Investigations on Mass Balance and Dynamics of Moreno Glacier based on Field Measurements and Satellite Imagery

Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaten an der Leopold-ranzens-Universitat Innsbruck

eingereicht von

Mag. Martin Stuefer

Innsbruck) im November 1999

11 ., . . ;; )C .• I 2. P.; Abstract The mass luxes and dynamics of Perito Moreno Glacier have been investigated by means of ield mesurements and remote sensing techniques. Moreno Glacier, covering an area of 257.3 km2, is one of the main estern outlet glaciers of the Southern Patagonian Iceield (SPI) . The climate in Patagonia is characterized by westerly winds and wet air rom the Paciic Ocean, which cause abundant precipi­ tation at the SPI; the formidable topographic barrier of the Andes produce sharp local contrasts of climate. High resolution optical images from Landsat and SPOT, as well as SAR images from ERS and RADARSAT satellites were used together with the cartographic maps to analyze the glacier boundaries and to estimate the position of the equilibrium line. The motion ield over the glacier terminus was derived from SIR-C data acquired during a shuttle light in October 1994, applying radar interferometry and amplitude cross-correlation. The ield work was carried out on Moreno Glacier in several campaigns between November 1995 and March 1999. It included ablation and ice motion mesurements at three proiles using stakes, the installation and maintenance of a climate station, and echo sounding of the lake depth close to the glacier front. Using the seismic relection method the ice thickness was mesured along two transverse proiles of the glacier terminus. The measurements along the upper proile, about 7.5 kilometers distant from the calving front, revealed a subglacial trough with an approximately parabolic shape. The annual transport of mass through this proile is about 707 109kga-1 . The maximum ice depth of 684 m indicated a glacier bed rising from 200 m below sea level towards the calving front. The majority of SPI glaciers hs been subject to glacier retreat during the last 40 years. The climate records in Patagonia do not reveal a clear long-term climatic trend. The climate stations in the vicinity of the SPI are sparse, and the few long-term records of air temperature and precipitation are partly inhomogeneous. The few temperature records suggest a warming trend for the region est of the iceields since about 1940. The maritime inluence on Moreno Glacier is shown by continuous meteorological measurements, which are available since November 1995 from the automatic climate station close to the glacier terminus. The mean monthly temperatures of the years 1996, 1997 and 1998 ranged from 0.7°C to 10.1°C. The total ablation of Moreno Glacier is estimated by calculating the mass lx through glacier cross-section and by spatial melting extrapolation. Assuming the glacier to be close to an equilibrium state, a speciic annual net accumulation of 5250 ± 660 kgm-2 is obtained. Small inter-annual variations of ice velocity, a comparatively steep surface in the region of the equilibrium line, and the high ratio of 0.36 of calving lux to net accumulation are probably the resons for the remarkable stability of Moreno Glacier, which is in contrast to other glaciers in this regiOn.

a Contents

Acknowledgements iv

Introduction 1 1 1.1 Background 1 1.2 Patagonian Ice Fields 2 1.3 Previous observations 0 2

Topographic and remote sensing data base 2 6 201 Topographic data of the Moreno area 6 202 Remote sensing data 0 0 0 0 0 0 0 0 0 0 0 0 0 7 20201 Landsat, SPOT and DISP images 9 20202 ERS SAR and RADARSAT acquisitions 10 20203 SIR-C/X-SAR experiment 10 202.4 Aerial photographs 13

Glacier characteristics 3 14 301 Glacier history 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 14 302 Glacier area and estimation of the equilibrium line altitude 17 303 Fluctuations of glacier terminus 0 0 0 0 0 0 18 3.4 Description of glacier surface 21

Field measurements 4 28 401 Climate station 30 402 Ablation 30 403 Stake velocities 44 4.4 Ice thickness 52 405 Geodetic lines and points 55 406 Lake depth 61

Regional climatology 5 64 501 Meteorological stations and data 64 502 Air temperature 66 503 Precipitation 70 5.4 Wind 75 Contents

5.5 Pressure, humidity and radiation at Moreno Base Station ...... 76

Surface motion by means of SAR 6 83 6.1 Methods ...... 83 6.1.1 Interferometric motion analysis 83 6.1.1.1 Phase coherence 85 6.1.1.2 Motion analysis 85 6.1.2 Analysis of low direction 93 6.1.3 Amplitude cross-correlation ...... 94 6.1.4 Comparison of the velocity ield from INSAR and amplitude cross-correlation ...... 95 n 6.1.5 Measurement uncertainties ... 98 6.2 The motion ield ...... 98 6.2.1 Comparison to stake velocities 99

On the dynamic behaviour of Moreno Glacier 7 102 7.1 Velocity variation with depth . 102 7.2 Glacier sliding ...... 106

Mass balance 8 112 8.1 Ablation ...... 112 8.1.1 Areal extrapolation of stake measurements ...... 113 8.1.2 Estimation of multi-year ablation with degree-days . 115 8.2 Calving rate ...... 116 8.2.1 Introduction ...... 116 8.2.2 Determination of calving speed . 117 8.2.3 Ice export ...... 117 8.3 Ice discharge through B proile . 121 8.4 Accumulation ...... 123

Comparison with other freshwater calving glaciers 9 125 9.1 Characteristics of SPI glaciers ...... 125 9.1.1 Ameghino Glacier ...... 126 9.2 reshwater calving glaciers in other regions . 132 9.2.1 Nordboglacier ...... 132

Summary and conclusion 10 135 Bibliography 139 A Chronological description of ield activities 147 A.1 Field measurements in November/December 1995 . 147 A.2 Field mesurements in March/ April 1996 . 148 A.3 Field measurements in November /December 1996 . 149 Contents

A.4 Field measrements in March/ April 1997 . 150 A.5 Field measurements in November 1997 .. . 150 A.6 Field measurements in March/ April 1998 . 151 A. 7 Field measurements in March/ April 1999 . 151

B Positions measured with GPS 152 B.1 Stakes ...... 152 B.2 Seismic stations . 152

C Meteorological station 159 C.0.1 Technical details ...... 159 C.0.2 Data capture, processing and transfer . 15lll9 C.0.3 Sensors ...... 160

D Seismograph 161 E Lake depth measurements 162 Acknowledgements

The investigations described in this thesis have been carried out within the coop­ eration of a number of people all of whom I am grateful to. My greatest debt of gratitude is to my supervisor Helmut Rott. He gave me insights into a very fsci­ nating and interesting ield. His valuable proposals and his tireless eforts were the driving force to write and inish this thesis. I want to thak Pedro Skvarca, Direcci6n N acional del Antartico of Argentina, for the collaboration and fruitful discussions. Pedro Skvarca's local relationships, knowledge of the region, and his eforts to provide all necessary facilities were of inestimable value for the ield activities. I am grateful to Heirich Miller, the second supervisor of the thesis, for his helpful comments and the uncomplicated logistic support by the Alfred Wegener Institute for Polar and Marine Research. Thanks to all my friends and colleagues of the remote sensing group at the Institut fUr Meteorologie und Geophysik, Universitat Innsbruck, for numerous dis­ cussions and creating nice and productive atmosphere. Special thanks to Andreas Siegel, who provided and applied the interferometry-software and showed a lot of patience in discussions. Special thanks to Thoms Nagler and Wolfgang Rack, who contributed to this work by joining me during ield mesurements on Moreno Glacier, by giving valuable comments, by proof-reading and providing programming abilities to improve the results. The ield activities were carried out with the help of several people. Among others, I wish to thank Alfons Eckstaller from the Alfred Wegener Institute for Polar and Marine Research, who worked in the ield under hardest circumstances and analyzed seismic data. I want to thank Teodoro Tocons, Juan Carlos Quinteros and Stephan Hoinger. Thanks also to Luciano Pera and the Minitrekking Organisation, to Nicolas Benedetti, Berni Roil, and all the mountain guides for their support and the enjoyable hours at the Moreno Base Camp. Especially, I am grateful to all people who helped in .the rescue operation after my accident on Moreno Glacier on 20 March 1996. I like to thank Martin unk for the valuable discussions and Manfred Nothegger for proof-reading. I am very grateful to Erich Heucke for providing an excellent steam-drill, and to ranz Weitlaner for all the electronic solutions. The depart­ ment head Michael Kuhn and numerous people of the Institut fir Meteorologie und Geophysik is gratefully acknowledged.

IV Chapter 1

Introduction

1.1 Background

Glaciers are climate indicators which ofer remarkable information on former en­ vironmental conditions. In view of increasing greenhouse gases and the difer­ ent regional impacts of global warming, glaciological investigations in the South­ ern Hemisphere are of signiicant interest. Despite the fact that the Patagonian Iceields are the largest ice masses in temperate climate of the Southern Hemi­ sphere, only few glaciological investigations have been carried out in Patagonia so far [Warren and Sugden, 1993]. Field research has focused on only a few outlet glaciers. The Patagonian Iceields hold key palaeoenvironmental information. The study of the glacier dynamics and the glacier mass balances is of main importance for the understanding of past environmental changes. This thesis was supported by the Austrian Science Fund (FvVF) projects Nr. P10709 and P12923 GEO. The ield work was carried out in cooperation with the Instituto Antartico Argentino (IAA). The Alfred Wegener Institute for Polar and Marine Research (AWl) was in charge of seismic measurements of Moreno Glacier. The satellite data were received rom the SIR-C/X-SAR Experiment of NASA/DLR/ ASI, the European Space Agency (ESA) with the project Nr. A02- A101 and from the Radarsat's Application Development and Research Opportunity (ADRO) program. Moreno Glacier is located at 50°30' southern latitude and 73°10' western longi­ tude in a region, where extreme weather conditions and diicult access restrict ield measurements. Wet air from the Paciic Ocean and the topographic barrier of the Patagonian Andes, lying athwart to the prevailing westerly winds, result in strong west-east precipitation gradients. The mountain range is characterized by extended iceields, with average elevations between 1500 and 2500 m a.s.l., and numerous outlet glaciers. Large glaciers, the largest of which covers about 1265 km2 in area [Aniya et a., 1996], transport ice masses from the iceields to elevations close to or at the sea level. The complexity of the dynamics of Patagonian glaciers is evident from the diferent behaviour of various glaciers.

1 Chapter Intoduction 2 1.

Calving glaciers play an essential role for the mass balance and dynamics of the Patagonian Iceields. Although most glaciers have been subject to retreat during the last 50 years, Moreno Glacier has been almost stable after an advance at the beginning of the century. The glacier terminates with calving velocities of partly more than 2 md-1 into Lago Argentino. Ioreno Glacier is noted for spectacular water outbursts after damming up the southern arm of Lago Argentino, Brazo Sur, by reaching the opposite lake bank. About 21 damming episodes occurred between 1917 and 1988.

1.2 Patagonian Ice Fields

The Patagonian Ice Fields are accounting for more than 60 percent of the Southern Hemisphere's glacial area outside of Antarctica. The Northern Patagonian lceield (NPI, Hielo Patag6nico Norte) is centered around 47°S and 73°30'W and extends over an area of about 4200 km2 [Warren and Sugden, 1993]. The major ice mass of Patagonia is stored in the Southern Patagonian Iceield (SPI, Hielo Patagonico Sur). The SPI stretches from 48°20'S to 51°3 0'S over a length of approximately 350 km. The average width of the SPI is 30-40 km, the narrowest part in 50°25'S has a width of 17 km (igure 1-1). Most glaciers extend from the Iceields down to low altitudes at or slightly above sea level. All of the main glaciers of the SPI, except two, calve into Paciic fj ords in the west and into proglacial lakes in the east. The low altitude of the ablation areas, the vicinity of the sea, and large mass turnovers are typical characteristics of these temperate glaciers. Field observations of frontal moraines, analyses of aerial photographs, with the oldest photographs dating back to the mid-1940s, and satellite image data revealed for SPI the predominance of glacier retreat since the last 50 years. Lliboutry used preliminary maps at a scale of 1:250000, published by the Instituto Geogniico Mil­ itar of Chile, to estimate the area of SPI at 13500 km2 [Lliboutry, 1956]. Re­ cent analyses using satellite images resulted in a total area of about 13000 km2 [Naruse and Aniya, 1992]. Due to the extreme weather and the rough topography of the Patagonian Andes, glaciological ield observations are sparse. Remote sensing techniques represent a useful tool for scientiic research in this region.

1.3 Previous ob servations

Previous glaciological ield work on Moreno Glacier has been restricted mostly to the glacier front (section 3.3) and to the lower part of the terminus. The irst documented visit to the region goes back to Hauthal in 1899 [Hauthal, 1904] . Moreno Glacier is noted for the repeated damming-up of Brazo Sur (chapter 3) since 1917. Most of the former expeditions to the region were interested in the damming, some pub­ lications describe spectacular water outbursts at the end of the damming episodes Chapter Introduction 1. 3

•JO.l •L

�¥ndall G.

Figure 1-1: Location map of the Southern Patagonian Iceield with the main glaciers and proglacial lakes in the Est.

Mt.ale

74•w Chapter Introduction 4 1.

[Heinsheimer, 1958] , [Liss, 1970]. One of the irst glaciological reports on Moreno Glacier was published by Rafo [1953] ; he estimated the total glacier area at 380 km2 , and the accumulation area at 255 km2 . Rafo's analysis was based on aerial photographs from 1937. A irst complete overview of the glaciers of the SPI was presented by Lliboutry [1956]. His book includes a sketch map of the SPI, which was based on U.S. aerial photographs acquired during the 2nd world war. rom this sketch map Heinsheimer derived a total area of 207 km2 for Moreno Glacier [Heinsheimer, 1958]. A similar number, 195 km2 , was published by Liss [1970] . More recent analyses are based on topographic maps published by the Institute Geograico Militar of Argentina (1989). Aniya derived a glacier area of 257 km2 from the map and aerial photographs taken in 1984 [Aniya and Skvarca, 1992]. Initial ice motion studies at Moreno Glacier were carried out by J.M. Rafo in 1948/49 and in 1950. Between 26 November and 2 December 1948 he set up a trans­ verse proile, consisting of 22 iron poles with 1.8 m length, located at a distance of approximately 5.5 km from the glacier front. The proile crossed about two-thirds of the glacier width starting from the orographically left margin [Rafo et al., 1953]. The positions of the poles were re-measured after 113 days on 24 March 1949 and after 491 days (two summers and one winter) on 6 April 1950. During the latter mesurements only 10 poles were found. A regular increase of velocities from the margin to the centerline was deduced, from 1948 to 1950 the fastest pole was dis­ placed by about 1.3 km, corresponding to an average motion of 2.64 m d-I, which is faster than the results of our mesurements at about the same positions (section 4.3). Repeated observations of selected Patagonian glaciers were initiated in 1967 by Japanese scientists. rom 1983 to 1986 and in the austral summers of 1990-91 and 1993-94 glaciological studies were performed through the Japan-Argentina-Chile joint Glaciological Research Project in Patagonia (GRPP). From 1983 to 1986 em­ phasis of the research on the NPI was at Soler and San Rafael glaciers, and on the SPI at Tyndall Glacier [Naruse and Aniya, 1992]. Subsequently, ield activities were initiated in November 1990 on Moreno, Upsala and Tyndall glaciers; the ield observations were typically carried out during periods of 1 to 3 weeks in early sum­ mer. The activities included geodetic measurements of vertical displacement and ice velocity at selected points of the ablation area, measurements of ice ablation with stakes, measurements of the frontal position and acquisition of meteorological data (section 5.2): About 500 m down-glacier from Rafo's proile, two survey lines arranged in the form of a cross were set up around 5 km rom the glacier front [Naruse et al., 1992]. The elevations of six marked srvey points along each line were measured on 25 November 1990. Flow velocities were deduced by repeat measurements of 11 survey points on 26 November 1990. The velocity rates increased from 0.38 md-1 near the right margin to about 2 md-1 at the center. Chapter Introduction 5 1.

