
Bristol Bay Sockeye Forecast Looks Normal This Time Around
Here’s more from the Alaska Beacon:
The forecast for 2027, released by the Alaska Department of Fish and Game last week, estimated a return of 49.31 million sockeye salmon in the coming summer season. That would support a harvest of 33.97 million fish, the department forecast said.
The forecasted totals are about normal compared to the last 20 years, and would follow another normal run this year, when 45.7 million sockeye returned to the bay. But it would be 31% larger than the long-term average dating back to 1963, according to the department.
This year’s forecast, like others produced by the department, includes a range of uncertainty. The run could be as low as 35.47 million and as high as 63.14 million fish, according to the department.
Estimating future salmon returns “is inherently difficult and uncertain,” the department forecast said. Over the past two decades, the forecasts have underestimated actual runs by an average 13%, though there have been occasions when the forecast total was higher or almost exactly the same as the actual run total.
Here’s a snippet from ADFG’s forecast:
Forecasting future salmon returns is inherently difficult and uncertain. The department has used similar methods since 2001 to produce Bristol Bay sockeye salmon forecasts, which have performed well when applied to Bristol Bay as a whole. In the last 20 years, our forecasts have on average under forecasted the run by 13%, ranging from 36% below the actual run in 2014 to 21% above the actual run in 2011 (Figure 1). Forecasted harvestable surplus has had a mean absolute percentage error of 16% since 2007.
Forecasts at the individual river scale exhibit greater uncertainty than the total Bristol Bay forecast. Since 2007, return estimates have been biased low for the Alagnak (-15%), Kvichak (-34%), Wood (-18%), Nushagak (-15%), Igushik (-9%), and Ugashik (-10%) Rivers, and biased high for Egegik (4%), Togiak (2%) and Naknek (0.3%) Rivers. These offsetting errors contribute to higher accuracy in the bay-wide forecast compared to forecasts for individual rivers.
