High resolution topography of rivers has facilitated quantification of physical habitat at the reach scale like never before. Such ecohydraulic models range from simple habitat suitability curve models to mechanistic bioenergetic models. However, it is rare that such models are upscaled meaningfully to assess entire fish populations. We present insights from applying such ecohydraulic models at over 700 sites in the Columbia River Basin (USA), to assess status and trends in anadromous salmon populations. We automated the simulation of hydraulic (Delft3D), and subsequent ecohydraulic models capable of predicting carrying capacity using cloud computing for 1000’s of visits over the past 5 years at over 700 sites.  We further characterize reaches in terms of reach type, habitat condition, primary production potential and thermal regime. We then independently produce similar drainage network-scale models (reach resolution, watershed extent) from much coarser remotely sensed data available across the entire Columbia River Basin. These co-variates give us a basis for imputation and extrapolation of reach-scale capacity estimates across entire watersheds. Capacity estimates combined with survival estimates allow a more robust quantification of the freshwater portion of the anadromous lifecycle through life cycle modeling.  Thus, we leverage the detail and insights afforded from advances in remote sensing and high resolution ecohydrualic models to simulate how fish populations are responding to habitat conditions.  More fundamentally, we explore the impact of tributary habitat restoration scenarios and whether the more than $8 Billion US spent on such restoration actions can actually help recover these populations or prevent their extinction.