New issue 1225: sz_kinetic field (probably) has the wrong sign https://bitbucket.org/yt_analysis/yt/issues/1225/sz_kinetic-field-probably-h... John ZuHone: The `"sz_kinetic"` derived field is given by this defintion: ```python def _sz_kinetic(field, data): scale = 0.88 * sigma_thompson / mh / clight vel_axis = data.get_field_parameter("axis") if vel_axis > 2: raise NeedsParameter(["axis"]) vel = data[ftype, "velocity_%s" % ({0: "x", 1: "y", 2: "z"}[vel_axis])] return scale * vel * data[ftype, "density"] ``` This, when projected, gives us a "y" parameter that is proportional to the distortion of the CMB spectrum (see, for example, Equation 6 of http://arxiv.org/abs/1304.6088. In this formula, the velocity has a positive sign if the plasma in question is moving away from the observer, and a minus sign if it is moving toward. This has to do with the interpretation of projections in yt. In, say, a projection along the z-axis, the positive z-direction is pointed at the observer, which means a projection of something proportional to the velocity will be positive if it is moving toward the observer and negative if it is moving away. This is the opposite sense of what is normally assumed by astronomers. The solution is to simply add a minus sign in the field definition. Responsible: jzuhone