[Coral-List] NOAA Seminar tomorrow: Machine Learning and Spatial Forecasting for Dynamical Insights into Coral Reef Systems

Tue Aug 8 14:11:24 EDT 2017

OneNOAA Science Seminar Series <https://www.nodc.noaa.gov/seminars/>
*Please forward to folks who might be interested; thanks.*

*Titlle: Machine Learning and Spatial Forecasting for Dynamical Insights
into Coral Reef Systems*

Speakers: Dr. Kenneth Ells, Project Scientist, Dep't of Physics and
Physical Oceanography, University of North Carolina at Wilmington (UNCW),
and Dr. Dylan McNamara, Chair/Associate Professor, Dep't of Physics and
Physical Oceanography, UNCW. Presenting remotely from Wilmington.

When: Tomorrow, Wednesday, August 9, 2017, 12 to 1 pm EDT

Where: NOAA SSMC4 Room 8150 or via Webinar - see login info below.

Sponsor: NOAA's National Ocean Service Science Seminar; seminar co-hosts
are Tracy.Gill at noaa.gov

Webinar Access: Mymeeting webinar uses phone for and internet. Audio is
only available
over the phone: dial toll-free from US or CAN: 1-877-708-1667. Enter code
7028688#
For the webcast, go to www.mymeetings.com  Under "Participant Join", click
"Join an Event", then add conf no: 744925156. No code is needed for the
web. Be sure to install the correct plug‐in for WebEx before the seminar
starts - the temporary plugin works fine.

Abstract: Coral reefs are among the most biodiverse and productive habitats
on the planet, and provide a host of economic services including shoreline
protection, fisheries and tourism. These ecosystems have experienced
unprecedented changes over the past quarter century from a combination of
anthropogenic and natural stressors. Recent advances in underwater imaging
have provided a deluge of data showing the configuration of coral and algal
species over spatial scales that encompass many thousands of individual
species, a scale that was previously impossible with existing underwater
survey techniques. However, current efforts investigating these images are
hampered by the time required to identify species at each pixel location
within the large photo mosaics. We describe the application of cutting-edge
machine learning technologies for image classification of coral reef
benthic species over large spatial scales and then apply the classification
technology to photo mosaic images taken from reefs in states ranging from
pristine to degraded.  We will also discuss how classified images can
answer dynamical questions about coral reef configuration. The
organizational rules for species across a reef seemingly have both random
and deterministic features.  For example, the landing of coral recruits in
a turbulent water column or periodic disturbances from wave events can be
considered random spatial influences while hierarchical competition for
space between coral and algal species would be considered deterministic. It
is not clear whether random or deterministic spatial dynamics dominate the
configuration of coral species in space and common aggregate spatial
statistics are not useful in making this distinction.  We will present a
novel spatial forecasting tool that can clearly distinguish randomness from
determinism and apply the technique to classified images of coral and algal
species from Palmyra Atoll, a relatively pristine reef in the Southern
Pacific Ocean. We find that the spatial distribution of coral and algal
species shows clear signs of determinism, suggesting that nonlinear spatial
dynamics dominate the battle for space in coral reef systems.

About the Speakers:

Dr. Kenneth Ells received his Ph.D. in Earth Sciences from Duke University.
After the completion of his Ph.D., Dr. Ells has been a postdoctoral
researcher and project scientist at UNCW in the Complex Adaptive Systems
Laboratory in the Department of Physics and Physical Oceanography.  Dr.
Ells research interests include large-scale coastal morphodynamics,
entropy, and machine learning.  Dr. Ells has contributed to projects funded
by the UNCW Center for Marine Science, the National Science Foundation, the
National Fish and Wildlife Foundation, and the National Oceanic and
Atmospheric Administration.

Dr. Dylan McNamara received his M.S. degree in Physics from San Diego State
University and his Ph.D. in Oceanography from the Scripps Institution of
Oceanography at UCSD.  After completing his Ph.D., Dr. McNamara was a
postdoctoral scholar at Duke University in the Earth and Ocean Sciences
Division before joining the faculty at UNCW in 2008. Dr. McNamara has broad
research interests with publications in many fields including; Optics,
Coastal Science, Coral Reef Ecology, Chaos Theory, and Environmental
Economics. His expertise lies in Complex Systems Analysis, Numerical
Modeling, and Nonlinear Forecasting. Dr. McNamara has received grants to
support his research from the National Science Foundation, the National
Oceanic and Atmospheric Administration, the National Fish and Wildlife
Foundation, the Gordon and Betty Moore Foundation, and the North Carolina
Sea Grant.

Subscribe to the OneNOAA Science Seminar weekly email: Send an email to
OneNOAAscienceseminars-request at list.woc.noaa.gov with the word `subscribe'
in the subject or body. See http://www.nodc.noaa.gov/seminars/

-- 

Tracy A. Gill   tracy.gill at noaa.gov   240-533-0349
Physical Scientist
, 
Biogeography Branch, Marine Spatial Ecology Division
National Centers for Coastal Ocean Science
, 
NOAA's National Ocean Service
1305 East-West Hwy, N/SCI-1, # 9208, Silver Spring, MD  20910-3278

*Check out NOAA's **NCCOS Website <http://coastalscience.noaa.gov/>*
*for more on our projects, products and news.*

          o                                  o

><(((o>                                    o

            <o)))><                    o

    <o)))><                  ><(((o>   ><(((o>