯性和质量其实也是一种宏观现象。 用文字来表达他的结果,就是: 1 1、引力是熵力。 2、加速度与熵的梯度有关,所以惯性是 无熵梯度的表现,质量与bits数成正比。 3、牛顿势是熵与bits数的比例。2011年9月19日星期一什么是熵力? 例子:虎克定律中的弹性力就是熵力。2011年9月19日星期一在微正则系综中有或热力学第一定律2011年9月19日星期一引力 Verlinde假设m2011年9月19日星期一所以,根据第一定律:利用Unruh公式得牛顿第二定律2011年9月19日星期一问题:Unruh公式是量子场论推出的,不用 如何? 答案:不用Unruh公式,但假设全息原理, 可得牛顿万有引力公式。 在球面上,假设bits数(自由度数):2011年9月19日星期一由推得代入 得2011年9月19日星期一总结: 1、基本假设 加Unruh公式 推出牛顿第二定律 2、基本假设 加全息假设推出牛顿万有引力2011年9月19日星期一问题:在熵变的基本公式中,Planck常数 出现,在Unruh公式和全息假设中,Planck 常数也出现,但牛顿第二定律和万有引力 公式是经典的,所以Planck常数相消。 我们可以用任何其他常数代替Planck常 数,结论不变,所以量子力学不是必须 的,虽然量子力学是隐含的。2011年9月19日星期一惯性和牛顿势 考虑将一个质量为m的粒子“融入”全息屏。 根据能量均分原则,有其中n是描述m需要的bits数。由于m是固定 的,T越低,需要的n越大。2011年9月19日星期一的确,在远离大质量物质M的地方,T较 低:利用基本假设 式,可得和Unruh公2011年9月19日星期一这个公式的右边是描述该粒子的每个bit所带的熵, 我们可以直观地想成每个bit的受激程度。 方程右边已与Planck常数无关。2011年9月19日星期一引入牛顿势得这个结果很重要,说明每个bit的熵与牛顿 势成正比。2011年9月19日星期一将变分符号去掉我们可以这样解释上面公式:牛顿势(绝 对值)越大的地方,bit的效率越高。对于 固定的系统,熵是固定的,所以牛顿势大 的地方,bits数少,被粗粒化得更多 (IR)。 很类似AdS/CFT中的UV/IR
    关系。2011年9月19日星期一有趣的是,量的取值范围是0到1。 在黑洞视界上,这个量最大,所以粗粒化 最厉害,或者说bits的效率最高。 在无限远处,这个量最小,bits的效率最 低,这是UV极限。2011年9月19日星期一一般的质量分布 引入牛顿势,自然就可以考虑一般的质量 分布了。我们无非要导出Poisson方程。 考虑等势面,并将等势面看成全息屏2011年9月19日星期一2011年9月19日星期一现在,取代Unruh公式,我们假设:以及全息假设:2011年9月19日星期一能量均分原则是得2011年9月19日星期一用Stokes定理,我们推出:注意,和前面导出牛顿公式不同,我们没 有用到熵变的基本假定,那里用熵变是为 了推出作用在试验粒子上的力,而不是 Poisson方程。2011年9月19日星期一最后,稍微复杂地是推导作用在试验粒子 上的力,这和前面推出牛顿万有引力公式 类似。 这里不复述。2011年9月19日星期一等效原理和Einstein方程 前面是非相对论引力的讨论,虽然出现了 光速甚至Planck常数。 要推广到一般情形,先从静态引力场开 始。在这个情况下,存在time-like Killing vector2011年9月19日星期一定义推广的牛顿势加速度的推广是2011年9月19日星期一考虑等势面,此时加速度与等势面垂直。 定义温度熵变假设为2011年9月19日星期一从热力学第一定律得熵力公式这确实是静态引力场中的正确公式。2011年9月19日星期一要获得Einstein方程,和推导Poisson方程 一样,我们需要全息原理和能量均分2011年9月19日星期一所以由于牛顿势与Killing vector有关,故2011年9月19日星期一用Stokes定理和得2011年9月19日星期一即使取任意曲面,我们只能得到和Killing vector 有关的方程。 要去掉Killing vector,我们可以利用局域的 任意坐标系中的任意Killing vector (很多 局域惯性系),这样我们就获得Einstein方 程。2011年9月19日星期一讨论 由此看来,引力确实是熵力,即非基本 的。 我想第一个问题是,引力要量子化吗? 我觉得可以量子化,如同声子要量子化一 样。2011年9月19日星期一从AdS/CFT来看,引力一边是闭弦理论, 如果引力是emergent的,那么闭弦也应该 是。 (我过去曾认为闭弦可以从非对易几何得 到,也许两者有关联)2011年9月19日星期一QCD,一些凝聚态物理系统对应于引力, 引力也应该是作为熵力出现的。 也许并不存在更加细致的全息原理,否则 我们无法解释为什么很多凝聚态系统也诱 导引力。2011年9月19日星期一最后,我们问,空间并不完全是emergent 的,我们还需要等势面,在这些面上有一 些bits。 如果我们假
    想所有空间都是emergent的, 我们需要考虑这些bits如何导出。2011年9月19日星期一另外,引力既然是熵力,为什么Eintein方 程,特别是Friedmann方程,是时间反演不 变的? 如何理解Penrose问题(宇宙初始时刻熵最 小) 如何理解暗能量,即斥力?(这里肯定涉 及引力量子化)2011年9月19日星期一Thanks!2011年9月19日星期一

No comments:

Post a Comment