Introduction

Mountain glaciers are key variables for early detection strategies in global climate-related observations. Their changes have been observed systematically in various parts of the world for more than 100 yr. During this time, however, various aspects involved have changed in a most remarkable way. Future perspectives must envisage the possibility of dramatic evolutions, including the rapid deglaciation of entire mountain chains within decades. Worldwide documentation of such developments represents a challenge that must be met using the best available process understanding, methodology and strategy. The Global Terrestrial

Network for Glaciers (GTN-G) recently established as part of the Global Terrestrial Observing System (GTOS/GCOS) and operated by the World Glacier Monitoring Service (WGMS) follows a global hierarchical observing strategy. This integrative approach is based on a combination of in situ, remote and numerical-modelling components and consists of observations at several levels which link detailed process studies at one extreme with global coverage by satellite imagery and digital terrain information at the other. The present contribution briefly outlines the historical background and present challenge of worldwide glacier observations, introduces the integrative concept of the applied multilevel monitoring strategy, summarizes some principal aspects of past measurements and potential future scenarios and concludes with recommendations relating to the most urgent needs. It is based on numerous discussions with, and feed-back from, the staff members, national correspondents and principal investigators of the World Glacier Monitoring Service (WGMS) as well as colleagues involved with the ongoing Global Land Ice Measurements from Space (GLIMS) project led by the US Geological Survey.

84.2 Background and perspectives

Fluctuations of glaciers and ice caps have been observed systematically for more than 100 yr in various parts of the world (Haeberli et al., 1998a) and are considered to be highly reliable indications of worldwide warming trends (cf. fig. 2.39a in IPCC, 2001). Mountain glaciers and ice caps are, therefore, key variables for early detection strategies in global climate-related observations. The internationally coordinated collection of information about ongoing glacier changes was initiated in 1894 with the foundation of the International Glacier Commission at the 6th International Geological Congress in Zurich, Switzerland. It was hoped that the long-term observation of glaciers would provide answers to the questions about global uniformity and terrestrial or extraterrestrial forcing of past, ongoing and potential future climate and glacier changes (Forel, 1895). Since then, various aspects involved have changed in a most remarkable way:

1 Concern is growing that the ongoing trend of worldwide and fast, if not accelerating, glacier shrinkage at the 100-yr time-scale is of non-cyclic nature—there is hardly a question any more of the originally envisaged 'variations périodiques des glaciers' (Haeberli et al., 1998a).

2 Under the growing influence of human impacts on the climate system (enhanced greenhouse effect), dramatic scenarios of future developments—including complete deglaciation of entire mountain ranges—must be taken into consideration (Haeberli & Hoelzle, 1995; IPCC, 2001).

3 Such future scenarios may lead far beyond the range of historical/Holocene variability and most likely introduce processes (extent and rate of glacier vanishing, distance to equilibrium conditions) without precedence in the Holocene.

4 A broad and worldwide public today recognizes glacier changes as a key indication of regional and global climate and environment change.

5 Observational strategies established by expert groups within international monitoring programmes build on advanced process understanding and include extreme perspectives.

6 These strategies make use of the fast development of new technologies and relate them to traditional approaches in order to apply integrated, multilevel concepts (in situ measurements to remote sensing, local-process oriented to regional and global coverage), within which individual observational components (length, area, volume/mass change) fit together enabling a comprehensive view.

An international network of glacier observations such as the World Glacier Monitoring Service (WGMS) of the International Commission on Snow and Ice (ICSI/IAHS) and the Federation of Astronomical and Geophysical Data Analysis Services (FAGS/ICSU), together with its Terrestrial Network for Glaciers (GTN-G; Haeberli et al., 2000) within the Global Terrestrial Observing System (GTOS) and the Global Climate Observing System (GCOS), is designed to provide quantitative and understandable information in connection with questions about process understanding, change detection, model validation and environmental impacts in a transdisciplinary knowledge transfer to the scientific community as well as to policy makers, the media and the public. This difficult but increasingly important task makes adequate perception of glacier changes a challenge of historical dimensions.

