We present a Fourier method to combine 160 $\mu$m to 500 $\mu$m publicly available Herschel data with the Planck foreground thermal dust model. The method eliminates pervasive negative fluxes present in the Herschel 160 $\mu$m archive data while preserving the angular scale dependence of the background flux at all wavelengths. We adopt an effective scale for transitioning from Planck power to Herschel power of $\kappa_{eff} = 40^\prime$. The noise properties of the deconvolved Planck image constrain $\kappa_{eff}$ to be large. We show that there is previously unrecognized systematic excess power at small Fourier modes in the Herschel 160 $\mu$m images that sets an upper limit on $\kappa_{eff}$ of $\sim 50^\prime$. This excess power at large scales is found in some but not all Herschel 160 $\mu$m images. We apply our method to three regions spanning a range of Galactic environments and image noise properties: Perseus, the B68 region of the Pipe Nebula, and the Galactic plane region around $l = 11\deg$. We find that Perseus and B68 exhibit the above mentioned large-scale excess power; in the case of Perseus our method correct this. In the case of B68 the data are too noisy to make a proper correction and thus we apply the constant--offset correction (Fourier mode 0,0). HiGal--11 is well behaved. For each region we post-process the combined dust continuum emission images to generate column density and temperature maps. We compare these to the previously adopted (0,0) mode Fourier corrections and show that the scale dependent Planck model corrections yield lower column densities and higher temperatures in diffuse regions, or regions surrounding strong emitting objects compared to the previous method. Discrepancies between both methods are reported over significant ($\sim$15%) areas of the images.
↧