The measurements of ice ablation and ice motion in 1990, 1991 and 1992 covered limited periods of the year. On 11 November 1993 aluminium poles were drilled into the ice at the position of for of the 1990 survey points. The pole positions and the heights above the surface were re-measured ive times in 1993 and in 1994 [Skvarca and Naruse, 1997] . These were the irst measurements over a complete year. The velocities rom November to December 1993 were slightly larger than the annual mean. The annual mean velocity near the center-line was 1.6 md-1. Elevation measurements during the expeditions of 1990-91 and 1993-94 and repeat measurements in April 1996 revealed no signiicant changes of ice thickness at all nine survey points [Skvarca and Naruse, 1997] . Within the GRPP the surface proile of a portion of the orographically right calving front at Brazo Sur was measured by a triangulation method; the measurements revealed heights above the lake level between 55 m and 77 m [Naruse and Aniya, 1992]. No direct measurements of accumulation at Moreno Glacier are available. Spa­ tially and temporally variable accumulation rates have to be expected due to the local topography and the strong winds. Schwerdtfeger assumed the annual pre­ cipitation of the SPI at heights above 1500 m a.s.l. to be at least 1.5 to 2 times the annual precipitation measured at coastal climate stations [Schwerdtfeger, 1958]; he estimated the annual precipitation sum at the western side of the SPI to be at least 7000 mm. In accordance with Schwerdtfeger, Escobar estimated a mean annual precipitation of 7000 mm over SPI and of more than 8000 mm on the central plateau of the Iceield [Escobar et al., 1992], these numbers were deduced from an estimation of the water balance of the SPI. The analysis of the only irn core taken rom the accumulation area of the SPI resulted in a much lower value [Aristarain and Delmas, 1993] . A mean value of 1200 mm water equivalent was de­ duced from the deuterium analysis and stratigraphy of the 13 m ice core, which was drilled at an elevation of 2680 m a.s.l. on Moreno Glacier. Possible reasons for this low value might be local anomalies in accumulation due to orography and wind, or uncertainties in the interpretation of the stratigraphic sequence. Deuterium levels in this maritime climate might not be subject to clear annual cycles, as increased deuterium levels may also be related to strong precipitation events with advection of warm oceanic air in winter. Chapter 2

Topographic and remote sensing data base

2.1 Topographic data of the Moreno area

A digital elevation model (DEM) of Moreno Glacier was generated using diferent sources. The elevation lines of two cartographic maps were digitized with auto­ mated and manual methods. Some control points and lines measured during the ield programs (chapter 4) by means of a Global Positioning System (GPS) were used to improve certain features of the maps. The glacier extent and the shore of Lago Argentino were derived from georeferenced Landsat and SPOT images (section 2.2.1). In 1982 the Institute Geogniico Militar (IGM) of Argentina issued 1:100000 scale maps covering the Argentinian part of the SPI and a small strip of Chilean territory. The IGM maps are in ransverse Mercator (TM, Gauss-Krueger) projection refer­ ring to the geodetic datum of Campo Inchauspe (=International 1924). The contour interval of elevation lines is 50 m. The map sheet "Glaciar Perito Moreno" (sheet number 5172-7) extends over 36 km from 50°20' to 50° 40' southern latitude, covering the major part of Moreno Glacier. The southernmost area of Moreno Glacier with altitudes above 2000 m a.s.l. is on the adjacent map sheet of "Lago Frias" (sheet number 5172-13) extending from 50°40' to 51 °00' southern latitude. The two maps were digitized, distortions due to the digitizing process were geo­ metrically corrected using the reference grid. rom the digitized maps a DEM, also in TM projection, was produced (igure 2-1), the speciications are shown in table 2.1. The original map datum was changed to the WGS84 datum in order to simplify the comparison with GPS measurements. Matching optical satellite images and vertical airphotos to the map revealed some map inaccuracies especially at the left margin of Moreno Glacier at elevations above 1200 m. Another error was detected at Cerro Perito Moreno (73d 04' 27" W, 50d 30' 55" S). Taking into account a height diference of 13 m between the altitudes of the WGS 84 ellipsoid and the reference height of the map, an altitude of 1734

6 Chapter Topogaphic and remote sensing data base 7 2.

Earth Ellipsoid WGS 84 (GPS) Georeferenced Units ransverse Mercator (Gauss-KrUger) Upper Left Corner 73d23'22.45" 50d20'58.83" w, s Upper Right Corner 72d59'46.46" 50d21'13.36" vv, s Image Centre 73d11 '49.77" 50d31'13.37" w, s Lower Left Corner 73d23'58.21" 50d41'12.07'' w, s Lower Right Corner 73d00'12.10" 50d41'26.77" w, s Latitudinal extent 37.5 km Longitudinal extent 28.0 km Horizontal raster 20 m x 20 m Table 2.1: Speciications of the DEM. m was measured with GPS for Cerro Perito Moreno in contrast to 1640 m a.s.l. the map. Elevation contours in this area were interpolated to reduce the errors of the DEM around Cerro Perito Moreno. Because the few GPS measurements, which were carried out in the Moreno region, revealed already some signiicant errors of the map, considerable elevation errors have to be expected also in other parts. The DEM, representing a rster image grid, was created by interpolating or illing in elevation data between the digitized contour elevations. A diagonal search algorithm (program grdint of PCI software) performed the grid interpolation.

2.2 Remote sensing data

Optical and radar images have been used. High resolution optical images, acquired by Landsat and SPOT satellites, as well s oblique photoprints from aircraft are available. Due to westerly winds from the Paciic forcing the wet air to rise on thme slopes of the Patagonian Andes, the SPI is often obscured by clouds and only few images from visible and infra-red sensors are available. Better temporal coverage is provided by Synthetic Aperture Radar (SAR) images, acquired by European Remote Sensing (ERS) and RADARSAT satellites, and by the Space Shuttle Endeavour. Radar data from Moreno Glacier became available within the following projects: ERS-1/ERS-2 Experiment A02.A101 of ESA on "Comparative investigations of climate sensitivity and dynamics of glaciers in Antarctica, Patagonia, and the Alps" . Radarsat ADRO Experiment of the Canadian Space Agency (CSA) on "Snow hydrology and glacier research in the Alps and in Patagonia" . SIR-C/X-SAR Experiment of NASA, the German Aerospace Research Center (DLR) and the Italian Space Agency (ASI): "The High Alpine SAR Experiment" .

•

•

• Chapter Topogaphic and remote sensing data base 8 2.

Figre 2-1: Digital elevation model representing a frontal view of the Moreno Glacier region towards W-S-W direction with the highest peak Cerro Pietrobelli (CP, 2950 m a.s.l.) . The mesh-distance in this igre is 500 m, vertical exaggeration 2.25. Chapter Topogaphic and remote sensing data base 9 2.

Satellite A-Date Instr. C-Lat. C-Lon Path Row LS-5 14 Jan 86 TM S50d17'00" W073d21'00" 231 95 SPOT 23 Aug 95 HRV-2 S50d22'08" W072d45'03" 680 452 Corona 15 Oct 65 Camera 50d20' W070d03' 1025(2) 10 s Table 2.2: Analysed Landsat-5 (LS-5) Thematic Mapper (TM), SPOT images and declassiied intelligence stallite photo (DISP); acquisition date (A-Date), center lat­ itude (C-Lat) and longitude (C-Lon), path and row of satellites.

Figure 2-2: Location of Landsat and SPOT scenes of table 2.2.

2.2.1 Landsat, SPOT and DISP images Two high-resolution images, acquired by Landsat-5 and SPOT, available for this investigation (table 2.2). The spectrum of the seven TM-bands ranges from 0.45 (band 1) to 12.50 ,m (band 7), the maximum spatial resolution is 30 m. An austral winter scene from 23 August 1995 ws acquired by SPOT with the "High Resolution Visible" (HRV-2) instrument in a single band in the visible part of the spectrum (0.51 - 0.73 ,m), the resolution is 10m. Less than 20% in each of the image of table 2.2 are obscred by clouds and their shadows. The Landsat TM image of 14 January 1986 shows the SPI between 49°30' and 51 (including Moreno Glacier; igure 2-2) in exceptionally good quality. Unfortunately Moreno Glacier is not completely covered by the SPOT scene and clouds afect the scene over parts of the glacier. Nevertheless, the image quality is very good in the lower glacier region with the calving tongue. In 1995 intelligence imagery acquired by the irst generation of U.S. photo- re­ connaissance satellites was declassiied by the U.S. government. More than 860,000 images of the Earth's surface, collected between 1960 and 1972, recan be obtained

o Chapter Topogaphic and remote sensing data base 10 2. from the U.S. Geological Survey (USGS). A Declassiied Intelligence Satellite Photo (DISP) acquired by the Corona satellite on 15 Oct. 1965 covers the whole Moreno glacier. The quality of the black-white negative ilm allows an accurate observation of Moreno Glacier's tongue, whereas clouds and lack of contrast due to snow cover afect higher parts of the glacier. The spatial details of this DISP image are sim­ ilar to Landsat TM (with 30 m resolution), though the best ground resolution is speciied at approximately 4.3 m.

2.2.2 ERS SAR and RADARSAT acquisitions A number of images covering Moreno Glacier were acquired by the Active Microwave Instrument (AMI) of ERS-1 and ERS-2 satellites. AMI SAR operates at C-hand (A =5.6 em/ f=5.3 GHz). In image mode the ERS-SAR provides high resolution images; the ERS PRI products are in ground range projection with a spatial reso­ lution of 28 m in ground range (nominal: <33 m) and 20.8 m in azimuth (nominal: <30 m)(with 3x1 looks). The available images cover the period from 16 February 1993 until 6 March 1998 (table 2.3), the ERS-1 images were acquired during the mis­ sion phases C to G (from: ESRIN usERServices: http://gds .esrin.esa.it/). Three ascending and two descending passes of ERS provide good coverage of the glacier. The location of the ERS rames is shown in igure 2-3. In addition, SAR images acquired by the Canadian RADARSAT 1 satellite are available from late austral summer and early autumn 1997 (table 2.3, igure 2-4). RADARSAT SAR operates also at C-hand, the two available images have been acquired in the standard beam mode with a nominal spatial resolution of 25 m in ground range and 28 m in azimuth direction (with 4 looks).

2.2.3 SIR- C /X-SAR experiment The SIR-C/X-SAR (Spaceborne Imaging Radar-C/X-Band Synthetic Aperture Radar) aboard the Space Shuttle Endeavour imaged parts of the SPI in April and October 1994. Repeat- pass SIR-C data acquired during the eleven-day mission from 30 September 1994 until 11 October 1994 are used for this investigation [et al., 1995]. SIR-C operated at 1-band (A =24.2 em) and C-hand (A =5.7 em), X-SAR at X-band (A =3.1 em). SIR-C image data covering the Moreno Glacier were acquired from nearly the same orbital position on 7, 9, and 10 October (table 2.4). The SIR-C data, centered at 50.6 degrees south latitude and 72.7 degrees west longitude, in­ clude all of the ablation area and the major parts of the accumulation area of Moreno Glacier. Figure 2-5 shows the coverage by SIR-C which was similar on all three days. X-SAR data show slightly diferent frame locations, the center longitudes are shifted 9 minutes eastward from 9 to 10 October 1994. The spatial resolution was identical in C- and L-band with 5.21 m along-track and 3.33 m in slant range (single look), the X-band resolution was 8.2 m (azimuth direction) and 7.3 m (slant range). Chapter Topogaphic and emote sensing data base 11 2.

Satellite A-Date Format Orbit rame rack Pass ERS 1 16 Feb 93 GEC 8314 46351 153 Des ERS 1 26 Aug 93 PRI 11048 46352 382 Des ERS 1 02 Sep 93 PRI 11142 61653 476 Asc ERS 1 22 Jan 95 PRI 18412 61654 1611 Asc ERS 1 01 Feb 96 SLC 23782 61655 247 Asc ERS 1 16 Oct 97 SLC 32706 46351 153 Des ERS 1 23 Oct 97 SLC 32800 61655 247 Asc ERS 2 02 Feb 96 SLC 4109 61655 247 Asc ERS 2 20 Oct 96 PRI 7845 61653 476 Asc ERS 2 24 Oct 97 SLC 13127 61655 247 Asc ERS 2 15 Feb 98 PRI 14765 46352 382 Des ERS 2 06 Mar 98 PRI 15037 46351 153 Des RADARSAT 15 Feb 97 6700 Des RADARSAT 06 Apr 97 7423 Asc Table 2.3: ES-1, ERS-2 and RADARSAT scenes covering Moreno Glacier. Ascend- ing {Asc) and descending {Des) pass, date of image acquisition (A-Date); superscript frame numbers refer to the following igre showing ERS-rames. Formats: Ellip- said Geocoded Image (GEC), Precision Image (PRI), Single Look Complex Image (SLC).

asc

52°S

Figure 2-3: Location ofES-1 and ERS-2 frames; the indicated numbers correspond to the diferent tracks {TR) and frames (r) of ERS images (table 2.3).

1: Tr153. Fr 4635 2: Tr382, Fr 4635 3:Tr476,Fr6165 - t---- · 4: Tr1611. Fr6165- - 5:Tr 247.. Fr 6165 12 Chapter 2. Topographic and remote sensing data base

74°W

Figre 2-4: Location of RADARSAT frames

.

+--�- -�·-­ .

52°S ------

Figre 2-5: rames of available SIR-C (red) and X-SAR (9 Oct. 1994, black) data

Datatake ID Incidence Angle

dt 121.30 (7 Oct 94) 34.320t, deg dt153.30 (9 Oct.94) 34.354 deg �-,.-. dt169.30 (10 Oct.94) 34.369 deg ...��

Table 2.4: SIR-C/X-SAR data identiication and SIR-C data incidence angle at the image center. Chapter Topogaphic and remote sensing data base 13 2.

Date Source Scale Section early 1947 IGM 1:38000 terminus 18 Nov 1968 IGM 1:65000 terminus 11 Mar 1970 IGM 1:70000 total 20 Dec 1984 FACH 1:70000 terminus Table 2.5: Aerial photographs covering parts of Moreno Glacier.

2.2.4 Aerial photographs

Vertical aerial photographs taken by the Instituto Geogniico Militar (IGM) of Ar­ gentina and the Servicio Aerofotogrametrico of the Fuerza Aereo Chileno (FACH) cover mostly the lower parts of Moreno Glacier. In March 1970 the IGM took aerial photographs from two light paths. They cover the total region of Moreno Glacier; the quality of this set is exceptional due to cloudless weather condition. The date of photographs and the nominal scale are listed in table 2.5. Chapter 3

Glacier characteristics

Moreno Glacier, one of the main eastern outlet glaciers of the SPI, stretches over a length of 30 km from the continental divide to Lago (lake) Argentino (180 m a.s.l.). The average height of the divide is about 2200 m with the highest point at Cerro Pietrobelli (2950 m a.s.l.) . The terminus extends in a valley conined by ridges which rise up to 2380 m a.s.l. in the south (Cerro Cervantes) and up to about 2000 m a.s.l. in the north. The front terminates with calving clifs of 50 to 80 m height in the southern arm (Brazo Sur, Brazo Rico) and Canal de los Tempanos of Lago Argentino (igure 3-1) , [Heinsheimer, 1958] , [Naruse and Aniya, 1992] . The central point of the snout is at about 50°28'20" S and 73°02'00" W. The water of Brazo Sur usually discharges near Peninsula Magallanes through a narrow channel or below the glacier into the Canal de los Tempanos. From 1917 until 1988 the glacier repeatedly dammed up Brazo Rico by reaching the opposite bank, the damming episodes were ended by spectacular bursts of water through the snout.

3.1 Glacier history

First glaciological observations of glaciers on the West side of Lago Argentino have been reported by Hauthal [Hauthal, 1904] , [Hauthal, 1911] , who in 1899 and 1900 visited the region on behalf of the La Plata Museum and of the Argentinian-Chilean border commission. When Hauthal arrived at the Moreno Glacier for the irst time in March 1899, he named the previously unknown glacier Bismarckgletscher (igure 3- 2) . In contrast to all other Patagonian glaciers visited by Hauthal, Moreno Glacier was in an advancing state. On 25 February 1899 he marked the position of the orographically right glacier margin, where the glacier touched Lago Argentino at the southern shore of Brazo Sur. Further observations on 4 March 1900, during Hauthal's second visit, revealed a glacier extension of 23.5 mat this point. Moraines and big stones conirmed this observation. Hauthal also measured the position of the front by means of a compass; it advanced from 1899 until 1900 by about 140 m. At that time the height of the calving glacier front reached approximately 30 m

14 Chapter Glacier chaacteristics 15 3.

Figure 3-1: Landsat TM (channel 4) image of Moreno Glacier from 14 Jan 1986, with topographic features and the main peas. The dotted line shows the approximate position of the equilibrium line (section 3.2) . Chapter Glacier chaacteristics 16 3.

Figure 3-2: A painting of E.T. Compton showing Moreno Glacier (Bismarck­ gletscher) from the opposite coast at Peninsula Magallanes; the painting was drawn after a photograph taken on 25 February 1899. above the lake level. In order to estimate the thickness of the glacier at the front also the water depth of a straight transverse proile ws mesured at a distance of a few hundred meters from the southern glacier ront. The mesurements revealed water depths up to 137 m (igure 3-5). The glacier advance continued in the years after Hauthals visits. Quensel, a member of the Skottsberg Swedish Expedition [Reichert, 1917], checked the frontal position in 1908. According to Reichert and Hicken the distance between the calving front and the coast at the Peninsula Magallanes decreased to 100 to 150 m until March 1914 [Reichert, 1917]. Liss published a summary of the observatios of Moreno Glacier from the irst visit in 1899 until 1970 [Liss, 1970]. Based on the hypothesis of Nichols and Miller [Nichols and Miller, 1952] an ex­ planation for the anomalously advancing behavior of Moreno Glacier was reported by Rafo [1953]: He assumed as a possible mechanism that glacier erosion changed the bed rock around the southern divide. This change led Moreno Glacier to tap the ice from the retreating ris Glacier. Rafo assumed additionally an ice low ac­ celeration due to erosion of rock thresholds at the bed. urther explanation factors were possible climate changes and small earthquakes. The hypothesis of Heinsheimer ws based on changes in ice low due to atm­ spheric warming [Heinsheimer, 1958]. He assumed that numerous rock outcrops became opposed dring warming periods, water lowing from the outcrops to the glacier ice penetrated to the bed by raising the basal low. Chapter Glacier chaacteristics 17 3.

3000

2500

2000 s :: 1500

1000

500

0 0 50 100 150 200 250 Area (n2) i

l

Figurei 3-3: Area altitude distribution of Moreno Glacier. The mean equilibrium line altitude (ELA) is estimated at 1170 m asl..

0 •I Lissw [1970] considered another hypothesis which dealt with the mass loss due to i calving.> Possible minor changes in the water temperature or in the current of water ) in the Canal de los Tempanos could have decreased calving. l

i 3.2 Glacier area and estimat ion of the eq uilibrium ------line altitude

The glacier extent was derived rom the map, geocoded optical satellite images and vertical aerial photographs. Due to errors in the available map (section 2.1) and because the surface topography is comparatively level at parts of the ice divide, the Landsat TM image (14 January 1986) and stereoscopic aerial photographs were used to improve the glacier boundaries on the map. Taking the terminus position of 6 March 1998 (ERS2), an area of 257.3 km2 was derived for Moreno Glacier. The average position of the equilibrium line (EL) , where the net gain of mass over the mass-balance year equals the net loss of mass [Paterson, 1994], was estimated from the airphotos, satellite images and ield observations. The mean equilibrium line altitude (ELA) was inferred from the mean position of the snowlines at the end of seven summers, and from ield observation. Snow lines were analyzed in the satellite images, oblique photographs and one aerial photo­ graph. Snow mapping procedures using remote sensing data were used as described by Nagler [Nagler, 1996]. As the dielectric properties and the surface roughness Chapter Glacier chaacteristics 18 3. change signiicantly from the bare ice surfaces to the snow-covered areas, SAR en­ ables the detection of snowlines on glaciers. Signiicant changes of SAR backscat­ tering at the snowline are evident on the Radarsat image acquired on 15 February 1997 (top of igure 3-4), when maximum daily temperatures of about 12°C at Moreno Base Camp (section 5.1) caused melting up to high altitudes. The mean values of the backscattering coeicients a0 calculated for two test ields of 2x2 km in size (the square ields are marked at the top of igure 3-4), are -11 dB for the ablation area and -19 dB for the snow-covered accumulation area, assuming a radar incidence angle of 46 degrees for both ields. During the months January until mid of March of 1970 (aerial photograph), 1986 (Landsat) , 1993, 1995, 1996, 1998 (ERS), and of 1997 (RADARSAT) the snow line positions were similar. Major variations, amounting to about 150 m in elevation, occurred only at the ice stream on the orographically right side. The comparison of the derived EL with oblique photographs taken during the ield campaigns at the beginning of April 1996 to 1999 revealed no signiicant changes in comparison to the radar images from February and March. The bottom of igure 3-4 shows band 4 (0.76 to 0.90 Lm) of the Landsat TM rom 14 January 1986; the snowline on 14 January 1986 almost coincided with the mean EL from the other images. A mean ELA of 1170 m a.s.l. was derived. Aniya et al. used the Landsat TM acquired on 14 January 1986, a topographic map, and Chilean aerial photographs taken on 20 December 1984 to analyze the ELA [Aniya and Skvarca, 1992], [Aniya et al., 1996]. A cluster analysis method was applied to Landsat TM bands 1, 4, and 5. Aniya estimated an ELA of 1150 m for Moreno Glacier. The area-altitude distribution (igure 3-3) was derived from the digital elevation model. Because the slope is comparatively steep above the EL, a climatically induced shift of the ELA would cause no signiicant change in the ratio of accumulation area to total glacier area (AAR) ; an ELA rise of 100 m in elevation would correspond approximately to an AAR decrease of 4%. The accumulation zone of Moreno Glacier, covering 181 km2 or 70.4% of the total glacier area, is interrupted by several prominent rock outcrops (igure 4-1). The rock outcrops separate the two main glacier streams which nourish the lower parts of the glacier.

3.3 Fluctuations of glacier terminus

Hauthal [1904, p.35] described the position of the calving front of the glacier at about 1 km behind the present position (igure 3-2). The glacier terminus advanced steadily from 1899 until 1917. R. A. Chiesa of Administraci6n Nacional de Parques Glaciares de Patagonia compiled the luctuations of the glacier tongue in 1917 and from 1928 to 1942 [Skvarca and Naruse, 1997]. Rafo et al. investigated the position of the glacier front from 1943 to 1953, it luctuated less than 200 m within this Chapter Glacier characteristics 19 3.

N 5 km A

\':

:� . . �- I

I

N 5 km A

Figure 3-4: Top: RadarsatL· image (15 February 1997) of Moreno Glacier; snow covered areas appear dark, the snowline is shown. Two squares mark the areas for

. t calculating -0. Bottom: Landsat TM band 4 (14 January 1986), the equilibrium line inferred from all ima. ,. ges is shown as black line.

C Chapter Glacier chaacteristics 20 3.

date of damming rupture date damming height looded area [m] [km ] 19171 1917 1934/351 1934/35 (Austral) winter 19391 17 Feb 1940 10.5 (Austral) winter 19411 21 Mar 1942 14.9 Dec 19462 (Austral) winter 19471 (Austral) spring 1947 2.6 Nov 19472 Apr-Dec 19482 Jul 19511 2 March 1952 12.7 (11.33) 66.7 Sep 19521 30 March 1953 14.4 {12.83) 74.0 Jul 19541 14 Sep 56, 10 Oct 56* 26.7, 25.6 (263) 88.0 Aug 19591 15 Feb 60, 31 Mar 60* 13.1, 11.6 Sep 19621 25 Feb 1963 15.7 Aug 19641 10 Feb 66, 25 Feb 66* 28.4, v27 93.6 1970 1972 1975 1977 1980 1984 17 Feb1988 Table 3.1: The damming of Brazo Rico (Brazo Sur of Lago Argentino). * Two rupture dates indicate major ruptures following partial outbrsts of water from the Brazo Rico to the Canal de los Tempanos. Sources: 1-Liss 1970, 2-Aniya 1992, 3-Heinsheimer 1958 (after Rafo, Colqui and Madejski). period [Rafo et al. , 1953] . Reported damming events are summarized in table 3.1; the information until 1966 was taken from Liss [1970, p.175]. The rise of the Brazo Sur lake level during the damming led to increased calving of the southern glacier front up to a rupture limit, when outbursts of Brazo Sur water to the Canal de los Tempanos occurred. On 9 and 10 October 1956 G.J. Heinsheimer [1958] was lucky to observe one of the main outbursts; he estimated the water, that entered into Brazo Sr during the damming event from August 1954 to October 1956, to have reached a volume of about 5.2 109 m3. Liss' sources of information concerning the snout behavior were the previous publication of J.M. Rafo [1953] and unpublished data from M.E. Madejski, Buenos Aires. rontal luctuations between 1899 and 1966 were compiled by Liss (igure 3- 5), [1970, p.176]. Liss used Hauthal's sketch map [1904, p.35], which reveals some inaccracy due to a wrong and irregular scale and an inaccurate north direction. An additional source for Liss's igure was the sketch map by Rafo [1953, p.331-332] . The positions and results of the water depth measurements carried out by Hauthal

x Chapter Glacier chaacteristics 21 3.

35,56 km2 40,88 km2 Table 3.2: Areas of the two roughness zones in the ablation area.

[1904, p.35, p.44) are included in igure 3-5. Information on recent terminus luctuations are presented by Aniya [1992) and in this thesis. Snout variations derived from vertical aerial photographs and satellite imagery are shown in the igures 3-6, 3-7 and 3-8. The earliest aerial photograph, taken between January and March 1947 [Rafo et al., 1953), shows a small part of the front touching Peninsula Magallanes. Between the date of the images of 1947 and 1965 the channel connecting Brazo Sur and Canal de los Tempanos closed 9 times, -the snout advanced in both channels. The satellite photograph of 1965 was taken during a damming period which lasted more than 1 years. The water level of Brazo Sur reached the vegetation line in 1965 (igure 3-! 6). The 1968 airphoto shows a channel between the snout and the opposite shore of Peninsula Magallanes, only minor terminus changes are visible at the sections pointing towards the water. During the following years until 1986 the two frontal sections behaved diferently. The Brazo Sr front advanced, whereas the northwestern part of the front in the Canal de los Tempanos retreated. The section at Peninsula Magallanes oscillated slightly by damming up Brazo Sur 6 times, over that period its position varied by some tens of meters. The damming interval ranged rom 2 to 4 years (table 3.1). The front has been stable since 1986 at Canal de los Tempanos, whereas it retreated slightly at Brazo Sr. Damming occurred the last time in 1988. Presently the ront is in an advanced position. Since 1988 the glacier has been touching Peninsula Magallanes during winter, but the water of Brazo Sur has been running out through a tunnel in the ice. In the 1990's the front at Brazo Sur had a similar extent as in 1947 (igure 3-8).

3.4 Description of glacier surface

The surface of the ablation area is clean, with a few, narrow longitudinal moraines (igure 3-9). The medial moraines indicate the low direction of the glacier ice below prominent rock outcrops near the ELA. The ablation area can be separated into two zones according to the surface roughness at the scale of meters to a few tens of meters: very rough and crevassed surfaces, which occur mainly in regions of strong shear in the vicinity of the glacier margins and above discontinuities of the bed rock, and comparatively smooth or gently undulating surfaces in regions with more or less regular glacier velocities. The areal extent of the smooth surface and the rough surface has been Sb,1 Sb,1 1 estimated from satellite and aerial imagery (table 3.2). Several lakes can be observed in the crevasse zones in summer (igure 3-10). In the central part, at elevations between about 450 and 600 m, two major water channels Chapter Glacier chaacteristics 22 3.

�

' z 0 I � • Figure 3-5: The terminus of Moreno Glacier 1899-1966; after C. C. Liss, 1970 0 Chapter Glacier chaacteristics 23 3.

, ...

