# Spherical Coordinates

1. Sep 3, 2005

### bomba923

Just curious, why is $$\phi$$ calculated as the angle between the +z axis and a position vector of a point of a function, as projected onto the yz plane? Why this convention?

In polar & cylindrical, $$\theta$$ is calculated from the +x axis to the +y axis (counterclockwise) for position vectors.

*So, why not extrapolate alphabetically, to have the $$\phi$$ be the angle between the +y axis and the position vector of a point as projected onto the yz plane?

Last edited: Sep 3, 2005
2. Sep 3, 2005

### arildno

Use whatever convention you like, there is no general agreement on this issue.

3. Sep 3, 2005

### bomba923

But nowhere have I found $$\phi$$ calculated as the angle between the +y axis and the position vector of a point, projected onto the yz plane...

4. Sep 3, 2005

### amcavoy

You will see why once you start working with problems involving these. For example, take the intersection between a sphere and a cone (I posted about this before). This can be computed by a double integral, but if you use a triple integral and spherical coordinates, it becomes much simpler (thanks to HallsofIvy on this).

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook