Transcript Presentation123.pptx

Understanding Glacier
Characteristics in Rocky
Mountains Using Remote
Sensing
Yang Qing
Introduction and Background
Glaciers are an essential part of the global hydrological cycle and play a
major role in global and regional climate change. Glaciers are a significant
factor of sea level rise and climate change, although mountain glaciers only
make up a small portion of the whole ice in the world, they still are an
essential part acting on sea level rise and climate change and other
environmental hazards.
Many efforts have been put to understand global glaciers using Remote
sensing techniques, with a focus on the Himalayas in central Asia, the Andes
in South America and the Alps in Europe.
The study area is Rocky Mountains which are a major mountain system in
North America, I’ll focus on northern Rockies in Montana and western
Canada where the glaciers are abundant.
Objectives
To understand the glacier characteristics using remote sensing imagery with
a focus on the glaciers classification and mass balance estimation.
Several different classification methods will be performed on remote sensing
imagery, different results will be compared and validated by in situ
observations.
Mass balance estimates will be compared with historical documents.
Which classification approach is the most appropriate one? Is there a
significant mass change during the years, and is it accurate?
Methodology
1.
Select a suitable remote sensing data source based on the
sensor’s ground coverage, revisiting period, spatial resolution,
spectral resolution. For example, ASTER or Landsat ETM+ / DEM
from SRTM.
2. A supervised classification will be performed on the image,
different classifiers such as PLDA will be used and compared, terrain
attributes such as slope and north-exposedness will be included to
yield higher accuracy.
3. An unsupervised classification using ISODATA method will be
performed on the image as well, the result classes will be named or
labeled based on in situ spectra collection and observation.
Methodology
4. In the classification process, NDSI or other band ratios thresholds will be
used to help identify and delineate boundaries. PCA and other linear
combinations may be used to decide the most important bands
information in classification.
5. To cope with debris cover classification problems, maybe hyperspectral
or thermal remote sensing imagery is needed, or the error will be largely
reduced by incorporating terrain attributes.
6. The geodetic approach will be used to estimate the mass balance. This
method consists in measuring elevation changes over time (δh/δt) from
various DEMs constructed over the glacier surface, hence the mass
change will be computed by multiplying the elevation change by pixel
size then by density.
Thanks.