0 m

198

1

Figure 3-6: The terminus of Moreno Glacier in Argentinian aerial photographs (1947 and 18 Nov 1968), a Corona satellite image (15 Oct 1965) and a Chilean (FACH) photograph (20 Dec 1984). All images have been coregistered to the map at the lake level. Chapter Glacier chaacteristics 24 3.

. 1

Figure 3-7: The terminus of Moreno Glacier in Landsat TM {14 Jan 1986), SPOT (23 Aug 1995) and ERS1 {16 Feb 1993) and ERS2 (6 Mar 1998) SAR images. The images have been coregistered to the map at the lake level. Chapter 3. Glacier chaacteistics 25

Figre 3-8: The position of the glcier ront rom 1947 until1995. Chapter Glacier chaacteristics 26 3.

Figure 3-9: Airphoto (11 Mar 1970) showing the terminus of Moreno Glacier; narrow moraines are starting from the rock outcrops. can be found, which meander and are located in about 15 m wide depressions, the water disappears in moulins. Chapter Glacier chaacteistics 27 .i.

Figre 3-0: Crevasses at the orographically right glacier margin. Chapter 4

Field aeasureaents

Between 21 February 1994 and 7 March 1994 an exploratory ield campaign took place in order to deine the general directions of the project. The participants of this preliminary visit were Pedro Skvarca from the Instituto Antartico Argentino (IAA), Helmut Rott and Thomas Nagler from the University of Innsbruck. They were supported by Teodoro Toconaz from the Gendarmeria Nacional. During this campaign 4 stakes, which had been installed on Moreno Glacier in late November 1993 by P. Skvarca (section 1.3), were re-measured. The possibility for setting up a transverse stake proile across the total width of the glacier tongue 7.5 km above the calving front was checked; the ablation area ws crossed from the orographically right to the left margin. Based on these observations and on former ield activities on Moreno Glacier (section 1.3) the following topics were deined for the ield program: 1: Installation and maintenance of an automatic climate station, 2: continuous ablation measurements at representative sites, 3: ice motion measurements, 4: determination of ice thickness using a seismic relection method, 5: GPS measurement of geodetic lines and ixpoints in the glacier surrounding, 6: echo sounding of lake depth near the glacier fronts.

Ice ablation, motion and seismic data are the basis for estimating the mass balance. Measured ice velocities were supplemented by the motion ield derived by remote sensing methods (chapter 6). GPS control points and lines were used in support of geocoding the images (chapter 2.2). Seven ield campaigns were carried out between November 1995 and March 1999 (table 4.1). The campaigns took place at the beginning and the end of the three summer periods 1995/96, 1996/97 and 1997/98 (November to March), and in March 1999. A chronological report of ield activities is included in appendix A. Two camp sites next to the right glacier margin were used in order to reach the locations of the ield activities in comparatively short time: the Moreno base camp and the Buscaini camp in a distance of 8 km up-glacier (igure 4-1).

28 Chapter 4. Field measurements 29

Campaign start date end date days I 14 Nov 95 10 Dec 95 26 II 16 Mar 96 6 Apr 96 21 III 16 Nov 96 10 Dec 96 24 IV 24 Mar 97 3 Apr 97 9 13 Nov 97 14 Nov 97 2 v VI 19 Mar 98 11 Apr 98 23 VII 3 Mar 99 31 Mar 99 28 Table 4.1: Field campaigns carried out on Moreno Glacier.

Stakes Climate station GPS base Camp sites

Figure 4-1: Location of the stakes (proiles A, B, Co and stake D) , of the automatic •••••••• climate station, the GPS base and the camp sites. +The calving ront is shown as a hashed line. o

0 2 4 6km

""""11 - -;;;r Chapter 4. Field measurements 30

Period Interval [ i ] n 15 Nov 95 until 31 Mar 96 15 31 Mar 96 until 3 Apr 97 20 3 Apr 97 - 30 Table 4.2: Averaging interval for meteorological measurements.

4.1 Climate station

An automatic climate station was installed on 14 and 15 November 1995 at the shore of Brazo Sur at a distance of 360 m from the orographically right terminus of Moreno Glacier (igure 4-1 and 4-2). On 15 November 1995 at 13 o' clock the irst datasets of air-temperature, humidity, air-pressure, global radiation, wind speed and wind direction were measured (section 5, appendix C). The temperatre and wind speed sensors were mounted at heights of 2mand 2.2 m above groundm (table C.1); section C.0.3 includes the instrument speciications. The meteorological measurements of the six diferent sensors were taken every 10 seconds and averaged over 15-minute intervals. One exception ws the wind direction, which was sampled only once at the end of each 15 minute interval. This 15 minute storage interval enabled a loating data record for 6 months. Dring following campaigns the storage interval was changed twice (table 4.2). The climate data were transferred to a PC during each campaign, providing a continuous data record from November 1995 onward. In March 1996 a second storage module ws connected to the datalogger in order to store the meteorological data for more than 6 months, and two solar modules were installed besides the meteorological station for additional power supply. rom 10 December 1996 to 3 April 1997 an electronic raingauge was added to the climate station.

4.2 Ablation

Ice ablation was measured at up to 19 ablation stakes, which were irst drilled between 16 and 25 November 1995 along two transverse and one longitudinal proiles (igures 4-1, 4-3). Wo oden poles with a diameter of 2 centimeter and a length of 2 m were connected with plastic tubes to obtain total lengths up to 16 m. The holes for the stakes were drilled to depths between 9.5 and 16 m, depending on the site and the measurement period. The drilling device is made up by a steam generator, 2 gasoline burners and a rubber hose, which leads the steam to a drilling pipe (igre 4-4), [Heucke, ress] . To reduce the buoyancy of the stakes in the mostly water illed holes the stakes were furnished with elastic steel wires to increase the riction. Short steel poles were additionally ixed at the end of some stakes to raise their weight. The lower transverse proile (A-proile) consisted of 5 stakes between the orographically right glacier margin and the middle of the glacier, the distance to the calving glacier front Ch apter 4. Field measurements 31

Figure 4-2: Automatic climate station near Perito Moreno Glacier. Chapter 4. Field measurements 32

2m·

.... :.·

.,.· A04x ·soa x x A03 B07 AQ2 x · � •. · · . Btt . . B06X .. . � . . ..· ·Aot ;�.����: :� .. �- ·. :-:::.� : ·. . -.-� -- · - ... B05x ···; !' .- -.- .. • -� ... A05 X .. , .,_ . . .. _ __ _ . : B04X .•. B03 x L03 x B02x 02 x BOt X . L .,.. . � -• � LOt X f . � . . ·i.. . •. . . .• •.

Figure 4-3: SPOT image (23 Aug.l995) with positions. . : .: ·· of.:·· 19 stakesX forming the transverse proiles A (stakes AOl to A05) and B (BOl·: to B11), and the longitudinal proile C (LOl to L03). c:"

· .. _. :-·:: ..

. � ......

-��

.. p" .• ; : r , ·. Chapter 4. Field measurements 33

Figure 4-4: H. Rott drills a hole for an ablation stake.

was 4.5 km. A heavily crevssed zone in the center of the glacier inhibits the access to the northern part of the tongue. The upper transverse proile (B-proile) was spanning the whole glacier width of 4.4 kilometers. 11 stakes formed the B-proile, its distance upstream from the calving glacier front ws 7.5 kilometers. The 3 stakes of the longitudinal C-proile (LOl, L02, L03) were placed along a lowline above stake B03, the mximum distance to proile B ws 2.3 km (L01) (igre 4-3). The stakes were redrilled at the original positions during each ield campaign to obtain comparable data for diferent years, and to avoid the propagation of the stkes into crevssed zones glacier-downward. All stakes were fo und again during the following campaigns. Ablation mesurements of two stakes were impossible, as the stakes lay at the bottom of crevasses (B10 in March 1996, B09 in March 1997). In March 1996 stake B07 of the B-proile was eliminated s the position very close to the adj acent stakes B08 and Bll (igure 4-3). On 1 December 1995 P. Skvarca installed one stake (stake D) drilling to a depth of 15.9 m with another, more heavy drilling device (position in igure 4-1). In March 1997 the stake net ws reduced to 3 stakes at A-proile and 5 stakes at B-proile to maintain a bseline for multi-year ablation.

The mesured ablation ws related to the air temperature to calculate degree­ days. Mean daily temperatures were measured at Moreno Bse (section 5.2) and at a second temperature sensor installed near stake A03 during the period from 20 November 1996 to 9 December 1996. A degree-day is deined a departure of one degree per day in the daily mean temperatre rom an adopted reference temperature. For estimating snow and ice

ablation the reference temperature is 0°C. The ratio of the total amount ofwas ablation

as Chapter 4. Field measurements 34

-A-profile ··-- - Base

··' I G : �

. �/ ·

e & � �

� � � 8 15 � 14 13 •\ Figure 4-5: 12Floating 3 hourly mean temperatures measured at the climate station L II near the� Base Camp and at the A-proile in November/December 1996. The A­ 10 proile temperature curve shows gaps from 2nd to 4th December 19I 96 due to technical 9 ' problemsE. 8 = .. 7 6 Bb(z) at a 5st ake to the cumulative daily mean temperatures above ooc ET(z) is called the positiv4 e degree-day factor k [Zingg, 1951]. The ablation is expressed 3 Bb(z) in em water2 equivalent assuming an ice density of 900 kg/m-3. I 0

...... z z z z z z z z z z z 0 0 0 0 0 0 0 0 0 (4.1) i i i ; D ..: O � i ..: O � ) ) ) ) ) ) ) ) ) ) )... )0 ) ) ) ) ) ) ) ) ------> > > > > > > > > > > u u u u u u u u u The 3-hour0 loating 0 0 0mea 0 n 0 air0 tempera0 0 0tures 0 at stake A03 ranged from 2.1 to ' 11.5°C during Nthe N 20 N daysN period N N in N NovembN N er/Decemb0 er 1996, the0 mean 0 0 temper­ ature was 6.2°C. The relationship between the temperatures at Moreno Base and at A03 is shown in igure 4-5. A mean temperature diference between A03 and the base was 2.15°C. The altitude diference between the 2 meteorological stations was 173 m corresponding to a vertical temperature gradient of 1.24°C/100 m. This increased gradient is probably caused by diferences in the surface/atmosphere heat exchange of the two diferent surface types. The thermal modiication of air by a glacier ws discussed by Braithwaite [Braithwaite, 1977] . The ice ablation from 20 November to 9 December 1996 at stake A03 was 85 em, a mean degree-day factor of 0.63 em water equivalent;oC day was derived (igure 4-6). Takeuchi and others established two measurement sites for air temperature close to the Moreno Base Camp and about 150 m higher at the orographically right-hand bank of Moreno Glacier [Takeuchi et al., 1996]. Takeuchi measured an average lapse Chapter 4. Field measurements 35

80

70

60

� 50 -

-· 40

� 30 k=0.63 cmrcday 20

.·

0

0 10 20 30 40 50 60 70 80 90 100 110 12-•0 Cumulative Air Temperature

Figure 4-6: The relation between the ablation of stake A03 and the cumulative temperature; the diamonds show the measured ablation values between 20 November 1996 and 9 DecE ember 1996. �....

rate of air temperature of 0.8°C/100 m during the. period from 1 December 1993 to

<::: 30 November0 1994. For calculating the (positive) degree-day sums ET(z) in equation 4.1 the temperatures at stake altitudes were computed assuming a linear vertical 0 ( ) temperature lapse rate l .

(4.2) T(z) T(zreJ) - l(z - ZreJ) � where Zref is the elevation of the meteorological station (192 m a.s.l. ). For igures 4-9 and 4-11 a mean lapse rate of air temperature of 0.8°C/100 m was assumed according to Takeuchi [1996] . The resulting degree-day factor k for each stake revealed distinct diferences between summer and winter and between central and marginal zones of the glacier (igures). Stake D drilled on 1 December 1995 provided short period ablation data for 151 days of summer 1995/96. Mountain guides measured the ablation of stake D 84 times, until 30 April the ice ablation was 10.99 m (table 4.3). The relation between z the amount of ablation and the cumulative daily mean air temperatures during the period from 1 December 1995 until 30 April 1996 was calculated (igure 4-7). Stake D was located at a distance of 600 m rom the climate station, the altitude diference was 40 m. The distance of stake D and the glacier margin was about 70 m. Neglecting possible local variations, the same daily mean temperatures at the

stake site at the climate station= was assumed and a degree-day factor k of 0.79

s Chapter 4. Field measurements 36

Period total ablation mean daily ablation December 1995 269 9.0 January 1996 270 8.7 February 1996 229 7.9 March 1996 218 7.8 28 March - 30 April 1996 113 3.4 1 December 1995 - 30 April 1996 1099 7.3 Table 4.3: Ice ablation (in em) at stake D drilled on 1 December 1995 near the glacier margin.

(em water equivalent;oc day) was obtained. The ice ablation at the A-proile for a period of 99 days in summer 1995/96 was similar at stakes A01, A03 and A04 (tables 4.4 and 4.9, igure 4-8). The highest ablation of 6.8 cmd-1 was measured at the marginal zone at stake A05, the ablation at stake A02 was 1.1 em less. The winter period of 1996 revealed ablation rates less than 2 em/day for all A-proile stakes. The ablation diference between the marginal stakes (A05 and A01) and the stakes in the central part was 34%. Though the mean temperature in summer 1996/97 was 0.5°C lower than in the previous summer, exceptionally high ablation rates were measured at stakes A01 and A02 (table 4.6). The ablation at stake AOl was 0.84 em water equivalent;oc day for a period of 106 days in summer 1996/97 (igure 4-9). Ablation rates lower than 1 em/day were observed during the comparatively cold winter 1997. The mean temperature from May until October 1997 ws 1.6°C lower than in winter 1996, the May temperature was even 3.3°C colder than in the previous year. The comparison of the two summer periods of 1995/96 and 1997/98 with similar mean temperatures (0.1 oc higher from December-March in 1997/98) revealed 0.6 em/day less ablation at A01 and A02 in 1997/98 (table 4.9) . Ablation was also measured near the A-proile at 4 stakes by Skvarca (section 1.3, [Naruse et al., 1995b]). These ablation measurements during a period of 110 days in summer 1993/94 and the measurements in the three summers of 1995/96 to 1997/98 (table 4.10) show comparatively little inter-annual variability. The mean ablation of 6.36 cmd-1 in summer 1993/94 agrees well with the mean of proile A (6.26 cmd-1) for summer 1995/96. Takeuchi measured the air temperature on the right glacier margin in the region of A proile and at the base camp during 14 days in November 1993 [Takeuchi et al., 1995b). Applying the degree-day method, a factor of 0.76 em water equivalent;oc day was obtained for the period from 12 to 27 November 1993 [Takeuchi et al., 1995a) . Naruse [Naruse et al., 1995b) extrapolated the temperature measurement period until March 1994, using temperatures from Lago Argentino Station, to calculate a degree-day factor of 0.69 for the whole 110 days period. As the stake positions in 1993/94 did not exactly coincide with A­ proile stakes, there is no signiicant diference to the ablation rates and degree-day factors which were observed in the summers 1995/96 to 1997/98 (table 4.12). Chapter 4. Field measurements 37

1000

900

800

.... 700

600 � ' k=0.79 cmfDCday ": - - wo �

400 � //- 300 J /' 1 200 j / 10� k<"

0 200 400 600 800 1000 1200 1400 ' Cumulative Air Temperature

.. . ' Figure 4-7: The relation between the ablation of staker•- D and the cumulative mean � r ' daily air tempe: rature during the period from 1 December 1995 until 30 April 1996; the resultinE g degree-day factor k is 0.79 (em water equivalent;oc day).

