GIS/EM4 - Modelling Peruvian Deglaciation

Modelling Peruvian Deglaciation

GIS/EM4 No. 38

Bryan G. Mark

Abstract

This project utilizes GIS technology to model both past and present aspects of deglaciation. Begun as part of an investigation to reconstruct past glacial activity, this project has developed into an integrated environmental modeling effort to evaluate the hydrological implications of modern deglaciation, and assist in geohazard mapping. A digital elevation model is developed by digitizing 1:25,000 and 1:100,000 scale topographic maps in Arc Info, v. 7.2 (UNIX). Digital elevation data are combined with geologic coverages of glacial geomorphologic features and surface material mapped with differentially corrected GPS measurements. Past glacier extent is modeled based on the constraints of present-day topography and basic dynamic flow laws of glacial ice. Using hydrological networking, glacier watersheds are delineated and percentage of modern glacial coverage established from integrated satellite imagery (Landsat TM and SPOT). Available meteorological data are incorporated to make calculations of mass balance to glaciers and model percent contribution to stream runoff. Evaluating the relative loss of ice for glaciers of different orientation is important for determining the factors controlling glacier mass balance. Many of these factors are closely coupled with the surrounding topography, and can be better constrained within a GIS. Using a DEM, the receipt of solar radiation is computed for each grid cell on a glacier surface.

Keywords

Peru, deglaciation, volumetric rates, hydrologic impact, geologic hazards, global warming, paleoclimatology, DEM, glacier modeling

Introduction

Most of the world's tropical glaciers are in the Cordillera Blanca, Peru, where they are melting at significant rates (Kaser and Georges, 1997; Hastenrath and Ames, 1995). Deglaciation is a climatically forced, spatial redistribution of the hydrological mass budget of serious concern to both the Peruvian engineer, planning for water resource management and geologic hazards, as well as the Earth Scientist interested in global climate dynamics. Based on research by the Andean hydroelectric companies, there is a pressing need to calculate and monitor the magnitude of glacial meltwater contribution to modern water supply. The Santa river drains the Cordillera Blanca, and provides the source of water for drinking, irrigation and hydroelectric power, all of utmost priority for present and future development of the region. Moreover, the deglaciation is producing dangerous pro-glacial lakes in unconsolidated sediments that present serious danger for residents in this tectonically active region (Ames, 1998). Devastating landslides have occurred in the past and continue to threaten the region as a result of falling glacier ice and moraine ruptures inducing catastrophic drainages of these lakes. Understanding the nature and spatial extent of deglaciation requires an integration of topography with glacial changes over time and space as well as climatic variables to which a GIS-based modeling approach is well-suited.

Modern deglaciation provides the context for both applied and theoretical research foci to which GIS-based modeling. Understanding the relative volumetric rates of loss of ice is important to climate research. Using GIS datalogging GPS receivers, glacial geomorphologic features are readily mapped. The GIS provides an effective means to manage a wide range of geo-spatial variables, and facilitate the modeling of both past deglaciation, but also predict future hydrological impact of this phenomenon in an important developing region. Not only do the tropical regions seem to be particularly sensitive predictors of global climate trends (Bradley, 1996), but as home to over 75% of the world's human inhabitants likely to face the consequences of global climate change (Thompson, 2000), the tropics are a critical region to implement effective environmental modeling.

Problem statement

Two central hypotheses are tested:

Modern deglaciation has been accelerated to rates unprecedented in the past as a result of enhanced tropospheric temperatures (potentially influenced anthropogenically).
The volume of meltwater produced by wasting glacier ice provides a significant contribution to surface water runoff.

The exploration of these central questions calls for the integration of spatial and temporal information well suited for the GIS environment. Further practical dimensions to be developed include geo-hazard mapping.

Study Site

The Cordillera Blanca is located in the north-central Andes of Peru between 9-11 degrees S latitude. The mountain range extends over 130 km along the eastern boundary of the Rio Santa watershed, and contains the largest glacier-covered area in the tropics (Kaser et al., 1990). The climate is typical of the tropical highlands, showing larger variations in diurnal temperature than mean temperature change over the seasonal cycle (i.e. Hastenrath, 1991). Annual precipitation shows distinct seasonal variations with oscillations in the position of the Intertropical Convergence Zone (ITCZ). This seasonal climate regime confines mass accumulation on the glaciers of the central Andes almost exclusively to the wet season, while ablation occurs on glacial tongues throughout the year (Kaser et al., 1990).

The Glacier Inventory of Peru (Hidrandina, 1988) described the total area of glaciers in the Cordillera Blanca to be 723.37 km2, based on aerial photography from 1962 and 1970. A recent study based on Landsat satellite imagery has shown the present ice coverage to be 611.48 km2, equal to an overall reduction in glacier area of 15.46% over the interval between the 1962-70 inventory and the 1997 imagery (INAGGA, 1998). This glacier coverage is divided between three different drainage systems, with the Rio Santa draining the majority of glaciated terrain (503.11 km2 of glacier area) to the Pacific Ocean. In its entirety, the Rio Santa watershed covers 12,200 km2 (ONERN, 1972). However, the location of a hydroelectric power plant at Huallanca delimits the upper Rio Santa watershed (referred to as the Callejon de Huaylas) to an area of 4,901 km2 that will be the focus of this analysis.

