 |
Global Land Ice Measurements from Space
GLIMS: Basis, goals, foundation, and organizational structure
Glaciers and ice sheets are the world’s chief repositories of fresh water. In
many situations, they are almost ideal natural reservoirs, as they store water
in frozen form during cold and cloudy (wet) periods and slowly release it during
sunny, hot periods when liquid water is most needed downstream. In common
non-ideal cases, glaciers store and suddenly release vast quantities of water,
sometimes wreaking havoc downstream if those areas are populated or utilized in
critical ways by human beings or the natural world.
It is clear that the world’s climate is undergoing rapid, heterogeneous
change, and with it, glaciers are also changing. Glaciers are also responsive to
climate change over their whole surface areas. Many glaciers are changing in a
steady, linear (and easily predictable) correlation with climatic perturbations.
Others have undergone rapid disintegration; feedbacks and nonlinear responses
can result in unexpected rapid responses of some glaciers and ice sheets.
Because of their compelling and obvious (albeit complex) climate linkages and
because they are so widespread in the world, glacier fluctuations are important
with regard to our understanding of past and ongoing climate change. In general,
it is understood that glaciers are melting at historically rapid rates. It is
critical to document and understand recent glacier fluctuations so as to enable
predictions of future changes and their impacts on people and nature. The larger
ice masses are important because of their major significance with respect to
water resources and hazards, their significance for sea level rise, and because
of their long-term records of past glacial and climatic fluctuations. Glaciers
are critical dynamic elements of some multi-national, multi-billion-dollar
regional hydrologic catastrophes, and so it is urgent for regional and global
political stability that we understand the state and dynamics and impending
future changes of glaciers in these regions (Kargel et al. 2005). It is also
important to document the extent and distribution of smaller ice bodies, since
many will not be around several decades from now, lending urgency to the
harvesting of their climatic information. Some of the many important practical
aspects of glaciers and glacier change, as well as scientific problems that
relate to the changing state of glaciers, are summarized in Figure 1.

Figure 1.
Diagrammatic justification for world glacier monitoring. Center of the
diagram illustrates key elements of basic glaciological science;
yellow-green elements are key areas of investigation by GLIMS. Stream flow
and glacier net mass balance (shown in purple in the “science” part of the
diagram) are the key links between glaciers and their practical importance
to people and natural ecosystems and our understanding of climate change, as
shown in the “Impacts” and “Fundamental Understanding” parts of the diagram.
|
Glaciologists, working individually and for large, well coordinated programs
such as the World Glacier Monitoring Service, have faced the daunting challenge
of inventorying the distribution of the world’s glaciers, characterizing their
state and dynamics, and quantifying basic glaciological parameters, such as
length and area. The Satellite Image Atlas of the Glaciers of the World
(henceforth, the Atlas, Williams et al. 2006) has nearly completed the task of
producing a pictorial record of the world’s glaciers and assessing it for
indications of glacier state and dynamics using chiefly satellite imaging. In
the time since the first civil earth resources satellite imaging of the world’s
glaciers a third of a century ago, satellite glacier imaging of has grown in
quality, resolution, coverage, and frequency of coverage; furthermore, the
technology available to analyze these images for glacier extent and dynamics has
also grown almost immeasurably. Starting with early manual assessments of
glacier extent, material units, and ice motion, increasingly sophisticated and
accurate semiautomatic and automated methods of glacier analysis have been
utilized to extract information difficult or impossible (and very expensive, if
possible at all) to obtain from the ground or by airborne surveys. Technology
and semi-automated applications include mapping glaciers and internal material
units such as lakes, mapping derivative ice flow fields and changes in glacier
extent , and mapping second-derivative changes in flow speeds and acceleration
of glacier retreat.
Notwithstanding these improvements, a steadily shrinking pool of personnel
dedicated to interpreting and understanding glaciological data, especially data
pertaining to glacier change, makes efficiency afforded by technology all the
more critical. It also requires organization and coordination, which can enhance
efficiency and geographic coverage of glacier changes, provide for checks on
accuracy and uniformity of reported data, and reduce excessive redundancy of
effort. This is the context in which Global Land Ice Measurements from Space (GLIMS,
http://www.glims.org) was developed as a team member project for ASTER (Advanced
Spaceborne Thermal Emission and Reflection Radiometer). GLIMS is seen as a
logical extension of the Satellite Image Atlas and the World Glacier Inventory.
Utilizing state of the art imaging and image-analysis technology, and drawing
from a global network of nearly 100 scientists in 25 countries, GLIMS has the
ambitious goals of (1) repeat imaging of the world’s glaciers by ASTER at up to
15 m resolution, in 14 bands spanning the visible through thermal infrared, and
in stereo in one of these bands; and (2) assessing these images for glacier
extent and changes. Though the project organization and some of the analysis
efforts are done as ASTER science team member activities, and overall there is
heavy reliance on ASTER, the project also makes use of imaging by the Landsat
series and many other satellites, historic map archives, and field studies. Many
members of the GLIMS consortium are further dedicated to practical application
of glacier analysis results to problems involving climate change, current and
future water resource availability, and current and future glaciological hazards
(Figure 1).
At its inception, GLIMS was a small ASTER Science Team project with major
ambition, designed by Hugh H. Kieffer, and no funding dedicated strictly to
GLIMS. The observational needs for multispectral and thermal glacier
observations were considered in the engineering development of ASTER. To make
progress toward glaciological analysis on a global scale, it was evident that a
global consortium effort was needed. The GLIMS Coordination Center at USGS/Flagstaff
gained small-scale funding through NASA as a Pathfinder project at about the
time of ASTER’s launch in 1999; development of the consortium was a major early
activity. Kieffer retired in 2003; recompetition for the ASTER Science Team
resulted in Kargel’s assignment as an ASTER Science Team member. GLIMS gained
substantial NASA funding starting in 2004 through separate but explicitly
linked, competitively reviewed proposals from USGS, the National Snow and Ice
Data Center, Portland State University, the University of Maine, and the
University of Nebraska at Omaha. Various project-wide coordination tasks,
database development, and glacier analysis tasks were funded that year. In 2005
Kargel resigned from USGS and brought GLIMS leadership to the University of
Arizona (Tucson). Foreign institutions are funded at various levels through
their own government agencies and other sources. These foreign investigators
provide substantial science leverage, since the sum of foreign support
substantially exceeds the sum of all American support for GLIMS.
Analysis results are being contributed by researchers around the world. To
organize and archive these results, a global glacier database has been designed
at the National Snow and Ice Data Center (NSIDC, Boulder, Colorado). Parameters
are compatible with and expanded from those of the World Glacier Inventory (WGI).
2. Types of ASTER image applications.
ASTER imagery has been used in multiple ways to achieve the objectives of
GLIMS. Transient and rapidly evolving glacier dynamical events have been
monitored using frequent repeat imaging by ASTER (Figure 2 in the case of
the Kolka-Karmadon disaster) or with other data (Figure 3 for the Columbia
Glacier breakup) to achieve a time series. Automated or semi-automated
classification of surface units (Figure 4) is a key to efficient extraction
of needed information. ASTER's major use in GLIMS is to track changes in glacier
extent between successive ASTER images or between other data sets and ASTER
imaging (Figure 5). Another vital area of work is use of multiple ASTER and
other image data sets to track ice motion; the motion of large glaciers can be
detected and mapped using ASTER images taken just weeks apart (Figure 6).
Use of good time series data permits not just rate of change to be measured but
accelerating or decelerating rates of change (Figures 7 and 8).