�- -' ..' ,' ; u Stake Date I Date II Height Ablation ..... A05 8 Dec 95 17 Mar 96 795 6.8 A01 8 Dec 95 17 Mar 96 747 6.3 A02 8 Dec• 95 17 Mar 96 748 5.7 A03 8 Dec 95 17 Mar 96 720 6.2

1 A04 8.. Dec 95 17 Mar 96 742 6.2 B01 9 Dec 95 19 Mar 96 787 6.6 B02 9 Dec 95 19 Mar 96 775 6.1 B03 8 Dec 95 19 Mar 96 730 5.9 B04 8 Dec 95 19 Mar 96 572 4.5 B05 8 Dec 95 19 Mar 96 626 5.1 B06 8 Dec 95 19 Mar 96 687 5.5 Bll 8 Dec 95 19 Mar 96 641 5.2 B07 8 Dec 95 19 Mar 96 620 5.3 B08 8 Dec 95 23 Mar 96 814 6.5 B09 9 Dec 95 20 Mar 96 746 6.0 B10 9 Dec 95 20 Mar 96 960 8.0 101 9 Dec 95 24 Mar 96 741 5.5 102 9 Dec 95 24 Mar 96 708 5.6 103 9 Dec 95 24 Mar 96 614 4.8

Table 4.4: Ablation during the summer period 1995/96 of stakes drilled in November 1995. The indicated ice ablation in centimeter/ day is the mean of the period reaching from date I to date II. Chapter 4. Field measurements 3B

Stake Date I Date II Height Ablation A05 2B Mar 96 17 Nov 96 443 l.B A01 31 Mar 96 17 Nov 96 37B 1.6 A02 31 Mar 96 19 Nov 96 32B 1.3 A03 29 Mar 96 19 Nov 96 290 1.2 A04 29 Mar 96 19 Nov 96 332 1.3 B01 19 Mar 96 22 Nov 96 57B 1.9 B02 23 Mar 96 25 Nov 96 530 l.B B03 27 Mar 96 22 Nov 96 300 1.0 B04 23 Mar 96 22 Nov 96 277 1.0 B05 19 Mar 96 22 Nov 96 302 1.2 B06 23 Mar 96 24 Nov 96 291 1.1 Bll 23 Mar 96 24 Nov 96 290 1.1 BOB 22 Mar 96 24 Nov 96 359 1.4 B09 22 Mar 96 24 Nov 96 376 1.4 BlO 22 Mar 96 24 Nov 96 577 1.9 L01 27 Mar 96 25 Nov 96 303 1.1 L02 27 Mar 96 25 Nov 96 302 1.2 L03 27 Mar 96 25 Nov 96 252 1.0 Table 4.5: Ablation during the winter period 1996 of stakes drilled in March 1996 [em/day] .

Stake Date I Date II Height Ablation A05 9 Dec 96 25 Mar 97 775 6.1 A01 9 Dec 96 25 Mar 97 B75 7.3 A02 9 Dec 96 25 Mar 97 755 6.5 A03 B Dec 96 25 Mar 97 647 5.2 A04 9 Dec 96 25 Mar 97 610 4.7 BOl 6 Dec 96 26 Mar 97 B75 7.0 B02 25 Nov 96 26 Mar 97 701 5.0 B03 5 Dec 96 26 Mar 97 716 5.B B04 24 Nov 96 26 Mar 97 545 4.3 B05 22 Nov 96 27 Mar 97 656 5.1 B06 23 Nov 96 27 Mar 97 560 4.3 Bll 23 Nov 96 27 Mar 97 636 4.9 BOB 24 Nov 96 27 Mar 97 753 5.9 BlO 24 Nov 96 27 Mar 97 B06 6.1 L01 24 Nov 96 26 Mar 97 524 4.2 L02 24 Nov 96 26 Mar 97 61B 5.0 L03 25 Nov 96 27 Mar 97 572 4.6

Table 4.6: Ablation during the summer period of 1996/97 of stakes drilled in Novem- her 1996 [em/day] . Chapter 4. Field measurements 39

Stake Date I Date II Height Ablation A01 2 Apr 97 13 Nov 97 228 1.0 A02 31 Mar 97 13 Nov 97 210 0.9 A04 30 Mar 97 13 Nov 97 221 1.0 B03 28 Mar 97 14 Nov 97 239 1.0 B04 27 Mar 97 14 Nov 97 200 0.8 B05 30 Mar 97 14 Nov 97 174 0.7 B06 30 Mar 97 14 Nov 97 224 0.9 B11 30 Mar 97 14 Nov 97 227 0.9 A01 13 Nov 97 20 Mar 98 940 5.6 A02 13 Nov 97 20 Mar 98 865 5.2 A04 13 Nov 97 20 Mar 98 967 5.9 B03 14 Nov 97 23 Mar 98 1050 6.3 B04 14 Nov 97 23 Mar 98 885 5.3 B05 14 Nov 97 23 Mar 98 800 4.8 B06 14 Nov 97 23 Mar 98 950 5.6 B11 14 Nov 97 23 Mar 98 875 5.0 Table 4.7: Ablation during winter and summer periods of 8 stakes drilled in March 1997 [em/day].

Stake Date I Date II Height Ablation A01 20 Mar 98 7 Mar 99 1180 3.3 A02 21 Mar 98 7 Mar 99 1117 3.2 A04 20 Mar 98 7 Mar 99 1120 3.2 B03 29 Mar 98 9 Mar 99 *** B04 29 Mar 98 7 Mar 99 1058 3.0 B05 29 Mar 98 7 Mar 99 880 2.5 B06 29 Mar 98 8 Mar 99 852 2.4 B11 29 Mar 98 8 Mar 99 956 2.7 Table 4.8: Yearly ablation of 8 stakes drilled in March 1998 [em/day] ; *** stake P03 was melted out.

A05 AOl A02 A03 A04 08 Dec 95 - 17 Mar 96 6.8 6.3 5.7 6.2 6.2 30 Mar 96 - 18 Nov 96 1.8 1.6 1.3 1.2 1.3 09 Dec 96 - 25 Mar 97 6.1 7.3 6.5 5.2 4.7 31 Mar 97 - 13 Nov 97 1.0 0.9 1.0 13 Nov 97 - 20 Mar 98 5.6 5.2 5.9

Table 4.9: Daily ice ablation (em/day) of A-proile stakes for 5 measurement periods. Chapter 4. Field measurements 40

8 ii A05 .AOl DA02 DA03 .A04 7 6 � 5

4 � 3 2 1 0 ,In , J 1_1 summer winter 96 summer winter 97 summer 95/96 96/97 97/98

Figure 4-8: Daily ice ablation at the A-proile for 5 measurement periods .

� A05 AOl OA02 0 A03 A04 • • • 0.9

0.8

0.7

0.6 ._ 0.5

0.4

e 0.3 '

0.2

0.1

0.0 Summer Winter 96 Summer Winter 97 Summer 95/96 96/97 97/98

Figure 4-9: Degree-day factor calculated for the A-proile stakes for 5 periods. The ablation is expressed in em water equivalent. Error bars correspond to pos­ sible inaccuracies in the degree-days due to diferent lapse rates of air temperature; higher degree-day factors result from calculations with dry adiabatic lapse rates (0.01°C/m), -lower values from wet adiabatic lapse rates (0.006°C/m). u

, S 0

0

u Chapter 4. Field measurements 41

Period Proile A Proile A Proile B Proile B Center Margin Center Margin 19 Nov 95 - 18 Mar 96 6.0 6.5 5.1 6.2 23 Mar 96 - 22 Nov 96 1.3 1.7 1.1 1.7 22 Nov 96 - 26 Mar 97 5.2 6.4 4.9 6.0 30 Mar 97 - 14 Nov 97 0.9 1.0 0.9 14 Nov 97 - 22 Mar 98 5.5 5.6 5.4

Table 4.10: Mean daily ice ablation (em/day) of A- and B-proile stakes for ive measurement periods; the center ablation represents the mean value of A02, A03, A04 and B03 - B11 stakes, the margin ablation refers to A05, AOl and BOl, B02, BOB- B10 stakes; ablation values after March 1997 refer to the reduced stake proiles.

l • B03 • B04 D B05 D B06 • Bll

j

� � � ! 8 � , ! f i

j !

Summer Winter 96 Summer Winter 97 Summer 95/96 96/97 97/98

7.0

Figure 4-10:6. 0Da -i ily ice ablation of the 5 central stakes of the B-proile for 5 measure­ ment periods. � 5.0

!

The mean4.0 i daily ice ablation rates at the B-proile for three summer and two winter periods ranged from 0.7 cmd-1 to 7.0 cmd-1 (tables 4.4 to 4.7). Figure 4- 3.0 10 shows the mean daily ablation at the central B-proile, ablation increased from these cen-tra2.0l st I akes towards the stakes B01 and B10 at the marginal glacier zones.

The greatest1.0 diference in ablation rates along the B-proile within one period was

observed in sui mmer 1996/97 when the ablation at stakes B04 and B06 was 2.7 cmd-1 lower0. 0th an at BOl. As for A01 and A02, an exceptionally high degree-day factor was observed for B01 during summer 1996/97 (igure 4-11); the ice ablation of B01 exceededj the value measured during the previous summer by 6.3%. Table 4.11 contains ablation measurements of the C-proile stakes. Three periods of ablation data were measured. The C-proile stakes were drilled along a lowline above stake B03 (igure 4-3). Whereas the surface at 103 was comparatively smooth, the region of 101 and 102 was characterized by surface undulations with a horizon- Chapter 4. Field measurements 42

1.1

1.0

0.9 BOl 0.8 •B02 0.7 DB03 � 0.6 DB04 � 0. 5 •B05 0.4 •B06 • Bll 0.3 DB08 0.2 •B09 0.1 D BlO

0.0

Summer 95/96 Winter 96 Summer 96/97

Figre 4-11: Degree-day factor calculated for the B-proile stakes for 3• periods; the ablation is expressed in em water equivalent. Error bars correspond to possi­ ble inaccuracies in the degree-days due to diferent lapse rates of air temperature; higher degree-day factors result from calculations with dry adiabatic lapse rates (0.01oC u /m), -lower values from wet adiabatic lapse rates (0.006°C/m). � 0 Period LOl L02 L03 9 Dec 95 - 24 Mar 96 5.5 5.6 4.8 28 Mar 96 - 25 Nov 96 1.1 1.2 1.0 25 Nov 96 - 26 Mar 97 4.2 5.0 4.6 Table 4.11: Mean daily ice ablation (em/day) at the C-proile for 3 measurement periods.

tal scale of about 30 m and a vertical scale of a few meters. According to these diferences in surface roughness the maximum C-proile ablation rates occurred at 102. Though the altitude diference between stake 101 and stake 103 was 90 m, the mean ice ablation at 103 during summer 1995/96 was lower by 0.7 cmd-1 than at 101 (table 4.11). The colder mean temperature of summer 1996/97 resulted in 13% less ice ablation than in summer 1995/96; the degree-day factors were similar during the two summer periods (0.73 :}0.71, table 4.12).

The measurements of ablation revealed distinct diferences between the central-, smooth parts of the glacier and the crevasse and boundary zones, and between A­ and B proiles for the period from November 1995 to March 1997 (table 4.10). At the central zones 11% less ablation was observed at proile B; the diference at the marginal stakes was 4%. During the period from March 1997 to March 1998 slightly less ablation occurred at B-proile than at A-proile. The ablation of the boundary Chapter 4. Field measurements 43

Proile A Proile A Proile B Proile B Proile C Period Center Margin Center Margin Center 9 Dec 95 - 17 Mar 96 0.65 0.70 0.65 0.80 0.73 30 Mar 96 - 18 Nov 96 0.27 0.36 0.28 0.42 0.34 9 Dec 96 - 25 Mar 97 0.62 0.77 0.65 0.78 0.71 1 Apr 97 - 13 Nov 97 0.26 0.31 14 Nov 97 - 20 Mar 98 0.62 0.70 Table 4.12: Mean degree-day factors (em we. I degree-day) for 3 summer and 2 winter periods.

Season Center Margin Nov - Mar 0.65 0.76 Apr-Oct 0.29 0.39 Table 4.13: Mean seasonal degree-day factors (em we. I degree-day) for the two sur­ face zones. zones is higher due to entrainment of warmer air from the ice-free surfaces, due to longw.ve radiative efects and partly also due to lower ice albedo. The ablation of the crevsse zones is higher due to increased turbulent mixing as result of increased roughness and larger surface are.. Ablation may take place throughout the year because even in winter temperatures stay above ooc over extended periods and the lower terminus is usually covered by snow only during short periods. On the average of the two transects, the ablation was lower by 5% in summer 1996197 than in summer 1995196. The mean temperatures for the two summers (December to March) were 9.6°C (1995196) and 9.1°C (1996197) [Rott et .l., 1998]. Tables 4.12 and 4.13 show the degree-day factors averaged over central and marginal zones, the corresponding degree-day sums were calculated with . lapse rate of 0.8°CI100 m. A degree-day factor of 0.91, as obtained for stake BOl during the summer 1996197, represents .n upper limit. Though meteorological parameters other than temperature, such as the precipitation, the amount of solar radiation, and the wind speed, also afect ablation, .n almost linear relation between ablation .t stake D and the cumulative daily temperatures was observed. The diference of average degree-day factors between winter and summer is 45% for the central zone and 51% for the glacier margin. The mean seasonal degree-day factors in table 4.13 can be compared with ablation measurements .t 8 stakes from March 1998 until March 1999. The calculated ablation exceeds the measured ablation by 1% (igure 4-12). Braithwaite and Hock summarized degree-day factors for ice and snow on various glaciers in the Alps, on the Greenland ice sheet and in Iceland [Braithwaite, 1995), [Hock, 1998]. Typical degree-day factors between 0.6 and 0.7 cm;oc day agree well with the values derived for the summer seasons in table 4.12. A mean of 0. 72 cm;oc day was derived by Braithwaite [1995] with ice ablation dat. mea­ sured on Nordbogletscher in South Greenland almost every day for six summer Chapter 4. Field measurements 44

•measured 0 calculated

�

1400

1200 Figure 4-12: Ablation measured at a reduced stake net at A- and B-proiles during

a one year periods 1000 from March 1998 until March 1999 in comparison with ablation calculated with mean degree-day factors. :: 800 .)u periods. Braithll waite and Olesen found a seasonal variation of degree-day factors

0 on Qamanarssup. .. in West Greenland; the derived mean values were 0. 79 cm;oc

day in summer-1 from400 June to August and 0.94 cm;oc day for September to May [Braithwaite and Olesen, 1993). Braithwaite found evidence for a tendency of high ) 200 degree-day factorsu to occur at lower temperatures [Braithwaite, 1995). This fact H stands in contrast to the present study on Moreno Glacier. A mean degree-day factor 0 of 0.77 cm;oc day was reported for Satujokull glacier in Iceland [Johannessen et al., 1995). A0 1 A02 A04 B04 B05 BOG Bll Hoinkes and Steinacker assumed a degree-day factor of 0.9 cm;oc day for melt- ing of ice from Hintereisferner in the Otztaler Alps [Hoinkes and Steinacker, 1975) . Hoinkes obtained a signiicant correlation between the mass balance of Hintereis­ ferner and degree day sums.

