Groundtrack velocity calculation update

Based on @naltobel comment on Feb. 7 tag-up meeting, the current definition of the groundtrack velocity needs to be updated.

More details are required.

changed title from Groundtrack velocity to Groundtrack velocity calculation update

Adding some details here as Livio hit the same issue in computing the velocity of the ground track.

We would expect the groundtrack_velocity attribute to correspond to the velocity in km/s of the sub-sc point on the ellipsoidal surface of the target.

The current implementation computes the velocity as:

 # Groundtrack speed
    gt_speed = np.sqrt(r**2 * ((vlon * np.cos(lat)) ** 2 + vlat**2) + vr**2)

Notice that at the end vr (radial velocity) is also added. We would suggest computing the velocity only on ground, excluding the radial (the sc-approaching) component. Hence:

 # Groundtrack speed
    gt_speed = np.sqrt(r**2 * ((vlon * np.cos(lat)) ** 2 + vlat**2))

Luca & Livio

I confirm that the previous formula was incorrect. Here is the results without the vr component vs. the Haversine distance between 2 consecutive points on the surface spaced with a 1 seconds sampling rate:

Fixed in !56 (merged) / 9f299c7d.

changed milestone to %Version 1.1

added need details label

mentioned in commit 0e053535

mentioned in commit 9f299c7d

mentioned in merge request !56 (merged)

closed with merge request !56 (merged)