84.3 An integrated observational strategy

Within the framework of the global climate-related terrestrial observing systems, a Global Hierarchical Observing Strategy (GHOST) was developed to be used for terrestrial variables. According to a corresponding system of tiers, the regional to global representativeness in space and time of the records relating to changes in glacier mass and area should be assessed by more numerous observations of glacier length changes as well as by compilations of regional glacier inventories repeated at time intervals of a few decades—the typical dynamic response time of mountain glaciers (Haeberli etal., 2000). The individual tier levels can be described as follows:

Tier 1 (multicomponent system observation across environmental gradients) Primary emphasis is on spatial diversity at large (continental-type) scales or in elevation belts of high-mountain areas. Special attention should be given to long-term measurements. Some of the already observed glaciers (e.g. those in the American cordilleras or in a profile from the Pyrenees through the Alps and Scandinavia to Svalbard) could later form part of Tier 1 observations along large-scale transects. Tier 2 (extensive glacier mass balance and flow studies within major climatic zones for improved process understanding and calibration of numerical models) Full parameterization of coupled numerical energy-/mass-balance and flow models is based on detailed observations for improved process understanding, sensitivity experiments and extrapolation to areas with less comprehensive mea surements. Ideally, sites should be located near the centre of the range of environmental conditions of the zone that they are representing. The actual locations will depend more on existing infrastructure and logistical feasibility rather than on strict spatial guidelines, but there is a need to capture a broad range of climatic zones (such as tropical, subtropical, monsoon-type, mid-latitude maritime/continental, subpolar, polar). Tier 3 (determination of regional glacier volume change within major mountain systems using cost-saving methodologies) There are numerous sites to reflect regional patterns of glacier mass/volume change within major mountain systems, but they are not optimally distributed (Cogley & Adams, 1998). Observations with a limited number of strategically selected index stakes (annual time resolution) combined with precision mapping at about decadal intervals (volume change of entire glaciers) for smaller ice bodies or with laser altimetry/kinematic GPS (Arendt et al., 2002) for large glaciers constitute optimal possibilities for extending the information into remote areas of difficult access. Repeated mapping and altimetry provide important data at lower time resolution (decades).

Tier 4 (long-term observations of glacier-length-change data within major mountain ranges for assessing the representativeness of mass balance and volume change measurements) At this level, spatial representativeness is the highest priority. Locations should be based on statistical considerations (Meier & Bahr, 1996) concerning climate characteristics, size effects and dynamics ('normal' flow versus effects from calving, surge, debris cover, etc.). Long-term changes of glacier length at a minimum of about 10 sites within each of the mountain ranges should be measured either in situ or with remote sensing techniques at annual to multi-annual frequencies. Tier 5 (glacier inventories repeated at time intervals of a few decades by using satellite remote sensing) Continuous upgrading of preliminary inventories and repetition of detailed inventories using aerial photography or—in most cases—satellite imagery should enable global coverage and the validation of climate models (Beniston et al., 1997). The use of digital terrain information in geographical information systems (GIS) greatly facilitates automated procedures of image analysis, data processing and modelling/interpretation of newly available information (Haeberli & Hoelzle, 1995; Kaab et al., 2002; Paul et al., 2002). Preparation of data products from satellite measurements must be based on a long-term programme of data acquisition, archiving, product generation and quality control.

This integrated and multi-level strategy aims at integrating in situ observations with remotely sensed data, process understanding with global coverage and traditional measurements with new technologies. Tiers 2 and 4 mainly represent traditional methodologies which remain fundamentally important for deeper understanding of the involved processes, as training components in environment-related educational programmes and as unique demonstration objects for a wide public. Tiers 3 and 5 constitute wide-open doors for the application of new technologies.

A network of 60 glaciers representing Tiers 2 and 3 is established. This step closely corresponds to the data compilation published so far by the World Glacier Monitoring Service with the biennial Glacier Mass Balance Bulletin and also guarantees annual reporting in electronic form. Such a sample of reference glaciers provides information on presently observed rates of change in glacier mass, corresponding acceleration trends and regional distribution patterns. Long-term changes in glacier length must be used to assess the representativeness of the small sample of mass balance values measured during a few decades with respect to the evolution at a global scale and during previous time periods. This can be done by (i) intercomparison between curves of cumulative glacier length change from geometrically similar glaciers, (ii) application of continuity considerations for assumed step changes between steady-state conditions reached after the dynamic response time (Haeberli & Holzhauser, 2003; Hoelzle et al., 2003), and (iii) dynamic fitting of time-dependent flow models to present-day geometries and observed long-term length change (Oerlemans et al., 1998). New detailed glacier inventories are now being compiled in areas not covered in detail so far or, for comparison, as a repetition of earlier inventories. This task is greatly facilitated by the launching of the ASTER/GLIMS programme (Kieffer et al., 2000). Remote sensing at various scales (satellite imagery, aerophotogrammetry) and GIS technologies must be combined with digital terrain information (Kaab et al., 2002; Paul et al., 2002; Bishop et al., 2004) in order to overcome the difficulties of earlier satellite-derived preliminary inventories (area determination only) and to reduce the cost and time of compilation. In this way, it should be feasible to reach the goals of global observing systems in the years to come.

84.4 Views to the past and to the future

Changes in glacier extent and volume provide important qualitative and quantitative information on pre-industrial variability, rates of change, acceleration tendencies and potential future scenarios relating to energy exchange at the earth-atmosphere interface over various time-scales. A brief examination of the main facts illustrates the key issues related to climate change questions.

Long-term mass balance measurements (Fig. 84.1) are available for about 50 glaciers and cover the past few decades. They provide direct (undelayed) signals of climate change and constitute the basis for developing coupled energy-balance/flow models for sensitivity studies. Extensive investigations explore complex feedback effects (albedo, surface altitude, dynamic response) and can be used in conjunction with coupled ocean-atmosphere general circulation models (model validation, hydrological impacts at regional and global scales, etc.; cf. Beniston et al., 1997). Simpler observations using strategically selected index stakes furnish evidence of regional developments. Both types of monitoring combine direct glaciological with geodetic/photogrammetric methods in order to determine changes in volume/mass of entire glaciers (repeated mapping) with high temporal resolution (annual measurements at stakes and pits). Laser altimetry com

Figure 84.1 Annual (left) and cumulative (right) mass balances based on World Glacier Monitoring Service data.

bined with a kinematic Global Positioning System (GPS) is applied for monitoring thickness and volume changes of very large glaciers, which are the main melt-water contributors to ongoing sea-level rise (Arendt et al., 2002).

The two decades 1980-2000 show a clear trend of increasingly negative balances with average annual ice thickness losses of a few decimetres. Because unchanged climatic conditions would cause mass balances to approach zero values, constantly non-zero mass balances reflect continued climatic forcing. The observed trend of increasingly negative mass balances is consistent with an accelerated trend in global warming and correspondingly enhanced energy flux towards the earth surface. There is considerable spatio-temporal variability over short time periods: glaciers around the North Atlantic, for instance, exhibited considerable mass increase during the recent past and the sensitivity of mass balance and melt-water runoff from glaciers in maritime climates is generally up to an order of magnitude higher than for glaciers in arid mountains (Oerlemans & Fortuin, 1992; Oerlemans, 2001). Statistical analysis indicates that spatial correlations of short-term mass balance measurements typically have a critical range of about 500km (Cogley & Adams, 1998; Rabus & Echelmeyer, 1998) but tend to increase markedly with increased length of time period under consideration (as it applies to meteorological variables in general): decadal to secular trends are comparable beyond the scale of individual mountain ranges, with continentality of the climate being the main classifying factor (Letréguilly & Reynaud, 1990) in adition to individual hypsometric effects (Furbish & Andrews, 1984; Tangborn et al., 1990).

Cumulative changes in glacier length are strongly enhanced and easily measured but indirect, filtered and delayed signals of climate change (Oerlemans, 2001). They represent an intuitively understood and most easily observed phenomenon to illustrate the reality and impacts of climate change. Trends in long time-series of cumulative glacier length represent convincing evidence of fast climatic change at a global scale, as the retreat of mountain glaciers during the 20th century is striking all over the world (Haeberli et al., 1998a). Total retreat of glacier termini during the

20th century is commonly measured in kilometres for larger glaciers and in hundreds of metres for smaller ones.

Characteristic average rates of glacier thinning (mass loss) can be calculated from cumulative length-change data using a continuity approach over time periods corresponding to the dynamic response time of individual glaciers (Johannesson et al., 1989; Haeberli & Hoelzle, 1995; Haeberli & Holzhauser, 2003; Hoelzle et al., 2003). Characteristic values are a few decimetres per year for larger temperate glaciers in wet coastal climates and centimetres to a decimetre per year for small glaciers and glaciers in dry continental areas with firn areas below melting temperature (Fig. 84.2). At retreating glacier termini, the total secular surface lowering is up to several hundred metres. The apparent homogeneity of the signal at the century time-scale, however, contrasts with great variability at local/regional scales and over shorter time periods of years to decades (Letreguilly & Reynaud, 1990). Intermittent periods of mass gain and glacier readvance during the second half of the 20th century have been reported from various mountain chains (IAHS(ICSI)/UNEP/UNESCO, 1988, 1993, 1998), especially in areas of abundant precipitation such as southern Alaska, Norway and New Zealand. Glaciers in the European Alps, on the other hand, have lost about 50% of their original volume between the middle of the 19th century and 1970-1980 when systematic glacier inventories were compiled (Haeberli & Hoelzle, 1995). Rates of change and acceleration trends comparable to the ones observed during the past 100 yr must have taken place before, within the framework of Holocene glacier fluctuations and, hence, during times of weak anthropogenic forcing. In analogy to the glacier shrinkage documented during the 20th century, the Holocene record of Alpine glacier advance/retreat (Fig. 84.3) mirrors a (regional, hemispherical global?) pre-industrial variability of integrated secular to millennial energy flux towards or from the earth surface. As indicated by the finding of the Oetztal ice man, the 'warm' or 'high-energy' limit of this Holocene variability range may now have been reached in the Alps and possibly in other mountain regions, too. In such cases, continuation or acceleration of the observed trend could soon lead to conditions beyond those occurring during the Holocene precedence.

Figure 84.2 Characteristic mass balances during the 20th century as a function of glacier size (top) and region (bottom); values reconstructed from cumulative length change and continuity. (From Hoelzle et al., 2003.)

With the information in detailed glacier inventories (highest and lowest point, area and length), continuity approaches in combination with assumed step changes in mass balance (Haeberli & Hoelzle, 1995) can be applied for calculations of climate-change effects over time periods of a few decades. This rough and simple approach enables realistic quantitative estimates for entire mountain ranges. For the European Alps, a loss in ice volume of at least one-quarter is estimated to have occurred since the 1970s (Haeberli et al., 2002). The extremely hot and dry summer of the year 2003 alone may have caused the melting— within one single year—of some 5% or even more of what is left by now (Frauenfelder et al., in press). With a realistic scenario of future atmospheric warming, almost complete deglaciation could occur within decades, leaving only some ice on the very highest peaks and in thick but downwasting rather than retreating glacier tongues (Table 84.1).

The complex chain of dynamic processes linking glacier mass balance and length changes is at present numerically simulated for only a few individual glaciers that have been studied in great detail (e.g. Greuell, 1992; Oerlemans & Fortuin, 1992; Schmeits & Oerlemans, 1997). A new possibility is to dynamically fit mass-balance histories to present-day geometries and historical length-change measurements of long-observed glaciers using time-dependent flow models (Oerlemans et al., 1998). This approach not only provides important insights concerning mass balances during past periods that are not documented by direct

Figure 84.3 Chronology of Holocene glacier length changes in the Swiss Alps. (Compiled by M. Maisch.)

measurements, but also indicates details of potential time-dependent future evolution including feed-backs from effects of flow dynamics. Wallinga & Van de Val (1998) show that Rhonegletscher could disappear within decades if the presently observed trend continues or accelerates. The extensive modelling study by Oerlemans et al. (1998) indeed confirms that this could be the case for many if not most other glaciers of the current worldwide mass balance network (Fig. 84.4).

84.5 Needs and recommendations

Worldwide monitoring of glacier changes can build on more than 100yr of systematic/coordinated observation within the framework of international scientific collaboration, a wealth of excellent—even though still by far incomplete—information, a highly developed understanding of the processes involved and well-reflected integrated/multilevel operational strategies. The task of documenting potentially dramatic developments in remote areas thereby represents a challenge that can be met only by using the best available techniques and concepts. The potential of highresolution satellite imagery with stereo capacity and in combination with geoinformatics for automated image processing or the use of atmosphere-ocean general circulation models (AOGCMs) for analysis and interpretation of the observed data are prominent examples of avenues for research and development.

The recent availability of high resolution Landsat-7 and ASTER images together with new methods for automated analysis based on GIS techniques in digital inventories of glaciers (Bishop et al., 2004) not only opens new possibilities for upgrading preliminary inventories and repeating earlier inventories in view to assessing regional and global developments but also provides important information on impacts such as rock/ice avalanches or hazards from glacier lakes (Kaab et al., 2003). The interpretation of regional aspects is assisted by the use of statistically downscaled AOGCMs together with seasonal sensitivity characteristics (SSC) on mass-balance models of intermediate complexity. A corresponding study by Reichert et al. (2001) demonstrates that mass balances in Norway and Switzerland, respectively, are highly correlated with decadal variations in the North Atlantic Oscillation (NAO). This mechanism, which is entirely due to internal variations in the climate system, can explain the strong contrast between recent mass gains for some Scandinavian glaciers as compared with the marked ice losses observed in the European Alps.

Regional scaling with advanced AOGCM calculations reflects part but not all of current process understanding. In particular, two fundamental physical aspects still await inclusion into simulations and assessments: the firn/ice temperature effect and the size/dynamics effect. Firn warming relates to latent heat exchange involved with percolation and refreezing of surface melt-water within cold accumulation areas; this process makes the rate of firn warming considerably higher than corresponding air temperature change (Hooke et al., 1983; Haeberli & Alean, 1985). Once the firn becomes temperate, mass loss starts taking place with continued warming of the air. This means that the mass-balance sensitivity of large and still cold firn areas in the Canadian Arctic or in Central Asia, etc., could (i) strongly increase during the coming decades and thereby (ii) reduce the regional differences in sensitivity. The large and relatively flat glaciers around the Gulf of Alaska or in Patagonia, which produce the most important melt-water contribution to sea-level rise, have dynamic response times beyond the century scale and cannot dynamically adjust by tongue retreat to rapid forcing but rather waste down with little area loss. This, in turn, causes the mass balance/altitude feed-back

Table 84.1 Analysis of glacier inventory data for the European Alps (From Haeberli & Hoelzle, 1995, updated.)

Situation 1970-1980

Simulation (moderate warming scenario)

Total glacierized area 1970-1980

2909 km2

Area reduction 1970-1980 to 2025

ca. 30% of 1970-1980

Total glacier volume 1970-1980

ca. 130 km3

Mass loss 1970-1980 to 2025

ca. 50% of 1970-1980

Sea-level equivalent

ca. 0.35 mm

Area reduction 1970-1980 to 2100

ca. 90% of 1970-1980

Number of glaciers > 0.2 km2

1763

Mass loss 1970-1980 to 2100

ca. 95% of 1970-1980

Average mass balance 1850 to 1970-1980

-0.25 m yr-1

Average mass balance 1980-2000

-0.65 m yr-1

Area reduction 1850 to 1970-1980

ca. 40%

Mass loss 1850 to 1970-1980

ca. 50% of 1850

Mass loss 1980-2000

>25% of 1970-1980

Mass loss 2003 alone

ca. 5-10% of 2000

tri

nlines

<l

simulated jy 1748 profile

19Ë

3 profile

1_

0 2 4 6 a 10 12 14 16 Distance along flow line (km)

