Abstract

Ku-band dual-polarization radar backscatter measurements from the SeaWinds on QuikScat (QuikScat) and SeaWinds on ADEOS-2 (SeaWinds) scatterometers are used to classify the melt state and estimate melt severity in Greenland. Backscatter measurements are organized into high temporal and high spatial resolution images created using the Scatterometer Image Reconstruction (SIR) algorithm and a new temporal data segmentation technique. Melt detection is performed using a layered electromagnetic model combined with a Markov chain model. The new melt detection method allows classification of the snow-pack into three states: melt, refreeze, and frozen. Melt severity and refreeze severity indexes are also developed. The melt detection methods developed in this thesis are verified by using a one-dimensional geophysical/electromagnetic model simulation of the snow-pack under melting conditions and by comparison with in situ weather station data at the ETH Camp in western Greenland. The diurnal cycle of backscatter measurements is also analyzed at this location. The melt detection and estimation method is applied to the entire Greenland ice-sheet. The resulting melt classifications and melt severity indexes are used to generate a number of maps outlining the features of the 2003 melt season. Good agreement of the melt severity and a 1978 SASS Greenland ice facies map is observed.

Degree

MS

College and Department

Ira A. Fulton College of Engineering and Technology; Electrical and Computer Engineering

Rights

http://lib.byu.edu/about/copyright/

Date Submitted

2006-07-20

Document Type

Thesis

Handle

http://hdl.lib.byu.edu/1877/etd1497

Keywords

melt detection, melt estimation, Greenland, scatterometer, QuikSCAT, SeaWinds, ADEOS-2, ice-sheet, diurnal variation

Share

COinS