Approach/Methods

Traditional methods of mapping deglaciation include terrestrial photogrammetry and surveying. These methods have had limited use over the latter part of the 20th century in the Cordillera Blanca, and are restricted in spatial extent. To extend the analysis of deglaciation spatially and temporally, regional topography was digitized to form a base map and DEM. Precise field mapping of modern glacier ice and ancient glacier moraines using differentially-corrected GPS point measurements allowed for the modeling of volumetric deglaciation rates for different time intervals (Figure 1). Topography is also used to delineate watersheds and calculate slope and aspect. Relative rates of deglaciation were computed for glaciers of different aspect. Incident solar radiation was modeled to test the energetic forcing on recent melting. On the larger scale, relative contributions of glacier melt water were calculated as a function of differential glacier coverage for different watersheds.

Figure 1: GPS mapped moraine positions draped over a hill-shade view of the DEM.

Findings

Modern volumetric deglaciation is accelerated well beyond rates estimated for discrete intervals in the late Pleistocene and Holocene (Figure 2). Glacier aspect seems to have a significant influence on late 20th century deglaciation (Figure 3). This is consistent with solar radiation modeling, indicating a potential shadowing effect of convective cloud cover. Watersheds with more glacier coverage contribute significant runoff ("negative storage") to surface water hydrology in the dry season as a direct result of modern deglaciation (Figure 4).

Figure 2. Table and chart showing that the modern measured rate of deglaciation (circled) is much greater than the estimated rates for previous deglaciation intervals. Also plotted are the estimated duration times of the glaciers, if the modern rate is applied to the modeled volumes.

Figure 3. (top): Glaciers of different aspect (SW, S, and E). In blue is the 1962 position, and white represents the 1999 position. GPS point measurements are shown with dots. (bottom): Gueshque East with an easterly aspect showed slightly more percentage change in volume.

Figure 4. Storage term calculated for watersheds of different glacier coverage: Querococha=4%; Llanganuco=35%; Olleros=12%; Chancos=22%; Paron=47%.

Discussion/Conclusion

While the deglaciation of the tropical Andes is not a new phenomenon, the rates of deglaciation seem accelerated in modern times. It is inconclusive whether or not the enhanced rates are a function of troposheric warming, as uncertainties in the volumetric modeling of deglaciation are large. However, the results suggest that previously observed shadowing effects may be less obvious, indicating an enhanced temperature effect. Glacier meltwater is seen to contribute a significantly (over 10%) to surface water runoff. This effect is especially important during the annual dry season, and for watersheds with larger percentage of glacier coverage. This work illustrates an excellent application of GIS-based modeling to both a scientific query and civil engineering issue.

Recommendations for Future Work

Focus on the application of radar imagery to reconstruct modern glacier volume loss over larger spatial scales. Also, integration of surface and bedrock geology to simulate landslide hazards as a result of new forming moraine-dammed lakes.

References

Ames, A., 1998: A documentation of glacier tongue variations and lake developments in the Cordillera Blanca, Peru. Zeitschrift fur Gletscherkunde und Glazialgeologie, 34 (1), 1-36.

Bradley, R. S., 1996: Are there optimum sites for global temperature reconstruction? NATO ASI Ser. I 41, 603-624.

Hastenrath, S., 1991: Climate dynamics of the tropics. Kluwer Academic Publishers, The Netherlands.

Hastenrath, S. and A. Ames, 1995a: Diagnosing the imbalance of Yanamarey glacier in the Cordillera Blanca of Peru. Journal of Geophysical Research, vol. 100 (D3), 5105-5112.

Hidrandina S. A., 1988: Glacier Inventory of Peru (Parts 1 and 2). Publicacion auspiciada por el Consejo Nacional de Ciencia y Tecnologia, 275 pp.

INAGGA (Instituto Andino de Glaciologia y Geoambiente), 1998: Vulnerabilidad de los recursos Hídricos de Alta Montaña, Unpublished report.

Kaser, G., A. Ames, and M. Zamora, 1990: Glacier fluctuations and climate in the Cordillera Blanca, Peru. Annals of Glaciology; 14: 136-140.

Kaser, G. and Ch. Georges, 1997: Changes in the equilibrium-line altitude in the tropical Cordillera Blanca, Peru, 1930-50, and their spatial variations. Annals of Glaciology, 24, 344-349.

ONERN (Oficina Nacional de Evaluacion de Recursos Naturales), 1972: Inventario, evaluacion y uso racional de los recursos naturales de la costa: Cuencas de los rios Santa, Lacramarca y Nepena. Vol. 1., Lima, Peru.

Thompson, L. G., 2000: Ice core evidence for climate change in the Tropics: implications for our future. Quaternary Science Reviews, 19, 19-35.

Author

Bryan G. Mark, Department of Earth Sciences
Syracuse University, 204 Heroy Geology Laboratory, Syracuse, NY, 13244, USA.
bgmark@syr.edu

Tel: +1-315-443-2672, Fax: +1-315-443-3363.