0 2 4 6 a 10 12 14 16 Distance along flow line (km)

, . . 1;

no change

-

10.01 K a"'; dp]

_

(0.01 K a 'i _

10.02 K a"1; dP|

.V * i; E i

|0 02K a ']

1950 2000 2050 2100 2150

Year AD

Graphics by J, Oerlemans

Figure 84.4 Dynamic fitting of Nygardsbreen and model calculation for future scenarios. (From Oerlemans et al., 1998.)

to become important. A cumulative surface lowering of about 50-100m within a century or so could, indeed, easily increase the mass-balance sensitivity by a factor of two, correspondingly doubling the surface lowering and, hence, leading to a runaway effect. The corresponding growth in size of the ablation area on such glaciers would probably by far overcompensate the effect of shrinking total areas on small glaciers elsewhere. This means that the sensitivity of the main melt-water producers with respect to

The international GLIMS project is a global consortium of universities and research institutes, coordinated by the U. S. Geological Survey (USGS) in Flagstaff, Arizona, whose purpose is to assess and monitor the Earth's glaciers from space. Specifically, the objectives of the GLIMS project are to ascertain the extent and condition of the world's glaciers so that we may understand a variety of Earth surface processes and produce information for resource management and planning. These scientific, management and planning objectives are supported by the monitoring and information production objectives of the United Nations scientific organizations (Kieffer et al., 2000; Bishop et al., 2004).

The GLIMS project entails:

1 comprehensive satellite multispectral and stereo-image acquisition of land ice;

2 use of satellite imaging data to measure interannual changes in glacier area, boundaries and snowline elevation;

3 measurement of glacier ice-velocity fields;

4 assessment of water resource potential;

5 development of a comprehensive digital database to inventory the world's glaciers, with pointers to other data and relevant scientific publications—the database is developed and located at the National Snow and Ice Data Center (Boulder, CO).

sea-level change is likely to strongly increase during the coming decades and strengthen regional differences accordingly. The effects on sea level, however, would be reduced to some degree by the fact that important parts of such large maritime melt-water producers are below sea level.

These examples clearly illustrate the key to a successful glacier observation programme: the combination of mapping, monitoring and modelling with advanced process understanding.

This work and the global image archive at the EROS Data Center (Sioux Falls, SD) will be useful for a variety of scientific and planning applications (Bishop et al., 2004).

The GLIMS project will primarily utilize multispectral imaging from the Landsat TM and ETM+ series, and the new ASTER sensor. Landsat TM and ETM+ data represent a well-established 'working horse' for glacier inventorying and monitoring from space (Kaab et al., 2002; Paul et al., 2002). The ASTER sensor, available since 2000 onboard the NASA Terra spacecraft, opens additional possibilities for glacier observation. The spectral and geometric capabilities of ASTER include three bands in VNIR (visible and near infra-red) with 15 m resolution, six bands in the SWIR (shortwave infra-red) with 30 m resolution, five bands in the TIR (thermal infra-red) with 90 m resolution, and a 15-m resolution NIR along-track stereo-band looking backwards from nadir. The stereo band 3B covers the same spectral range as the nadir band 3N. Of special interest for glaciological studies are the high spatial resolution in VNIR and the stereo and pointing capabilities of ASTER. With topography being a crucial parameter for the understanding of high-mountain phenomena and processes, digital elevation models (DEMs) generated from the ASTER along-track stereo-band are especially helpful (Fig. 85.1; Kaab, 2002). Imaging opportunities by ASTER are governed by Terra's 16-day nadir-track repeat period and the fact that the ASTER

Was this article helpful?

0 0

Post a comment