i
Temperature Data from the Eastern Bering Sea Continenta Shelf Bottom Trawl Survey as Used for Hydrodynamic Model Validation and Comparison
-
2021
-
By Kearney, K.
[PDF-3.35 MB]
Details:
-
Personal Author:
-
Corporate Authors:
-
NOAA Program & Office:
-
Description:Over the past four decades, temperature data have been collected across the eastern Bering Sea shelf as part of the annual bottom trawl survey conducted through the Alaska Fisheries Science Center. This dataset includes a spatially resolved annual time series of surface and bottom temperature, and serves as a primary observation-based temperature dataset against which regional ocean models of the region are validated. This report provides an overview of the data processing used to prepare the survey observations for model validation purposes. We then used the survey-derived temperature values for a thorough skill analysis of simulated surface and bottom temperature in the Bering10K model, an implementation of the Regional Ocean Modeling System (ROMS) covering the Bering Sea region. Overall, the Bering10K hindcast simulation captures observed patterns in eastern Bering Sea shelf bottom temperature well, with high correlation, low bias, and comparable interannual variability to the survey data. The exception to this is in the vicinity of the shelf break, where the model performed relatively poorly. This decrease in bottom temperature skill is attributable to bottom topography mismatches between the real and simulated shelf break location and is unavoidable in sigma-coordinate models like Bering10k; users should carefully consider the effects of the shelf break displacement whenever using model data extracted from this narrow region, particularly when attempting direct comparison with observations. Model skill was also generally higher in the southeastern portion of the shelf compared to the northern shelf region, though this may be an artifact of the low number of samples collected in the north relative to the south. Surface temperature performs with similar high correlation and comparable interannual variability, but simulations tended to be biased warm across much of the domain. This is likely due to a shallow bias in the simulated mixed layer that concentrates surface heating near the surface.
-
Keywords:
-
Series:
-
DOI:
-
Document Type:
-
Place as Subject:
-
Rights Information:Public Domain
-
Compliance:Submitted
-
Main Document Checksum:
-
Download URL:
-
File Type:
