Tracking Ice Flow in Two Greenland Outlet Glaciers Michael Jefferson Jr. Elizabeth City State University Research & Discover Summer 2010 Mentor: Mark Fahnestock.

Tracking Ice Flow in Two Greenland Outlet GlaciersTracking Ice Flow in Two Greenland Outlet Glaciers

Michael Jefferson Jr.Elizabeth City State UniversityResearch & Discover Summer 2010

Mentor: Mark Fahnestock

Michael Jefferson Jr.Elizabeth City State UniversityResearch & Discover Summer 2010

Mentor: Mark Fahnestock

ObjectivesObjectives

• To identify the best pairs of files to investigate ice flow– There has to be a pair of images in order to see a change in the data

• Create a new program to process the images

• Gather the images and process them with the appropriate parameters

• Compare findings from each of the processed image pairs

• To identify the best pairs of files to investigate ice flow– There has to be a pair of images in order to see a change in the data

• Create a new program to process the images

• Gather the images and process them with the appropriate parameters

• Compare findings from each of the processed image pairs

Glaciers and Remote SensingGlaciers and Remote Sensing

• The Landsat 7 satellite orbits the earth and takes pictures of earth’s features

• The data collected by this satellite was used in my research

• The Landsat 7 satellite orbits the earth and takes pictures of earth’s features

• The data collected by this satellite was used in my research

LocationLocation

•Latitude(62.4), Longitude(-49.6)Kangiata Nunata Sermia(KNS)

LocationLocation

Nunatakavsaup Sermia(NKS)

• Latitude(74.6) ,Longitude( -56.0)

•The distance was 1300 km between the two glaciers

NKS

Selecting Image filesSelecting Image files

• Files where downloaded from the United States Geological Survey (USGS) GLObal VISsualization Viewer (GLOVIS) website

• There were many images with too much cloud cover– If clouds cover features then the changes can not be tracked

• 2004, 2005, 2007 used for KNS

• 2001, 2002, 2005, 2010 used for NKS

• Files where downloaded from the United States Geological Survey (USGS) GLObal VISsualization Viewer (GLOVIS) website

• There were many images with too much cloud cover– If clouds cover features then the changes can not be tracked

• 2004, 2005, 2007 used for KNS

• 2001, 2002, 2005, 2010 used for NKS

The ProgramThe Program

• To track features a program was used – The idea of tracking glaciers through pairs of images was developed 20 years ago• Dr. Bob Bindschalder and Dr. Ted Scambos

• Dr. Mark Fahnestock wrote the program– Restrictions: Only output the changes for one iteration over a single image

• To track features a program was used – The idea of tracking glaciers through pairs of images was developed 20 years ago• Dr. Bob Bindschalder and Dr. Ted Scambos

• Dr. Mark Fahnestock wrote the program– Restrictions: Only output the changes for one iteration over a single image

The Program(Cont’d)The Program(Cont’d)

• The program works by reading a line from a .txt file– A text file includes the two images and the geographic locations where the moving should be present

• The program works by reading a line from a .txt file– A text file includes the two images and the geographic locations where the moving should be present

The Program cont’d The Program cont’d

srcimgdel: 10 destimgdel: 25

Source 20x20Destination 50x50

Program IterationProgram Iteration

• Example of a single point of an iteration

• Describes the correlation over the source and destination

• The closer to the color red means the more “good” unique points

• Example of a single point of an iteration

• Describes the correlation over the source and destination

• The closer to the color red means the more “good” unique points

Creating a New ProgramCreating a New Program

• The previous program only ran one set of iterations per image

• Not all sections of the image had distinct features

• Not all images were 100% cloud free

• A new program was created to display more than one

iteration on a single image

• The previous program only ran one set of iterations per image

• Not all sections of the image had distinct features

• Not all images were 100% cloud free

• A new program was created to display more than one

iteration on a single image

Pic of one iteration

New Program: Multi-Velocity

New Program: Multi-Velocity

• Display multiple iterations with different parameters

• Display the average velocity of the “good points”

• Good points were the iteration points that had a unique correlation surface values over a specified amount

• Display multiple iterations with different parameters

• Display the average velocity of the “good points”

• Good points were the iteration points that had a unique correlation surface values over a specified amount

Output from New Program

Output from New Program

Green Plus- Best Correlation

Red Hexagon - Worst Correlation

Mapping ChangesMapping Changes

Results for NKSResults for NKS

This line represents the beginning of the comparable data

2003-20052003-2005

•In years there was a large Ice front retreat this is what drove the large change in velocity over the years

•Figure by: Ryan Cassotto

20052006200720082009

20002001200220032004

Results for KNSResults for KNS

Around 1 to 1.5 Km from the ice front my data was approximately the same through the years

ConclusionConclusion

• There was a large change in velocity in the northern glacier NKS during the years of 2002-2005

• There was no apparent changes in the southernmost glacier KNS

• There was a large change in velocity in the northern glacier NKS during the years of 2002-2005

• There was no apparent changes in the southernmost glacier KNS

Future WorkFuture Work

• Research more glaciers to investigate more velocity changes using satellite data

• Change the program to be more efficient to run on larger collections of data

• Research more glaciers to investigate more velocity changes using satellite data

• Change the program to be more efficient to run on larger collections of data

AcknowledgementsAcknowledgements

• Dr. George Hurtt(Research & Discover)

• Dr. Mark Fahnestock (UNH Researcher)

• USGS GLOVIS (Image Database)• Ryan Cassotto(Graduate Student)• Stephen Hale(UNH Professor)

• Dr. George Hurtt(Research & Discover)

• Dr. Mark Fahnestock (UNH Researcher)

• USGS GLOVIS (Image Database)• Ryan Cassotto(Graduate Student)• Stephen Hale(UNH Professor)

Questions/CommentsQuestions/Comments