Closed AlecThomson closed 4 years ago
In makecube.py, the smoothing kernel is currently:
g = Gaussian2DKernel(hpbw / (2. * np.sqrt(2. * np.log(2.))) / grid) data = convolve(data, g, boundary='extend')
This was written for absolutely calibrated data (i.e. Kelvin-scale). As the data are in Jy/Beam, this normalization/kernel is incorrect.
The kernel should take into account elliptical beams. Additionally, I think we need to include the normalization factor:
data = convolve(data, g, boundary='extend') * (2 * np.pi * sig_min * sig_maj)
Where,sig_min and sig_maj are the standard deviations of the minor and major axes, respectively.
sig_min
sig_maj
In makecube.py, the smoothing kernel is currently:
This was written for absolutely calibrated data (i.e. Kelvin-scale). As the data are in Jy/Beam, this normalization/kernel is incorrect.
The kernel should take into account elliptical beams. Additionally, I think we need to include the normalization factor:
Where,
sig_minandsig_majare the standard deviations of the minor and major axes, respectively.