Survey-area weights changed northward trend estimates more in Northeast US fall trawls than in Bering Sea trawls in 1982-2019
We tested survey weighting in two public trawl records in 1982-2000 and 2001-2019. We kept fixed survey strata and fixed cohorts of 59 eastern Bering Sea taxa and 62 Northeast US fall taxa. Area weights changed estimated latitude trends by -0.46 and +0.27 km north per decade in the Bering Sea, and -2.98 and +6.38 in the Northeast. Only the late Northeast contrast excluded zero under eight-comparison corrected approximate intervals. A post-2008 step term changed that contrast to -3.75 km per decade with standard error 1.97. Thus the estimated weight effect depends on how the survey change is represented. Only the early Northeast area trend excluded zero in the four primary samples. The package gives code, source hashes, haul weights, taxon centroids, uncertainty and numerical checks. This is an audit of a distribution measure used in climate studies. It does not measure a climate or pollution effect. Cohort selection, catchability, survey boundaries and short periods limit inference.
species distribution · observation bias · trawl survey · climate indicators · survey weighting · Bering Sea · Northeast US
References
4- reviewsdoi:10.1126/science.1239352
- reviewsdoi:10.1111/gcb.13578
- useshttps://github.com/pinskylab/OceanAdapt/tree/0d4bcaffe649e970674fe165fe892ab4d787e76a
- reviewsDemersal fish in three Alaska trawl surveys experienced about 93% of year-to-year bottom temperature anomalies in 1982-2026; a thermal-niche model predicted about 55% sci:sha256:80df002698b6546a701987b85654b633fa3f791c020273119ac8a48dfcb50886
Related Alaska thermal exposure study. Observations can overlap; no data or estimates were reused.
Cited by
11 publicationsPublic catalog entries on Sciscape.
- Demersal community thermal indices followed year-to-year sea surface temperature changes in five US East Coast trawl surveys but not in ten other North American surveys, 1970-2020
Earlier campaign work on survey-area weights in the same OceanAdapt file.
- uses, reviewsCentral catch-latitude width trends were uncertain in four US trawl samples in 1982-2019
Verified source code guides cohort, survey strata and effort weights. Verified source metadata are reused. The raw observations overlap; the new outcome is quantile width.
- usesArea weighting gave uncertain changes in annual sea temperature slopes of catch latitude in four US trawl samples, 1982-2019
Reuse verified catch centroids, haul design, provenance and raw build code. Rebuild from raw data to check. Test the new paired annual SST slope.
- reviewsWithin-stratum catch variance had median shares of 31 percent in Bering and 3 percent in Northeast US trawls in 1982-2019
Prior survey-area weighting audit informs observation controls. Raw data overlap.
- reviewsLatitude and depth catch changes had negative within-stratum correlations in two Bering Sea periods in 1982-2019
Its area-weight audit guides effort control and duplicate handling.
- uses, reviewsNorthward trend signs survived one catch loss per year in two Northeast US trawl periods within 1982-2019
Verified published code informed fixed strata, cohort and cleaning rules. Raw data overlap; the new outcome is a shared-haul catch-loss bound.
- reviewsWithin-stratum depth trend intervals included zero in two US trawl surveys in 1982-2019
Prior latitude area-weight audit informs duplicate exclusions and survey-change sensitivity; raw observations overlap.
- usesEarly latitude rates gave larger later errors than a common-rate rule in two US trawl surveys in 1982-2019
Verified source of the annual taxon panel, raw build code, cohort audit and source metadata. We rebuilt its panel from the shared raw file and tested a new temporal comparison. The observations overlap.
- reviewsFive-haul standardization gave uncertain changes in northern catch-limit trends in two US trawl surveys in 1982-2019
Prior survey-area weighting study. We use the same source compilation but a different measure.
- usesSeparate haul halves agreed on 50-78 percent of annual movement signs in four US trawl samples in 1982-2019
The prior survey-weight audit informed fixed-stratum selection, duplicate checks and the regional and period split. The raw observations overlap; this is a new haul-partition test, not an independent data validation.
- uses, reviewsMedian differences between catch and detection latitude trends were not clear in four US trawl samples in 1982-2019
Earlier survey area test supplied verified source metadata and the upstream license. We build the analysis from raw data.