Page that contains the error or needs to be improved (paste link): https://www.fatiando.org/harmonica/latest/gallery/forward/tesseroid_layer.html#sphx-glr-gallery-forward-tesseroid-layer-py
Description of the problem/suggestion:
It's subtle but the example calculates the geocentric radius of the ellipsoid and then uses that as the reference surface and uses it to calculate the topography radius. But when the layer is made, the geodetic latitudes are used instead of spherical ones. There's an assumption there that we can approximate the geocentric latitudes for geodetic ones but that's not mentioned anywhere. There are actually multiple ways that this can be approximated. We could say the reference is the mean radius and ignore the flattening, we could do what is done right now, or we could convert the topography to geocentric spherical coordinates (which would be the most accurate). But there's no place for that in the Gallery and that can be misleading (hence #654).
Page that contains the error or needs to be improved (paste link): https://www.fatiando.org/harmonica/latest/gallery/forward/tesseroid_layer.html#sphx-glr-gallery-forward-tesseroid-layer-py
Description of the problem/suggestion:
It's subtle but the example calculates the geocentric radius of the ellipsoid and then uses that as the reference surface and uses it to calculate the topography radius. But when the layer is made, the geodetic latitudes are used instead of spherical ones. There's an assumption there that we can approximate the geocentric latitudes for geodetic ones but that's not mentioned anywhere. There are actually multiple ways that this can be approximated. We could say the reference is the mean radius and ignore the flattening, we could do what is done right now, or we could convert the topography to geocentric spherical coordinates (which would be the most accurate). But there's no place for that in the Gallery and that can be misleading (hence #654).