Historical data generated from reconstructed estimates of daily marine abundances of Fraser River sockeye salmon form the basis for critical components of pre-season planning and in-season stock assessments. Pre-season, historical time-series of timing and diversion are used to parameterise forecast models and fisheries simulations. In-season, reconstructions are used to update estimates of run abundance, timing, diversion rate and test fishery catchability (expansion lines). For fisheries planning and assessment purposes, PSC Secretariat biologists assess these estimates at a refined stock resolution, often subset by cycle-line. Over time, the stock-resolution, file structure, and assumptions applied to inseason run reconstructions has evolved, leading to inconsistencies across the time series. In addition, post-season adjustments, including updated stock ID and corrected/updated catches, can lead to a disparity between the in-season run reconstruction files and postseason estimates.
We propose to update and standardize historical run reconstruction files using recently updated post-season catch and Mission passage information and currently adopted data standards. We also propose to build a new data warehouse to store post-season estimates of daily run reconstructions, from which time series of fish migration behaviour and fishery harvest parameters can easily be updated and queried at various stock resolutions and year-aggregations (e.g. by cycle line or dominant year). These improvements will bring the quality of the run reconstruction estimates in line with the recent improvements to the post-season catch, passage and escapement records, and help the Secretariat provide the Fraser River Panel with key inputs to both the preseason and in-season management processes. Updated stock-specific information on migration timing and diversion rate would also be shared with DFO to facilitate improvements to stock-specific forecasts for these important inputs in fisheries planning.