Figure 2.
Use of ASTER with other imaging systems (Landsat 7 and 5 and DOE MTI) to
observe and document a rapid breakup of the floating tongue of Columbia
Glacier (Alaska) and then seasonal changes. Contributed by Rick Wessels
(Alaska Volcano Observatory). |

Figure 3.
Use of ASTER imaging under emergency high-frequency mode to document before
and after effects of a glaciological disaster in the Russian Caucasus
Mountains. The top panel shows before and after scenes of the disaster,
which occurred in Sep. 2002. The upper right panel shows the debris deposit
due to a massive glacier avalanche that killed over 120 people. The
avalanche traveled 18 km (panel in lower right) from its source in the Kolka
Glacier area (lower left). The main debris deposit (panel in upper right)
rests on the village of Karamadon; the deposit has dammed streams, which
formed lakes and presented further hazards to recovery crews. This main
deposit also expressed a mud flow, which traveled a further 15 km, killing
people along the way. Contributed by Andreas Kääb and Rick Wessels. |

Figure 4.
Llewellyn Glacier, western British Columbia. (A) ASTER image (red-green-blue
composite from bands 3-2-1) draped over an ASTER DEM. (B) Portion of the
ASTER scene whose classification is shown in (C). Contributed by Rick
Wessels. |

Figure 5.
Glacier change (mainly retreat) on the Graham Coast of the Antarctic
Peninsula between acquisition of a Landsat TM scene on 26 November 1989 and
an ASTER scene on 4 January 2001. Contributed by Frank Rau (University of
Freiburg). |

Figure 6.
Glacier flow field of Kronebreen, Svalbard, measured in two ASTER images
obtained just 6 weeks apart (June 26, 2001 and August 6, 2001). Contours
show speeds in m/year and in m/day. Contributed by Andreas Kääb (U. of
Zurich and U. of Oslo). |

Figure 7.
Byrd Glacier, Antarctica. (top) ASTER image mosaic. (bottom) Surface flow
speeds determined by a 1978 theodolite survey (left), and feature tracking
in ASTER image analysis using two recent images (center) and the velocity
change (right) during two recent time periods. Contributed by Gordon
Hamilton and Leigh Stearns (University of Maine). |

Figure 8.
ASTER image used with historic map archives to reconstruct the recession of
the terminus of the Gangotri Glacier (India), which is considered to be the
Holy source of the Ganga River. The graphs show accelerating retreat, which
is ideally consistent with effects due to recent, progressive, and
accelerating climate change. However, the irregularity in recession makes an
unambiguous attribution of cause problematic. Records are needed for many
ice masses. Contributed by Jeff Kargel and Rick Wessels. Data source for
historic records is C.P. Vohra, 1989, Gangotri Glacier, Indian Mountaineer,
published by the Mountaineering Foundation of India (New Delhi). |
For detailed presentations on selected aspects of GLIMS, we refer the reader
to an extended showcase of results in Kargel et al. (2005). Population of the
GLIMS database remains the primary goal of GLIMS. Among other future directions,
we are striving to understand recent glacier variations and project likely
future changes on the basis of climate models and computed assessments of
uncertainty in climate models.
3. References
Kääb, A., Paul, F., Maisch, M., Hoelzle, M., Haeberli, W. (2002). The new
remote sensing derived Swiss glacier inventory: II. First results. Annals of
Glaciology, 34, 362-366.
Kargel, J.S., Abrams, M.J., Bishop, M.P., Bush, A., Hamilton, G., Jiskoot,
H., Kääb, A., Kieffer, H.H., Lee, E.M., Paul, F., Rau, F., Raup, B., Shroder,
J.F., Soltesz, D., Stainforth, D., Stearns, L., and Wessels, R., Multispectral
Imaging Contributions to Global Land Ice Measurements from Space, Remote Sensing
of Environment, 99, 185-223.
|
 |