West Coast Groundfish Observer Program Data Processing
Northwest Fisheries Science Center Observer Program procedures for processing West Coast Groundfish Observer Program and PacFIN fish ticket data for analysis and use in research and management.
To process West Coast Groundfish Observer Program data for further analysis, we:
- Expand the subsample of species composition to the tow/set-level.
- Translate observer species codes to the corresponding fish ticket data codes used in the Pacific Fisheries Information Network (PacFIN) database.
- Match observed trips to landed fish tickets.
- Query and process PacFIN fish ticket data for 2002 to the most recent year of observer data.
- Merge observer data and fish ticket data.
- Remove incompletely sampled and failed data.
The West Coast Groundfish Observer Program species composition sampling procedure allows for subsampling, so tow/set-level expansions are needed to estimate the total retained and discarded weight for each species. We calculate the subsample's weight by summing across the observed weights of the individual species using the equation:
xks= observed weight of the species s in catch category k in the subsample
wk = weight of the subsample from catch category k
The sampling ratio (Rk) used to scale the subsample weights to the total weight of the catch category is calculated by dividing the weight of the subsample by the total weight of the catch category:
yk = the total weight of catch category k
The tow/set-level expanded weight of species s in category k is calculated by dividing the species weight in the subsample by the sampling ratio:
Xks= the weight of species s in catch category k
Tallying the weight (Xks) of the species (s) across all categories (k) within a tow/set estimates the total weight of the species retained or discarded.
Merging the retained catch records from the observer data with fish ticket data provides more accurate retained estimates at the haul-level. Observers record fish ticket number(s) for each trip, which we confirm using vessel and port information and the return date recorded by the observer with the dates of fish tickets from the vessel. We use all fish tickets for a given trip. For trips with missing fish tickets, we do not adjust the observer obtained catch data. The translation of our species codes to those used by PacFIN allows a seamless match of observer with fish ticket data and provides consistent information for calculations.
To match retained data appropriately, we remove all observed discarded catch data, as well as observed trips:
- with tows/sets with no retained or discarded information,
- without a matching fish ticket, and
- where the observer monitored less than 100% of landed catch.
We next match species codes between West Coast Groundfish Observer Program and PacFIN and adjust our data so that the total trip pounds of retained fish in a catch category matches the total trip pounds on the fish ticket. The fish ticket weight is often more accurate as well as legally binding documents. To match the total trip pounds, we scale the weights within each observer retained catch category by the ratio of fish ticket and observer trip weights for that category based on the adjustment factor:
xmtk = lbs in catch category k in tow/set t in trip m
Amtk = adjustment factor used for catch category k in tow/set t in trip m
West Coast Groundfish Observer Program data is then adjusted based on:
C mk = lbs in catch category k for trip m recorded on the fish ticket
When we cannot match a catch category in our observer data to a fish ticket catch category, we do not adjust the observer data. We distribute catch categories found only on the fish tickets across the observed tows/sets using the proportion of the observed catch per tow/set divided by the total observed catch per trip by applying the adjustment factor:
Bmt= the proportion of observed catch in tow/set t in trip m
in the equation:
Cmtk = lbs in catch category k for tow/set t in trip m recorded on the fish ticket
After the observer data are merged and adjusted to fish ticket data, any data removed during the matching process are re-incorporated into the matched data file.
Prior to analysis, we remove data that failed quality standards and tows/sets no retained or discarded information was recorded or where the species composition of discarded catch was not known.
The Pacific Fisheries Information Network (PacFIN) has a description of their database.