This study presents results from models that are designed to simulate the underwater light field, to simulate phytoplankton primary production, and to estimate the fate of phytoplankton carbon in continental shelf waters of the west Antarctic Peninsula (WAP) and Ross Sea. Simulation of the underwater light field required derivation of new coefficient sets for power function-type cloud cover correction algorithms, which were found to be influenced by multiple reflections between the bottom of clouds and the surface. The coefficient sets indicate that the spectral effect of clouds on the properties of the surface irradiance was spectrally-neutral for wavelengths greater than 330 nm. The regional dependency of the newly-derived coefficient sets provide an approach for developing general cloud cover correction algorithms for Antarctic coastal waters. Next, a bio-optical production model that was forced with the simulated surface irradiance fields, corrected for cloud conditions, and the simulated underwater light field was used to estimate primary production and subsequent carbon flux at several sites along the western Antarctic Peninsula and in the Ross Sea. The parameterizations used in the bio-optical production model included depth-dependent photosynthesis-irradiance relationships that involved different patterns of diel variation. Sensitivity studies showed simulated primary production estimates were increased by up to 130% when photosynthetic parameters with a diel periodicity were used in the production model. Inclusion of spectrally-resolved quantum yields increased primary production estimates by as much as 300%, relative to a reference simulation that used constant parameters. The fate of newly-produced phytoplankton carbon obtained from simulations for the WAP and Ross Sea was investigated using budget calculations that included the effects of grazing, advection, and sinking. For the western Antarctic Peninsula region, horizontal (across-shelf component) advection is the dominant process controlling primary production carbon in the outer-shelf areas in all seasons. Depending on season, advection can remove up to 40% of the phytoplankton carbon in the shelf waters. Grazing, however, is as important as across-shelf advection during the summer and can be an order of magnitude greater in inner shelf waters than in mid- and outer-shelf waters. Sinking is also a dominant process that can remove up to 6% of primary production carbon, except in the austral winter season. Similar calculations for the Ross Sea show that zonal advection is the dominant process controlling phytoplankton primary production carbon (up to 57%) in the outer-shelf regions in all seasons. Grazing is an important removal process in the summer in the inner- and mid-shelf areas of Ross Sea continental shelf waters, but was found to be less of a control relative to advective removal. Sinking is also an important process for removing phytoplankton carbon, with 20% and 220% of the daily primary production being removed by this process in the summer and winter, respectively. The results of carbon budgets show that advective processes provide a dominant control on the fate of primary production, which suggests that primary production estimates for Antarctic coastal waters should be based on observational studies or models that incorporate circulation as well as biological processes.