4.3 Stake velocities

A rimble GPS Pathinder Pro was used for measuring the positions of the ablation stakes (information with technical speciications is available on rimble homepage: http:// www.trimble.com/). The GPS system provides real-time sub-meter accu­ racy, the measured positions are referred to the WGS84 ellipsoid. The diferential global positioning system (DGPS) operates with 2 antennas, one antenna is ixed to a base station with known coordinates (section 4.5). The 2nd antenna measures simultaneously the distance to the base antenna and to at least 4 satellites. The available rimble phase processor software provided with the exception of Chapter 4. Field measurements 45

Stake Date expected accuracy A04 19 Nov 96 1.0 m B02 22 Nov 96 1.0 m L02 25 Nov 96 1.0 m Bll 24 Nov 96 5.0 m Table 4.14: Stake coordinates with accuracy lower than 0.3 m.

98

...

<

<

<

<

BOl B02 B03 B04 B05 B06 B07

Figure 4-13: 2.5Mean daily velocities of B-proile stakes; the velocities are averaged 2.06 over 5 (4) days from 23.(24.) Nov. 1995 to 28 Nov.1.1995 . Error bars correspond to possible inaccura2.0 cies of GPS measurements.

for stake .posit 1.5ion s table 4.14) horizontal point measurement accuracies between 10 ll ( and 30 em,0 the GPS logging intervals were between 10 and 15 minutes. The accuracy values were: sp1.0eciied by the software; repeated measurements of the position of the climate station (section 4.1) revealed a GPS accuracy of 3 em for the short baseline 0.5 of 94.1 m (table 4.22). Bad GPS accuracies might result due to a poor spatial coniguration of the Navstar satellites, which is indicated at the GPS receiver by a 0.0 high PDOP (position dilution of precision) . Between 1995 and 1999 165 positions of ablation stakes were measured. The co­ ordinates of the stakes with the date of the mesurements are listed in the appendix (section B.1) . Stake velocities during the diferent seasons varied from 0.15 md-1 in the vicinity of the glacier margins to 2.39 md-1 at stake LOl (tables 4.16 to 4.19) . The mean number of days between the measurements is listed in table 4.15. A irst estimation of ice movement was possible by re-mesuring the positions of 7 stakes along the B-proile in November 1995 (table 4.16, igure 4-13). The stake velocities increased from the glacier margin towards the center line of the glacier; stake B06 moved about 10.3 m within 5 days. Large error bars (up to 0.28 md-1 for stake B03) in igure 4-13 are due to the speciied position measurement accuracy of 0.3 m and the short time period. The daily average of stake displacements at Chapter 4. Field measurements 46

Summer 95/96 Winter 96 Summer 96/97 1.8

1.6

1.4

1.2

1.0

� 0.8

0.6

0.4

0.2

0.0

A05 AOl A02 A03 A04

Figure 4-14: Seasonal distribution• of surf• ace velocities0 at the lower transverse stake proile (A-proile). The velocities do not correspond to tables 4.17 and 4.18 due to the impact of slightly diferent stake positions and diferent paths. transects A and B for two summer and one winter season is shown in igures 4-14 and 4-15. Tables 4.16, 4.17 and 4.18 include the dates of stake measurements and the corresponding displacements. i The velocity of the A-proile stakes increases steadily from the marginal stake

l A05 towards A04... . In March and November 1996 the A-proile stakes were set up at distances up to 207 m (stake A05 in November 1996) from the original positions in November 1995. In order to compare seasonal velocity variations (igure 4-14), the impacts of diferent stake positions and diferent displacements were considered using the motion ield derived from SIR-C/X-SAR data by means of interferometry (section 6.2) . The A-proile velocities were in the average 10% higher during the summer seasons 1995/96 and 1996/97 than during winter 1996. In summer 1996/97 the velocities were lower by 6% than in summer 1995/96. The summer velocities of the B-proile stakes exceeded winter 1996 velocities by 7% (igure 4-15). The velocities of the stakes LOl to L03 of the C-proile were similar during the summer periods. In winter the velocities were lower by 4%. The C-proile velocities varied from 2.38 md-1 at L01 in summer 1996/97 to 2.06 md-1 at the stake L03 in winter 1996 (igure 4-16). The average diference in ice motion measured at all stakes between the two summer periods November 1995 to March 1996 and November 1996 to March 1997 and the winter period March 1996 to November 1996 was 7%.

Average daily velocities at 8 stakes were obtained for two following annual periods from the end of March 1997 until March 1999 (table 4.19 and 4.20, igure 4-17) . In March 1997 the A-proile was reduced to 3 stakes (section A.4) , the stakes were set up at distances of 192 m (A01), 144 m (A02) and 124 m (A04) from the original Chapter 4. Field measurements 47

summer winter summer yer yer 95/96 96 96/97 97/98 98/99 A-proile 117 235 125 354 351 B-proile 118 246 124 360 344 C-proile 122 247 122 Table 4.15: Average number of days between GPS measurements of stakes.

Stake Date I Date II 6t Displacement Daily Displacement A-proile (d) (m) (md 1) A05 21 Nov 95 17 Mar 96 116.8 23.7 0.20 A01 21 Nov 95 17 Mar 96 116.9 56.1 0.48 A02 21 Nov 95 17 Mar 96 117.0 124.6 1.07 A03 21 Nov 95 17 Mar 96 117.0 170.1 1.45 A04 21 Nov 95 17 Mar 96 117.1 191.4 1.63 B-proile (d) (m) (md 1) B01 23 Nov 95 28 Nov 95 5.3 1.6 0.30 ±0.11 B01 28 Nov 95 19 Mar 96 111.6 30.3 0.27 B02 23 Nov 95 28 Nov 95 5.3 5.7 1.07 ±0.11 B02 28 Nov 95 19 Mar 96 111.7 106.7 0.96 B03 24 Nov 95 28 Nov 95 4.3 7.6 1.77 ±0.14 B03 28 Nov 95 19 Mar 96 111.9 185.6 1.66 B04 24 Nov 95 28 Nov 95 4.7 8.9 1.89 ±0.13 B04 28 Nov 95 19 Mar 96 112.1 203.4 1.81 B05 23 Nov 95 28 Nov 95 4.9 9.7 1.98 ±0.12 B05 28 Nov 95 19 Mar 96 112.1 205.1 1.83 B06 23 Nov 95 28 Nov 95 5.0 10.3 2.06 ±0.12 B06 28 Nov 95 19 Mar 96 112.1 203.4 1.81 B11 23 Nov 95 19 Mar 96 117.1 199.1 1.70 B07 23 Nov 95 28 Nov 95 5.0 7.5 1.51 ±0.12 B07 28 Nov 95 19 Mar 96 112.2 180.7 1.61 BOB 23 Nov 95 23 Mar 96 121.0 182.7 1.51 B09 23 Nov 95 20 Mar 96 117.9 91.3 0.77 B10 23 Nov 95 20 Mar 96 117.9 32.4 0.28 C-proile (d) (m) (md-1) 101 23 Nov 95 24 Mar 96 121.7 289.7 2.38 102 23 Nov 95 24 Mar 96 121.9 281.1 2.31 103 23 Nov 95 24 Mar 96 121.8 262.1 2.15 D-stake (d) (m) (md-1) D 17 Mar 96 8 Apr 96 21.7 3.6 0.17

Table 4.16: Velocities of stakes drilled in November/December 1995. The daily stake-displacement refers to the period between date I and date II. Possible dis- placement errors refer to the accuracy of GPS position measurements, for periods more than 100 days the possible errors are less than 1 em/day. Chapter 4. Field measurements 48

Stake Date I Date II .t Displacement Daily Displacement A-proile (d) (m) A05 29 Mar 96 19 Nov 96 234.9 36.0 0.15 A01 29 Mar 96 19 Nov 96 235.0 94.2 0.40 A02 29 Mar 96 19 Nov 96 234.9 240.0 1.02 A03 29 Mar 96 20 Nov 96 236.1 322.2 1.36 A04 29 Mar 96 19 Nov 96 234.9 357.4 1.52 B-proile (d) (m) BOl 27 Mar 96 22 Nov 96 240.0 55.3 0.23 B02 27 Mar 96 22 Nov 96 240.1 181.3 0.76 B02 22 Nov 96 26 Nov 96 3.3 5.9 1.80 ±0.18 B03 19 Mar 96 22 Nov 96 248.1 398.2 1.61 B04 23 Mar 96 22 Nov 96 244.0 420.2 1.72 B05 19 Mar 96 22 Nov 96 248.1 435.7 1.76 B06 23 Mar 96 24 Nov 96 245.8 425.7 1.73 Bll 23 Mar 96 24 Nov 96 246.0 391.8 1.59 Bll 24 Nov 96 28 Nov 96 3.9 12.6 3.23 ±0.15 B08 23 Mar 96 24 Nov 96 246.0 358.8 1.46 B09 20 Mar 96 24 Nov 96 249.3 182.2 0.73 BlO 20 Mar 96 24 Nov 96 249.3 50.3 0.20 C-proile (d) (m) L01 24 Mar 96 25 Nov 96 246.1 571.7 2.32 L02 24 Mar 96 25 Nov 96 246.1 550.8 2.24 L02 25 Nov 96 28 Nov 96 3.0 6.5 2.20 ±0.20 L03 24 Mar 96 25 Nov 96 246.2 505.6 2.05 Table 4.17: Velocities of stakes drilled in March 1996.

Summer 95/96 Winter 96 Summer 96/97 • • D 2.0 1.8 1.6 1.4 � 1.2 .... 1.0 0.8 0.6 0.4 0.2 0.0 BOl B02 B03 B04 B05 B06 Bll B08 B09 BlO

Figure 4-15: Sesonal distribution of surface velocities at the upper transverse stake proile (B-proile) . The velocities do not correspond to tables 4.17 and 4.18 due to the impact of slightly diferent stake positions and diferent paths.

0 e Chapter 4. Field measurements 49

Stake Date I Date II .t Displacement Daily Displacement A-proile (d) (m) (md-1) A05 19 Nov 96 25 Mar 97 125.8 22.3 0.18 AOl 20 Nov 96 25 Mar 97 124.7 61.2 0.49 A02 20 Nov 96 25 Mar 97 124.8 117.5 0.94 A03 19 Nov 96 25 Mar 97 125.8 181.6 1.44 A04 20 Nov 96 25 Mar 97 124.9 188.4 1.51 B-proile (d) (m) (md-1) BOl 22 Nov 96 26 Mar 97 124.0 30.0 0.24 B02 22 Nov 96 26 Mar 97 124.0 110.3 0.89 B03 22 Nov 96 26 Mar 97 124.0 216.0 1.74 B04 22 Nov 96 26 Mar 97 124.0 223.4 1.80 B05 22 Nov 96 27 Mar 97 124.9 229.3 1.84 B06 24 Nov 96 27 Mar 97 123.2 219.0 1.78 B11 24 Nov 96 27 Mar 97 123.1 197.5 1.60 BOB 24 Nov 96 27 Mar 97 123.0 184.7 1.50 B09 24 Nov 96 27 Mar 97 122.7 89.2 0.73 BlO 24 Nov 96 27 Mar 97 122.8 26.0 0.21 C-proile (d) (m) (md-1) L01 25 Nov 96 26 Mar 97 121.1 289.7 2.39 L02 25 Nov 96 27 Mar 97 121.8 281.8 2.31 L03 25 Nov 96 27 Mar 97 121.8 258.2 2.12 Table 4.18: Velocities of stakes drilled in November 1996.

• Summer 95/96 •winter 96 D Summer 96/97 2. 4 . 2.2 2.0 1.8 1.6 m 1.4 d 1.2 1.0 0.8 0.6 0.4 0.2 0.0 LOl L02 L03

Figure 4-16: Seasonal distribution of surface velocities at the longitudinal stake proile ( C-proile) . The velocities do not correspond to tables 4.17 and 4.18 due to the impact of slightly diferent stake positions and diferent paths. i

�

e

' Chapter 4. Field measurements 50

Stake Date I Date II 6t Displacement Daily Displacement A-proile (d) (m) (md-1 ) A01 31 Mar 97 20 Mar 98 353.8 224.1 0.63 A02 31 Mar 97 20 Mar 98 353.9 402.1 1.14 A04 31 Mar 97 20 Mar 98 353.9 513.4 1.45 B-proile (d) (m) (md 1) B03 27 Mar 97 23 Mar 98 361.2 608.1 1.68 B04 27 Mar 97 23 Mar 98 360.9 653.7 1.81 B05 28 Mar 97 23 Mar 98 360.2 662.6 1.84 B06 30 Mar 97 23 Mar 98 358.0 642.1 1.79 Bll 30 Mar 97 23 Mar 98 358.1 588.6 1.64 Table 4.19: Velocities of stakes drilled in March 1997. Stake Date I Date II 6t Displacement Daily Displacement A-proile (d) (m) (md-1) A01 20 Mar 98 19 Dec 98 274 166.3 0.61 A01 20 Mar 98 07 Mar 99 352 207.8 0.59 A02 21 Mar 98 07 Mar 99 351 367.3 1.05 A04 20 Mar 98 07 Mar 99 352 509.8 1.45 B-proile (d) (m) (md-1) B03 29 Mar 98 09 Mar 99 345 571.0 1.66 B04 29 Mar 98 07 Mar 99 343 603.0 1.76 B05 29 Mar 98 07 Mar 99 343 613.9 1.79 B06 29 Mar 98 08 Mar 99 344 601.1 1.75 Bll 29 Mar 98 08 Mar 99 344 556.0 1.62 Table 4.20: Velocities of stakes drilled in March 1998. stake positions of November 1995. The maximum A-proile velocity of 1.45 md-1 at stake A04 as mean from 31 March 1997 until 20 March 1998 agreed with former velocities. No signiicant inter-annual variations in low velocities occurred also at B proile, even though diferences in stake velocity may occur due to slightly diferent drilling locations. B proile stakes were set up in March 1997 at distances between 112 and 241 m above the original positions. The average yearly displacement of the ive stakes of B proile between end-March 1997 and 23 March 1998 exceeded the respective yearly mean of the two summer periods 1995/96, 1996/97 and the winter period 1996 by 2.5%. Slightly slower velocities were observed from March 1998 until the beginning of March 1999 (igure 4-17).

In addition to stake measurements the velocity of a stone with a length of about 8 m moving at the glacier surface was obtained (table 4.21). In March 1997 the stone reached a position close to stake B06, similar low velocities were measred.

The obtained stake velocities show partly diferences to previous measurements (section 1.3). The high low velocities of 2.64 md-1 measured by Rafo [1953] during Chapter 4. Field measurements 51

2.0 • 97/98 1.8 0 98/99 1.6 1.4 t 1.2 d 1.0 s 0.8 0.6 0.4 0.2 0.0 AOl A02 A04 B03 B04 B05 B06 Bll

Figre 4-17: Flow velocities of A proile (A01-A04) and B proile (B03-Bll) stakes during the periods from the end of March 1997 until March 1998 and from March 1998 until March 1999. Date I Date II t (d) Displacement ( m) Daily Displacement ( md 1) 6 25 Nov 96 30 Mar 97 124.9 229.3 1.84 30 Mar 97 o 23 Mar 98 358.0 624.3 1.74 Table 4.21: Displacement of stone at the centerline of proile B.

.... 1 years have never been observed at A and B proiles. The motion ield from SAR da� ta (section 6.2) showed velocities lower than 2 md-1 in the region of Rafo's proile, which was set up less than 1 km from the A proile glacier-upward. Flow acceleration may not be expected due to mass continuity, as the glacier showed no narrowing at this part. A possible reason for the discrepancy may be inaccurate positions of Rafo's iron poles in 1950, as the poles were less than 2mlong and might have drifted along the surface after melting out of the ice. In addition, long-term changes may not be excluded. Flow velocities were also obtained within the Japan-Argentina­ Chile joint Glaciological Research Project in Patagonia (GRPP) in November 1990 and between November 1993 and December 1994 (section 1.3, [Narusc et al., 1995b] , [Skvarca and Naruse, 1997]). In November 1990 11 survey points were established close to the A proile and from the A proile glacier downward [Naruse et al., 1992]. Velocities at the center up to 2.2 md-1 were measured. In contrast to this short mesurement period of one day in 1990, the average daily velocities at four stake sites were between 0.4 and 1.6 md-1 during the one year period 1993/1994; these values correspond to the velocities measured at A proile (igure 4-14). Velocities in November/December 1993 were slightly larger than the annual mean 1993/94 [Skvarca and Naruse, 1997] . Chapter 4. Field measurements 52

Figure 4-18: Juan Carlos Quinteros triggers of 2 kg of explosives at the lower transverse A-proile

4.4 Ice thickness

Ice thicknesses of Moreno Glacier were determined by using the seismic-relection method along A and B proiles in November 1996 (igure 4-19). A. Eckstaller was responsible for the seismic work. Registration was carried out on a 24-channel "Strataview" (series R, technical speciications in appendix D) seismograph. Seismic signals were picked up using two geophone-strings, each with twelve geophones, at a spacing of 10 meters along lines in direction of the proiles. During the measurements 100 kg of the fast burning plstic explosive "Austin Powder" were used s source for the seismic waves. The explosive charges, each containing between 1 and 2 kg explosives, were placed in little crevasses or in water puddles, and detonated of by the explosive expert Juan Carlos Quinteros (igure 4-18). 12 seismic shots with a total of 23 kg of explosives were used at 4 seismic stations of the A-proile, 35 seismic shots were detonated at 14 stations of the B-proile. The positions of the 18 seismic stations forming the seismic proiles are shown in igure 4-19. A iat glacier bed in low-direction, a horizontal line of geophones and a ho­ mogeneous isotropic medium are ssumed in order to reduce the data analysis to a two-dimesional formulation [Clarke and Echelmeyer, 1996]. Input data are the travel time of the seismic waves, the shot-to-geophone and shot-to-seismograph of­ sets. The depth of the glacier bed and the shape of the basal cross section are obtained. The way of the wave from the shot point to the relector and back to the geophone, vpt, is given as 53 Chapter 4. Field measuements

Figure 4- 19: SPOT image {23 Aug.1995) with A- and proiles shown dotted B- lines and positions of seismic stations.

� b � s

Figre 4-20: Sketch of the geometry of the seismic waves, showing the parameters used for data analysis, after (Clarke and Echelmeyer, 1996] .

Shot Point

Geophone

Seismgraph

Glacier Bed Chapter 4. Field measurements 54

(4.3)

where Vp is the P-wave velocity and t is the two-way travel time for a given geophone, dis the length of the perpendicular line from the bed to the shot point, s is the horizontal distance along the surface from the shot point to the geophone, and 3 is the bed slope (igure 4-20). The paths from the shot point to two geophones ((vpt1 ), (vpt2)) leads to the solution for the distance d:

(4.4)

For a given distance s the glacier-bed slope 3 can be obtained from equation 4.3. The glacier depth results for a certain shot-to-seismograph ofset rom 4.5, y assuming a constant slope 3. The calculation of a bed-slope in low direction would require additional geophones oriented in low direction.

d- = ysin j 4.5 _ ( ) COS j A. Eckstaller estimated the seismic P-wave velocity by analyzing the running time of the direct P-waves between the shotpoints and the geophones (d 0 in igure 4-20). For the diltancel between the irst and the last geophones the seis­ mograms showed a time lag of 29.24 msec for the direct P-waves in the average of 34 seismic shots along the B-proile. For a maximum distance of 110 m between the geophones of the two geophone-strings this time lag corresponds to a velocity of

3762 ms-1. Inaccuraciesx in the determined velocities may occur due to diferences between the theoretical maximum geophone distance of 10 m each (with totally racked cables) and an efective distance due to surface undulations and slight di­ rection deviations in crevassed zones. The reduction of the maximum geophone distance of 110 m by 10% would result in an average P-wave velocity of 3386 ms-1. Assuming a slight reduction of thXe geophone distance due to the undulating sur­ face and neglecting variations of wave propagation due to variations of the elastic properties of the ice [Miller, 1982], the P-wa_--ve velocity is assumed to amount to 3600 ms-1 for the temperate ice of Moreno Glacier. Kohnen summarized seismic� velocity data rom the Greenland and Antarctic ice sheet in order to ind a rela­ tion between seismic velocities and temperature of ice [Kohnen, 1974] . According to Kohnen the mean velocity gradient dvpfdT amounted to -2.3 ms-1 deg-1 , the ex­ trapolation of measured velocities in ice with temperatures below -16.5°C resulted in vP =3795 ms-1 for temperate ice of 0°C. This value is in good agreement with measurements reported by Deichmann, who obtained a Vp =3770 ms-1 for Unter- Chapter 4. Field measurements 55 aargletscher [Deichmann et al., in press] . These reported velocities are higher than 3600 but for ice temperatures close to the melting point a linear relationship between seismic velocities and the temperature of ice fails [Thyssen, 1967]. P-wave velocities decrease signiicantly with increasing water content of ice. The P-wave velocity measurements on Alpine glaciers summarized by Thyssen [1967] resulted in values of about 3600 The relected seismic waves at proile A revealed a steady lowering of the glacier bed from the margin towards the centerline; 1650 m from the glacier margin an ice depth of 540 was measured (igure 4-21). This depth and the surface eleva­ tion of 360 m a.s.l. at the respective seismic station indicate a glacier bed going down to about 180 m below sea level (igure 4-24) and rising towards the calving terminus. The measrements along proile B revealed a subglacial trough with an approximately parabolic shape (igure 4-22). Ambiguous returns of the seismic P­ waves within 0.08 at the top of igure 4-22 resulted from atmospheric echo and - 1 surfacems wave, energy trapped by crevsses and undulations. The average surface al­ titude along this transect was 495 m a.s.l., the two-way travel time of the deepest relected signal describing the ice-glacier bed interface ws 0.38 seconds (shotpoint nr.26, igure 4-23). Assuming a seismic wave velocity of 3600 in temperate ice, 1 this deepest signal corremssp- onds. to 684 in depth; it is shifted to the north of the centre of the glacier. A strong rise of the glacier bed from the deepest point towards southern direction and a small plateau in the center of the transect gives evidence of geomorphologic diferencesm between the opposite sides of the valley. The northern glacier bed seems to be less resistant against glacier erosion (igure 4-24). Diferent relecting interfaces below the glacier bed may indicate diferent layers of moraines.

Geodetic lines and points 4.5 s On 25 November 1995 a geodetic reference point was set up at the base camp by means of DGPS relative to known coordinates of a point at a distance of 15.46 km at the Estancia Geronima. A screw was concreted on a rock near the camp 1 (igure 4-1) in order to ix the bse antenna's position. The coordims- nates of Estancia la Geronima are included in the Posicimones Geolesicas Argentinas (POSGAR), a geodetic network of 127 points which deine the National Geodetic System. The Instituto Geogniico Militar established POSGAR in May 1997. The geodetic points are uniformly distributed all over Argentina, their coordinates are related to the WGS84 ellipsoid. The coordinates of the summits of Cerro Perito Moreno (H. Rott) and Cerro Buenos Aires (P. Skvarca) were mesured on 4 and 5 December 1996 (table 4.22). Inferred heights of previous lake levels were measured at points in the surrounding the GPS base station at Brazo Sur on 4 April 1997 (table 4.23). The lower bound­ of ary of the dense vegetation ws ssumed as indicator of the maximum height of the lake level Brazo Sur during previous lamming periods. The estimated maximum of Chapter 4. Field measurements 56

Distance to margin km) South Center

-·· ·-

-

- "

- -

-· ···· -

- ...

0. 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 0.000 0.00 Figure 4-21: Seismograms along A-proile; the distance of the shotpoints from the south0.ern100 glacier margin is indicated. For the depth below the surface a propagation0. 100 velocity of seismic waves in ice of 3600 m/sec and below ice of 2800 m/sec was assum0.2ed.0 A band-pass ilter with a corner frequency of 175-300 Hz was applied0.20 to theE signal. (analysis by Alfons Eckstaller). 0.300 0.300 5 0.400 0.40 :0

. � O.S(X) O.S

0.60 0.()

0.7(l 0.70 Fixpoint Latitude Longitude HAE BL G PS Base Moreno 50d 29' 22.9615" 73d 02' 43.2741" 215.3 0.00 s w Est. la Geronima 50d 33' 42.1703" 72d 51' 31.7820" 248.6 15463.4 s w Co. Perito Moreno 50d 30' 54.6593" 73d 04' 26.7325" 1746.9 3812.5 s vv Co. Buenos Aires 50d 26' 39.7965" 72d 59' 25.8816" 1484.4 6495.8 s w Metstation 50d 29' 22. 7079" 73d 02' 48.0017" 204.8 94.1 s w Table 4.22: WGS-84 coordinates ofixpoints; the baseline length (BL) is the distance to the reference station in meter, HAE height above ellipsoid (m).

= Chapter 4. Field measurements 57

Depth (km)

8 8 8 § § § .,8

8

8

8

8

j 8

p p p 8 J

.. 0 0 8 0 ., § � 8 3

> § �:r � 3 8

�0 8

.. 80

§

.. ii" §. .. 8 0-

.. �) - 8 § § 8 8 8 § N

- Figure 4-22: Mosaic of seismograms along B-proile; the distance -to the southern glacier margin is indicated. A band-pass ilter with a corner requency of 175-300 Hz was applied to the signal. The white line shows a second layer of P-wave relection below the glacier bed (analysis by A. Eckstaller). .,

N

w N p p p p

.. 0 > N Chapter 4. Field measurements 58

·· · ·· · · · . . . -U1 . . 450 . . : . . : 400

- ····-- --- ...... 350 ·· ··-·· ·· ·····- ·· ·i �e ;; 300 e 250

200 ­ _ ISO ) 100 100 � 50 50 5000 ------�----����---. 0 1· · 0 ;� 1000 2000 . . 3000 4000 South Distnce [m] ... ·· �:� North

2 e : e 12 14 17 : II 5 & 24 27 3 . . I 5 10 911 13 ' .o 1-. .IO . ' ·� .aa � Figure 4-23: Seismic data at transect B. Points mark the, 31 -�-�0- intercept (two-way-travel) time of the seismic shots for the ice-bed interface. The interpolation line is shown. i Columns at the bottom� --� repr-- esent the ofset distance between the irst geophone and the·:� shotpoin t.

lake level Latitude Longitude RAE Baseline Length max 50d 29' 23.4726" 73d 02' 39.5832" 216.72 74.47 s w 1988 50d 29' 23. 7278" 73d 02' 42.9300" 212.19 24.82 s w 1997 50d 29' 19.8702" 73d 02' 32.9843" 193.05 225.31 s w Table 4.23: GPS measurement sites of actual and inferred previous lake levelsl , the ellipsoidal heights (RAE) are contained; the baseline length corresponds to the distance between the GPS measurement site and the GPS reference station. damming-height was about 23.5 m above the lake level of April 1997. This height stands in contrast to former publications; Liss [Liss, 1970] reported a maximum damming-height of about 28 m in January 1966, Rafo reported a damming-height of more than 26 m in 1956 [Reinsheimer, 1958], (section 3.3). Some relicts of vegeta­ tion and soil, covering the coastal rock nearby the GPS reference station, indicated the approximate height of the maximum lake level in 1988 (personal communica­ tion P. Skvarca), when Brazo Sur was dammed the last time by the glacier. The damming-height in 1988 was about 19 m above the lake level in April 1997 (table 4.23). 12 lines were measured with DGPS along the coast of Brazo Sur and along parts of the glacier margin; these data were used for georeferencing satellite images (igure 4-25, table 4.24). The last GPS line in table 4.24 (Nr.12) corresponds to lines 6, 7 and partly Chapter 4. Field measurements 59

i l ]._\\ .. · · , ... g ·

g ____ _ I ::: j : jI ·-�-�,, NW - -·- A- Profile: Cross Se(m)ctio n

600 ..

