Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

What are Rosen Coordinates

  1. Mar 12, 2014 #1
    I see a number of gravitational wave analytic solutions with the metric given in terms of Rosen coordinates. I have no idea what these coordinates are. How do I perform a coordinate transformation from Rosen coordinates to traditional (t,x,y,z) Euclidean\Cartesian coordinates? Also, is there a difference between Rosen coordinates and Eisenstein-Rosen coordinates?

  2. jcsd
  3. Mar 13, 2014 #2


    User Avatar
    Science Advisor

    A coordinate system for the description of gravitational plane waves:

    ds2 = 2 du dv + gij(u) dyi dyj.

    As opposed to Brinkmann coordinates for the same wave:

    ds2 = H(u) du2 + 2 du dv + dx2 + dy2.

    Here's a reference
    that discusses both, and how to convert one to the other.
  4. Mar 14, 2014 #3
    That doesn't directly answer my question. Also, I am not familiar with Brinkmann coordinates. If we suppose the line element in Rosen coordinates in flat space-time is:

    ds2=2 du dv+dy12+dy22

    Then we can deduce the possible coordinate mapping to Cartesian coordinates of

    t = (u-v)/√2
    x = y1
    y = y2
    z = (u+v)/√2,

    Assuming a <-1,1,1,1> metric signature

    Is this the correct mapping from Rosen coordinates to Cartesian coordinates?
  5. Mar 15, 2014 #4


    User Avatar
    Science Advisor

    Sorry. What IS your question? You say you want to know what Rosen coordinates "are", but that question has no answer. Coordinates in a curved spacetime do not always have a simple interpretation.

    If you're interested in gravitational waves, you should learn about Brinkmann coordinates. Rosen coordinates have several drawbacks. One: they are not unique. Two: they can develop coordinate singularities (caustics). That's why they are not generally used.

    That's one possibility. Here's another instance of a plane wave in Rosen coordinates that is also flat:

    ds2 = 2 du dv + u2(dy12 + dy22)

    In a flat spacetime, (u, v, y1, y2) are called light-cone coordinates, and (t, x, y, z) are called Minkowski coordinates. The terms Euclidean and Cartesian do not apply!
  6. Mar 15, 2014 #5
    My problem is I have a formula for the metric tensor in Rosen coordinates and I would like the metric tensor in Minkowski spacetime. To go from one to the other I need the coordinate mapping, thus, how do I transform the coordinates from Rosen coordinates to Minkowski Coordinates?
  7. Mar 16, 2014 #6


    User Avatar
    Science Advisor

    Can't be done. A gravitational wave is a curved spacetime, and Minkowski coordinates exist only in flat space. You can't turn a curved spacetime into a flat spacetime just by changing the coordinates. A gravitational wave is not just Minkowski space written in some weird set of coordinates.

    You can certainly do the z = (u + v)/√2, t = (u - v)/√2 thing if you like, and that will make at least part of the metric look more familiar, but there will be terms left over.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook