Figure 1. Click image to enlarge

Working with Annette Dougherty (RACE) on a poster presentation entitled "Geographic comparison of growth and hatch dates of age-0 walleye pollock (Theragra chalcogramma) in the Gulf of Alaska" (Figure 1), I utilized a variety of tools in ArcGIS to produce four maps (images can be found under the temperature section of the poster). Data were selected for and represented on a series of maps to relate spatial and temporal variability in temperature along the Shelikof Strait and southern edge of the Alaska Peninsula where sampling takes place for larval walleye pollock. A temperature sensor that records data at all depths of the tow is deployed alongside the larvae collection net during the research cruise (Vance 2007). Depending on the methods of the scientist's research, the dataset associated with each sample site can also contain a variety of recorded environmental factors other than temperature.

For Annette's research, the versatility of ArcGIS allowed me to create a new copy of the sample sites containing only temperature data at forty meters depth, yet any recorded environmental factors (salinity, CPUE, pH, total length, etc.) can be selected and displayed for any depth of the tow. From this methodology comes the visualization of environmental data at a three-dimensional level, allowing the user to analyze the surrounding oceanic factors at specific periods in an organism's life history (Vance 2007).

Figure 2. Click image to enlarge.

Utilizing ArcScene for three-dimensional visualization of the maps presented above, I was able to present the data at a finer resolution (Figure 1). Figure 1 represents temperature data symbolized at one meter increments for every depth of the tow at each study site in May of 2001. The enlarged image of the highlighted study site relates warmer pools of water at a greater depth than cooler waters.

Representing environmental data in this manner clearly describes three-dimensional patterns or trends that may not be obvious when viewing the dataset in two-dimensional plots. Visualization of biological oceanographic data at this level is a recently applied technique and formed a significant portion of my internship at the AFSC.

Most of my experience working in three-dimensional space came when Janet Duffy-Anderson introduced me to her work visualizing output data from the Northeastern Pacific Regional Ocean Model System (ROMS), a free-surface, hydrostatic primitive equation ocean circulation model. The ROMS was used to simulate larval Greenland halibut (Reinhardtius hippoglossoides) pathways as passive particles (called drifters) under local ocean currents in the Bering Sea. Greenland halibut are thought to spawn in submarine canyons of the Bering Sea (Bering, Pribilof, and Zhemchug Canyons), with the larvae

Figure 3. Hypothesized passive larval Greenland halibut trajectories at depth for Zhemchug Canyon in the Bering Sea.

being advected to their nursery grounds on the shelf via currents during their early life history (Figure 2).

The model was initialized with 10 drifters deployed horizontally across each of three submarine canyons, at 11 discrete, vertical depths (0, 10, 20, 30, 40, 50, 60, 70, 80, 90 & 100m). Drifter float tracking occurred in multiple years (1995, 1997, 2003, and 2004) between April and September, encompassing the larval pelagic period for Greenland halibut. Fixes for each simulated drifter was once per day, resulting in a final, working dataset of over 500,000 points (Figure 2). For each year the model was run, the output points were reduced to only twenty-five linear pathways in each canyon (five at each depth of 0, 10, 20, 30 & 50m) to visualize any variation in trajectory of the drifters at stratified depths (Figure 3).

The process of three-dimensional visualization of drifter trajectories relied heavily on data selection, and data management to produce the desired results. Working with large datasets creates a need for efficient techniques to select for subsets of data specific to project goals as well as strong data management skills to organize the data clearly and concisely for multiple users to access.

Collaboration between multiple scientists and technicians was a key component of each project and my internship as a whole. Working at this level of professionalism was the single most impressive factor I have gleaned from my experience. I gained a great deal of knowledge from my time at the AFSC and a great deal of respect for the scientists, technicians, administrators, and fellow interns I interacted with. The people behind the AFSC embody much diversity in terms of knowledge, age, professional stature, and tenure, yet they are all working towards a common goal: furthering innovation in fisheries science and honing tools to sustain fisheries for generations to come. Alaskan waters are hailed as "the bread basket" for our domestic fisheries resources and although there is a seemingly limitless amount of work needed to sustain those resources, my relatively brief experience working with the AFSC has assured me that there is as much capacity to achieve long-term sustainability as there is support.

Sources:
-Figure 1 poster: Dougherty, A., K. Bailey, T. Vance. 2009. Geographic Comparison of Growth and Hatch Dates of Age-0 Walleye Pollock (Theragra chalcogramma) in the Gulf of Alaska. NOAA/NMFS/AFSC/RACE. Available: ftp://ftp.afsc.noaa.gov/posters/pDougherty02_pollock-goa.pdf (September, 2009)

-Figure 2 poster: Blood, D.M., J.T. Duffy-Anderson, A.C. Matarese, D. Sohn, P.J. Stabeno, W. Cheng. 2009. Early Life History of Greenland Halibut (Reinhardtius hippoglossoides) in the Eastern Bering Sea Based on Recent Field Studies (2007-2009): Spawning, Distribution, and Physical Processes Affecting Drift and Dispersal. NOAA. Available: ftp://ftp.afsc.noaa.gov/posters/pBlood02_greenland-halibut.pdf (September, 2009).

-Vance, T. 2007. If You Build It, Will They Come? Evolution Towards the Application of Multi-Dimensional GIS to Fisheries-Oceanography. Doctoral dissertation. Oregon State University, Corvallis, Oregon. Available: http://marinecoastalgis.net/tiffany07 (September, 2009).