We combined field data and the output from a climate-to-fish coupled biophysical model to calculate weekly climatologies and 1971–2009 time series of physical and biological drivers for 16 distinct regions of the eastern Bering Sea shelf and slope. We focus on spatial trends and physical-biological interactions as a framework to compare model output to localized or season-specific observations. Data on pollock (≥8cm) diet were used to evaluate energy flows and zooplankton dynamics predicted by the model. Model validation shows good agreement to sea-ice cover albeit with a one month delay in ice retreat. Likewise, the timing of spring phytoplankton blooms in the model were delayed approximately one month in the south and extend further into summer, but the relative timing between the spring and fall bloom peaks was consistent with observations. Ice-related primary producers may shift the timing of the spring bloom maximum biomass earlier in years when sea ice was still present after mid-March in the southern regions. Including the effects of explicit, dynamic fish predation on zooplankton in the model shifts the seasonal spring peak and distribution of zooplankton later in the year relative to simulations with implicit predation dependent only on zooplankton biomass and temperature; the former capturing the dynamic demand on zooplankton prey by fish. Pollock diets based on stomach samples collected in late fall and winter from 1982–2013 show overwintering euphausiids and small pollock as key prey items in the outer and southern Bering Sea shelf; a characteristic not currently present in the model.The model captured two large-scale gradients, supported by field data, characterizing the overall dynamics: 1) inshore to off-shelf physical and biological differences with a gradient in inter-annual variability from higher frequency inshore to lower frequency offshore; and 2) latitudinal gradients in the timing of events. The combined effects of length of day, bathymetry, and tides, which are consistent from year to year, and the two large-scale gradients, characterize the environment on which regional differences were based and restrict their inter-annual and seasonal variability. Thus, the relative timing and sequence of events remained consistent within regions. The combination of model outputs and observational data revealed specific ecosystem processes: (1) The spatial progression in the timing, peaks and sequence of events over the shelf is driven by wind, sea ice, and stratification and creates a seasonal expansion and contraction of the warmer pelagic and bottom habitat suitable to pollock. (2) The seasonal warming of air temperature and the spring-summer expansion of the warm pelagic and bottom habitats influence the ice retreat and the associated ice edge and open water spring blooms, as well as subsequent production/abundance of copepods and euphausiids. (3) These warmer conditions favor pelagic energy flows to pollock (≥10cm) and allow their distribution to expand shoreward and northward along the shelf break. (4) The fall-winter expansion of the seasonal ice cover drives the contraction of warmer waters towards the outer and southwest shelf and favors benthic energy flows over most of the shelf. There, fall blooms allow for additional lipid storage by large copepods and euphausiids that sink close to the bottom where they either go into diapause or have a restricted diel migration over winter. (5) During these cold months, the preferred pollock habitat shifts and contracts towards the outer and southwest shelf where their increased density and reduced prey availability leads to winter pollock cannibalism and consumption of overwintering euphausiids. Our project highlights the benefits of linking continuous and long-term field work with the development and implementation of highly complex models. In the face of uncertainty, simulations such as these, tightly coupled to field programs, will be instrumental as testbeds for process exploration and management evaluation, increasing their relevance for future fisheries and ecosystem management and strategic planning.