(pdf of agenda with map)

Northwest Washington GIS User Group Agenda
September 16, 2005

North Cascades National Park, Sedro Woolley Office
810 State Route 20, Sedro Woolley, WA

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

9:00 - 9:45     Round Table

9:45 - 10:00    Break

10:00 - 10:45   Comparison of Pixel-Based and Object-Oriented Approaches to Vegetation Mapping in Hawaii Using High Resolution IKONOS Imagery

        Mike Leech, WWU Graduate Student, GIS Analyst, Upper Skagit Indian Tribe

10:45 - 11:30   GIS Methods for Incorporating Catastrophic Risk Assessments in Chinook Recovery Planning

        Jeremy Davies, Conservation Biology Division, Northwest Fisheries Science Center, NOAA Fisheries

11:30 - 12:00   Wrap-up

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Comparison of Pixel-Based and Object-Oriented Approaches to Vegetation Mapping in Hawaii Using High Resolution IKONOS Imagery - Mike Leech, WWU Graduate Student, GIS Analyst, Upper Skagit Indian Tribe

IKONOS 1-m panchromatic and 4- multi-spectral images were used to map plant communities (native and invasive) at the alliance level on the Island of Oahu in the Hawaiian Islands.  We examined classification methods using the Maximum Likelihood Classifier (MLC) at the pixel level and a combination of Nearest Neighbor (NN) and fuzzy logic at the object level.  We hypothesized that the ability to delineate Hawaiian plant communities, particularly with invasive species, would be improved using the object-oriented approach.   Among the classification methods examined in this research, the object-oriented classification method achieved the highest classification accuracy.  The pixel-based method received an overall classification accuracy of 64% while the object-based approach received an overall accuracy of 75%.  The goal of this research was to provide resource managers with a useful tool for landscape-level monitoring and mapping native and invasive plant communities in Hawaii.  While both methods did not achieve a classification accuracy of > 80%, this research has provided direction for using a combination of IKONOS imagery and the object-based approach for modeling vegetation in the Hawaiian Islands and other Pacific Island regions.


GIS Methods for Incorporating Catastrophic Risk Assessments in Chinook Recovery Planning - Jeremy Davies, Conservation Biology Division, Northwest Fisheries Science Center, NOAA Fisheries

Catastrophic environmental events may have a great influence on extinction risk for Pacific salmon and steelhead.  However, the potential threats from catastrophic events (e.g., landslides, earthquakes, oil spills, etc.) are often overlooked when developing recovery plans for threatened and endangered species.  To assess the risk of significant losses to threatened Puget Sound Chinook salmon due to environmental catastrophes, we compiled data for a variety of potential natural and anthropogenic events, including volcanoes, earthquakes, landslides, toxic storage facilities, transportation routes for toxic substances, hatchery releases and unplanned dam breaches.  Using spatial overlays, simple modeling techniques, and a variety of statistical methods, we estimated the relative level of risk from these events for each of the 22 independent populations of Puget Sound Chinook.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Driving Instructions: 
o       Take Interstate 5 North or South to Exit 230, north of Mount Vernon. 
o       Take Highway 20 East (The North Cascades Highway) thru Burlington to Sedro Woolley (heading towards Concrete and Marblemount, approximately 5 miles).
o       As Highway 20 comes thru Sedro Woolley it will take a bend to the left followed by a turn to the right. About 4 blocks after crossing the railroad tracks you will find the North Cascades National Park Office on your left, next to the Arco AM/PM Gas Station (if you come to Highway 9 North you've come ½ block too far...).
o       Parking: Park in the Church parking lot next door (just to the west of the Park office). Please park only in spaces marked with yellow.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~