Workshop on the North Sea reopening protocol (WKNSROP)
WKNSROP met to reconsider the autumn reopening protocol for North Sea stocks. The work-shop evaluated the historical performance in relation to delivering improved recruitment estimates for short-term forecasts, and found it acceptable for cod, haddock, plaice and sole, but less so for whiting and saithe (although there were some mitigating factors for these). Increasing the D threshold used to trigger reopening above its current value of 1 did not lead to an increase in overall performance, and therefore it is recommended to stay at 1, unless there was a desire to lower the frequency of reopening, in which case it could be set at 2. The consistency of application, method used, and justification for ages used in the reopening protocol check were also evaluated. The continued use of RCT3 under default settings for the reopening check (apart from taper weighting that may be required for saithe), using an R package (rct3) maintained on the ICES GitHub repository, was recommended, along with using only a single age (usually the youngest age in the assessment, except for saithe) in the reopening check, and using the entire time series of corresponding assessment estimate in the RCT3 regression, apart from where the most recent estimate is considered too uncertain to be used (e.g. for cod, plaice and sole). One of the primary objectives of the workshop was to consider whether any information available during the intermediate year (the year WGNSSK meets and formulates advice for the following year) could be used to improve short-term forecast assumptions about fishing mortality. The most useful, and potentially readily available, source of information was reported landings up to the third quarter of the intermediate year, which were found historically to correlate strongly with landings for the full year (with r2 values exceeding 0.9 in most cases). If such information were available to be used, then an additional reopening check for the intermediate year fishing mortality assumption would be needed, and several thresholds were tested (comparing the spring landings assumptions with the predicted annual landings based on the quarter 1-2 or 1-3 landings); the choice of threshold (e.g. 10, 15 or 20% difference for triggering reopening) would de-pend on the desired trade-off between frequency of reopening and accuracy of forecast assumptions. The workshop also discussed the need for reopening in the first place, given that it would be better, from the point of view of using the most up-to-date data and information, and saving workload, to only run the assessments and forecast, and provide advice once in the autumn. This was particularly the case for Nephrops, for which the spring advice only differs slightly from the preceding autumn advice, because the summer surveys have the biggest impact on the assessments, and data for these are only available for the autumn. It was therefore recommended that ICES once again discusses with clients the need for spring advice. The final result of the workshop was a revised reopening protocol, which includes the use of landings data for the intermediate year, if available.
Published under the auspices of the following ICES Steering Group or Committee