广义相对论中所有参考系下光速是否恒定？对比狭义相对论特性

Great question—this is a super common point of confusion when moving from special relativity (SR) to general relativity (GR). Let's break it down clearly:

先回顾狭义相对论的情况 #

In SR, the core postulate is that the speed of light in vacuum is a constant $c$ in all inertial reference frames, no matter how those frames are moving relative to each other. This holds perfectly for flat, gravity-free spacetime.

广义相对论里的关键区分：局部vs全局，固有vs坐标 #

GR deals with curved spacetime (thanks to gravity), so we need to split this into two key scenarios:

• Local inertial frames (small, free-falling reference frames)
  If you're in a tiny, freely falling frame (like an astronaut in a small capsule falling toward Earth, far from any tidal effects), gravity is effectively canceled out locally. Here, the laws of physics match SR exactly—measure the speed of light in vacuum, and you'll get $c$ every time. This is a direct consequence of the equivalence principle: local gravity can be "transformed away" by choosing the right free-falling frame.

• Global non-inertial frames / large-scale gravitational fields
  If you're in a frame that's not locally inertial (say, standing stationary on Earth's surface, or observing light near a black hole from a distant point), things get different. We have to distinguish between two types of "speed":

  • Coordinate speed: This is the speed calculated using global coordinate systems (like the Schwarzschild coordinates around a black hole). Depending on the spacetime curvature and your choice of coordinates, this value can deviate from $c$. For example, a distant observer watching light approach a black hole would measure its coordinate speed as slowing down as it gets closer to the event horizon.
  • Proper (local) speed: This is the speed you measure right at the location of the light, using rulers and clocks that are at rest relative to the light's immediate surroundings. No matter where you are in curved spacetime, if you measure the speed of light locally, you'll always get $c$. This is because the local physics of spacetime is always Lorentz-invariant—curvature only affects global relationships, not local measurements.

总结 #

To wrap it up: In GR, the locally measured (proper) speed of light in vacuum is always $c$, just like in SR. But the coordinate speed of light can vary when you're looking at light from a global, non-inertial perspective or across curved spacetime. So it's not true that all reference frames (especially large, non-inertial ones) will measure the same speed of light—only local inertial frames guarantee the constant $c$ we know from SR.

内容的提问来源于stack exchange，提问作者PhyEnthusiast