The Twenty-one centimeter emission has the potential to trace the ionization of neutral Hydrogen in the early universe and map its structure. Detection of the signal will require a model of the radio bright foreground. The foreground, which is mostly radio galaxies, is 4-5 orders of magnitude brighter than the signal and will therefore play a crucial role in its detection. Previously, most radio sources have been modeled as bright points to be subtracted from the data and used in calibration. Our approach combines observations of various objects over multiple nights to build more accurate models of the radio sky. We do this by using a software package called FHD (Fast Holographic Deconvolution), which models objects with complicated internal structures as several points. We combine these point sources over multiple observations by treating them as Gaussian surfaces after extracting the brightest points to be omitted from this smoothing. We believe that using our new models for calibration of the observations could reduce contamination of the reionization signal from foreground emission. Although this research was conducted as part of the HERA (Hydrogen Epoch of Reionization Array) group, the data used was taken at the MWA (Murchison Widefield Array). The MWA has a much higher resolution than HERA and allows for the creation of better models. If successful, this approach will help narrow the probable limits for 21cm detection.