Tristan Sebens

Tristan Sebens


M.S. Student


College of Fisheries and Ocean Sciences
17101 Point Lena Loop Road
Juneau, Alaska 99801-8344


University of British Columbia
B.S. Biology
University of British Columbia
B.S. Computer Science



Comparison of statistical techniques to integrate variable gear-type fishery-independent survey data




Born and raised in Haines, Tristan has lived in Alaska his entire life except for his college years. He worked for 6 years as an associate software engineer for a local Juneau contract development firm, during which he worked primarily as a GIS analyst and data scientist for the NMFS compiling sonar data for high resolution sea-floor terrain models, and analyzing the fishing behavior of the Alaskan fleet to better understand commercial fish species distribution. Tristan began his Masters degree in the fall of 2020. His work is focused on using statistical models to estimate the relative catch efficiencies and size-selectivity of multiple fishery-independent survey data sources, with the intent of blending them together and producing relative indices of abundance with better accuracy and precision than any one data source could provide. Tristan is an avid fisherman and mechanic, and is happiest in his garage tinkering with his vehicles, or in a computer lab writing miles of code.


Selected Publications

Lewis, S., Pegus, C., & Sebens, T. (2016). Development of a comprehensive, high-resolution bathymetric dataset of the US\Alaska Exclusive Economic Zone and surrounding waters using all available and relevant ocean depth data sources. Development10, 28.


  • Spatial statistics
  • Species distribution
  • Data intercalibration


Research Overview

Fishery-independent survey data is a critical part of accurate population assessment and fisheries management. Surveys which are conducted with different gear-types and/or protocols produce observations in their own unique units of effort, or CPUE, and are subject to their own size- and age-selectivity biases. Multiple statistical models offer the potential to account for these differences and allow multiple datasets to be blended together into larger, more robust datasets. These combined datasets can then, in theory, provide relative indices of abundance with greater accuracy and precision than any single survey dataset. Tristan's research is focused on comparing the relative performance of some of these models, with the hope that future population assessments may benefit from more accurate and precise abundance indices.



  • NOAA