500

400 I I 300

200 ... ______. ___

100 11 �-. I g ///./' · · · ··-.. , · ''· · --,� __ �- ..... , - 3 00 J . �,· 0 500 1000 1500 2000 2500 3000 3500 4000 4500

. . . � . - llll-- SE Distance

B-Profile : Cross Sec(m)tio n 600

500

Figure400 4-24: Cross sections at A- and B proiles; vertical exaggeration 2.12. _ ...... _ .... _ .. _ .. _ ... � _ .. ___ _ .. _ .._ . _ - _ _ _ · _ _ _ 300 - _ _ _ . ______. __ 200 .. , 100 -·· ········ 11 ·

0 :: 0 .-� ...... -100 · · J · . t···· 11 ...... �� · ...... > , 1 -200 .. "' .. · · Nw - l · · . · -300 - · 0 500 1000 1500 · 2000 •2500 3000 3500 4000 4500 · Distance Chapter 4. Field measurements 60

_:- ! , ···-··-··-······-················· ·--··-··-· ·-··-··-·····-·· ······ ····-······-·· 1' -· ··- ·-··-··-··-······-·o· · . . .

Figure 4-25: SPOT image from 23 August 1995 with geodetic reference lines along the coast and the glacier margin. Lines 1-10 were measured in March and April 1996, line 11 in November 1996; projection: TM, datum: WGS84; grid: 2 km2 • !

+ : � "' + , ;· � Nr. Date Length (m) Pos. Nr. Description 1 25 Mar 96 5097.28 2002 shore Brazo Sur (north cost) 2 26 Mar 96 3180.96 480 shore Brazo Sur (south coast) 3 26 Mar 96 2794.06 457 shore Brazo Sr (south coast) 4 27 Mar 96 544.69 144 right margin above Buscaini 5 27 Mar 96 3831.58 803 right margin Buscaini to A-proile 6 28 Mar 96 2114.38 628 right margin from shore upward 7 28 Mar 96 884.13 255 right margin to A-proile 8 30 Mar 96 3339.91 807 left margin 9 30 Mar 96 1187.42 249 shore of P. Magallanes 10 6 Apr 96 3363.79 880 shore of P. Magallanes 11 29 Nov 96 1666.09 480 left glacier margin 12 1 Apr 98 4266.09 1378 right margin from shore upward Table 4.24: GPS lines of shores and glacier margins. The date of measurement, the length of the lines in meters and the number of measred positions are included; the Buscaini camp is located up-glacier at a distance of 8 km from the base camp. Chapter 4. Field measurements 61

Figure 4-26: SPOT image with GPS lines at the orographically right glacier margin.

Nr. date length pos.nr. echoes nr. description 1 28 Mar 96 12670 698 132 Brazo Sur 2 1 Apr 97 7185 1662 91 Canal de los Tempanos Table 4.25: GPS locations describing the trace of the rubber boat during echosound­ ing; date, length of the lines in meters, the number of positions and the number of echoes are indicated

2. The repeat mesurements were carried out to detect possible changes of the orographically right glacier margin. The line reached a length of 4.26 km from the GPS base station along the shore of Brazo Sur and the glacier margin (igure 4-26). hereas the main part of the repeat line (Nr.12) revealed only minor changes since March 1996 (igure 4-25, lines 6 and 7) , a glacier advance of 45 m sideways occurred where the right margin extends into Brazo Sur.

4.6 Lake depth

The lake depths were mesured close to the calving front. An echo sounding instru­ ment was used, the mesurements were carried out in March 1996 and April 1997. Table E.1 (section E) shows the technical speciications of the instrument, which hs originally been manufactured for boats to mesure water depth down to 199 m. Bathymetric lines are plotted in igure 4-27, the positions were mesured with GPS. Chapter 4. Field measurements 62

The measured maximum depth in Brazo Sur was 110 m, whereas Hauthal mea­ sured a mximum of 137 m in approximately the same region of Brazo Sur (section 3.1) . On 1 April 1997 echo sounding was carried out in the Canal de los Tempanos during a period of weak winds and a smooth lake surface. The rubber boat crossed the channel twice with a minimum distance of 40 m to the glacier front . The deepest of 91 measured points was 164 m in the eastern part of Canal de los Tempanos (igure 4-27) . Chapter 4. Field measurements 63

Figure 4-27: Lake depths at selected points along the plotted tracks of the rubber boat. Lines 1 in Brazo Sr: 28 March 1996; lines 2 in Canal de los Tempanos: 1 April 1997. The positions of the depth measurements are marked with black triangles. Chapter 5

Regional climatology

Climatological trends may partly explain the prevailing glacier retreat in Patagonia during the 20th century, though a non-climatic terminus response of calving glaciers is well recognized [Warren, 1994] . Precipitation and air temperature data of some stations in the Santa Cruz province show long term trends [Naruse and Aniya, 1995]; Burgos et al. described possible climate changes [Burgos et al., 1991]. Thorough descriptions of the climate of Argentina and Chile were given by Prohaska and Miller [Prohaska, 1976], [Miller, 1976]. Patagonian glaciers are inluenced by Antarctic and mid-latitude atmospheric circulation patterns [Warren and Sugden, 1993], and terminate in a diversity of cli­ matic environments. The irregular topography of the Andes athwart the westerlies causes sharp local contrasts of climate. Abundant precipitation in the accumula­ tion areas and high melt rates at the low altitudes of the outlet glacier termini are responsible for steep mass balance gradients. Wet air from the Paciic Ocean and the formidable topographic barrier of the Andes dominate the regional climate. The extraordinary steep west-east precipita­ tion gradients are typically of the order of meters per year over tens of kilometers [Warren and Sugden, 1993].

5.1 Meteorological stations and data

In Patagonia long-term climate data are sparse. The most important meteorologi­ cal stations used for the description of Patagonia's climate are comprised in table 5.1 (igure 5-1). The Argentine Meteorological Service operates 29 meteorological stations in the Province of Santa Cruz east of the southern Patagonian Andes. The longest record of climate data comes from Punta Arenas at Fagnano, which is situ­ ated 250 km south of the SPI; the Salesianeian priests operate this climate station in their monastery since 1888. Lago Argentino station (L.A.) is located 60 km to the east of the front of Moreno Glacier at the airport of Calafate. Climatic records of L.A. are available since 1937, although there are several gaps. Studies of Patagonia's long-term climate are diicult

64 Chapter Regional climatology 65 5.

74°W 66°W

Figure 5-1: Map of southern Patagonia with the locations of meteorological stations. SPI: Southern Patagonia Iceield, NPI: Northern Patagonia lceield. Chapter Regional climatology 66 5.

Station record since lat os lon ow alt (m) country Cabo Raper 1928 46.5 75.4 40 Chile Isla San Pedro 1932 47.4 74.6 22 Chile Torres del Paine 1964 51.2 72.9 46 Chile Islote Evangelistas 1899 52.4 75.1 49 Chile Bahia Felix 1915 53.0 74.1 15 Chile Punta Arenas* 1905 53.0 70.9 34 Chile P. Arenas at Fagnano 1888 53.1 70.5 28 Chile Ushuaia 1931 54.8 68.3 6 Arg. Punta Dungenes 1900 52.2 68.3 5 Arg. Rio Gallegos* 1927 51.6 69.3 17 Arg. Lago Argentino (L.A.) 1937 50.3 72.3 220 Arg. Moreno Base Camp (B.C.) 1995 50.5 73.0 192 Arg. San Julian* 1937 49.3 67.8 62 Arg. Puerto Deseado 1937 47.7 65.9 80 Arg. Comodoro Rivadavia 1931 45.8 67.5 61 Arg. Table 5.1: Meteorological stations located around the Patagonian Iceields south of 46rS. Marked stations (*) have been moved from urban areas to the respective airports. as the climatic records are mostly short or the stations were moved in the past. For example, the meteorological station of Punta Arenas moved from the Radioestaci6n Naval to the airport at Bahia Catalina and to the present airport at Chabunco, which is 24 km north of Punta Arenas. Some stations are situated in towns, where increasing industrialism afect the measurements. Homogeneous climate data are available from the climate station at the Moreno Bse Camp (B.C.) since November 1995 (section 4.1). In the following sections the data are compared with L.A. and with Ventisquero Moreno (precipitation) .

5.2 Air temperature

A warming trend south of about 46°S during the second half of the 20th century ws reported by Hofmann [Hofmann et al., 1997] , Rosenbluth [RosenblUth et al., 1995] and Ibarzabal [Ibarzabal et al., 1996]. At io Gallegos the mean annual air temper­ atures increased from about 6.5°C to 8.0°C from 1938 to 1988. The climatic signif­ icance of this observed warming is not deinitely proved, as the temperature record reveals a discontinuity due to a shift of the meteorological station at the beginning of the ifties. Additionally, there are gaps in the record until 1960. This warming tendency almost disappears at Punta Dungenes, situated only 100 km southeast of Rio Gallegos (igure 5-1). Weak warming occurred at L.A. from 1937 until 1990. The warming trend observed by RosenblUth and Ibarzabal results mainly from some relatively cold mean annual temperatures around the year 1971 and higher temper­ atures in 1983, 1985 and from 1987 to 1990. This implied warming trend at L.A. Chapter Regional climatology 67 5.

9, 0

8, 5 � ., 8, 0 .. . 1 ' ·. .. · I'. : . . � I ' II . 7, 5 . . ·. . ·. ·. . •' .. .. � , �·�. . . -. : ·. : . .. . . �· . ' ·. . 7,0 \ '. � ! i 6, 5 �

6, 0

1940------,------� 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995

� Figre 5-2: Fluctuations of the mean annual air temperature at Lago Argentino Station for the period from 1941 to 1998. t

) : becuomes less pronounced by extending the observation period to; more recen' ..·:t years 0 �· � : �: • • • until 1997, the mean annual temperatures from 1991 to 1997 were. · equal • 1993 • .. ) • :: H ( ) or beI low 8°j C igure• 5-2 . In 1998 the mean annual temperature at L.A. reached a J) ! • ; • maximumd value( of 8.4°C) especially due to the very warm months of February:: and H . ••

July; in July the mean temperature was 3.3°C •above the long term mean value.

obtained at 2 site•s within the Japanese Glaciological Research Pro ect in Patagonia GRPP from December 1993 to November 1994 and at B.C.. Thej 2 Japanese mea­ surement( ) sites were located at about 200 m a.s.l. and 330 m a.s.l. in the forest close to the Moreno Glacier. Hourly mean temperatures were recorded continuously with a data logger for 1 year Takeuchi et al., 1996 . The monthly mean air temperatures at the lower station ranged[ from 0.3°C in June] 1994 to 11.3°C in January 1994 igre 5-6 . No seasonal variations of the lapse rate between the 2 stations were observed( ; a) mean annual value of 0.8°C 100m was detected.

Fluctuationss of the monthly mean air/ temperatures and the monthly extreme values at B.C. are shown in igure 5-3 for the period from December 1995 until February 1998. The warmest and the coldest months were January 1999 11.3°C and July 1997 0.7°C . The comparison of the years 1996, 1997 and 1998( reveals) similar mean temperatures( ) during the months March '9.1°C , April '6.6°C and ( ) ( ) December '9.3 oc). Mean temperatures between 8.9°C and 10.1°C were measred during the (summer months December to February. The seasonal temperature luc­ tuations at B.C. are given in table 5.3. The mean diurnal cycle of temperatures shows signiicant diferences between the 3 winter seasons igure 5-4 . With the ( ) Chapter Regional climatology 68 5.

2 5 . . . '. . 2 0 L.i .. . • I • · . u 1 5 'i , .

� 10

5

· •• .• " I. • • • •• • � l • � 0 , �- .4 . l· 4 'l . "+ Jan Mar Ma, Jul � Jan Mar. May Jul srp Jan Mar �y Ju· Gp Jan

• ·. . A . j .. -10

Figurel 5-3: Monthly air temperatures at the Moreno BaseA Camp rom December

1995 unj til.. February.· ' 1999. The diamonds mark the mean temperatures; triangles A mark the monthly_ maximumA : .i (upper line) and minimum temperatures (lower line). . I temperatureA (°C) mean A. maximum minimum 1996 6.9 21.6 (12 Dec 96) -5.1 (5 Jul 96) 0 1997 5.9 22.1 (4 Nov 97) -9.5 (6 Jul 97)

1998 7.1 24.3 (21 Feb 98) -5.2 (15 Jun 98) ...... ::I A-.. w A' Table 5.2:I I Meao n- * and I I extremeA : I values : : : ..of' airI I temp! :erature i s : during : l I3 years i : I at l Moreno: : ! ,Se "Nov Nov A. . Nov . . .· . Bse Sta1996Ation. .._ .A ·.1997. ·h ·: 1998·A · � 1999 -5 6" •• : i

. exception of August 19.A'96 the Austral winter �(Ma y-Sep.) of 1997 was .'compa ratively. cold, a mild winter occurred in 1998. Diurnal amplitudes of extreme: temperatures were in the average 8.1°C in the summer season (Dec.-Feb.) and 6.1°C in the winter season (Jun.-Aug.) (period Dec 1995 - Feb 1999). The mean annual and extreme air temperatures of the three years (table 5.2) relects the cold temperatures of the year 1997, which were in the average 1 oc colder than in 1996 and 1998. The daily mean air temperatures were compared for the 2 stations L.A. and B.C. for the period from 16 November 1995 until 1 September 1998. The moving average of 10 days is shown in igure 5-5. The correlation coeicients are given in table 5.5. Seasonal temperatre luctuations are higher at L.A.. During the cold seasons the temperatures are similar or even slightly colder at L.A., whereas in summer L.A. exceeds B.C. (table 5.4) . A similar tendency was obtained from the Japanese temperature measurements from 1 December 1993 until 30 November 1994 [Takeuchi et al., 1996] , (igure 5-6). The reson for this diference is the location of the station L.A. in a dry zone, whereas B.C. is located at the lee of the SPI in a transition zone between the wet climate and the semi-arid climate to the east. B.C. may also be inluenced by the large glacier area to the west and the damping efect of the lake. ( :hapter Regional climatology 69 5.

12 l 11 10 9 8 7 3 ! 6 a 5 � 4 <

2 1 0 0 6 9 12 15 18 21 0

-sumer 95/96 (dec- feb) •--- winter 96 ( jun-aug ) sumer 96/97 (dec-feb ) winter 97 ( jun-aug )

u + sumer 97/98 (dec- feb) -winter 98 ( jun-aug )

�:- sumer 98/99 (dec-feb) ....

0 .... Figure 5-4E: Mean diurnal cycle of temperatures during diferent winter- and sum­ mer seasons at B.C.. The hourly values are the mean of all temperature measure­ ments (10 sec. 3interval within the previous hour. The time scale corresponds to .. ) Argentinian Time AT=UTC-3 hours . -· ( • ) • • • • • • • •

• • 3

•

.

• Chapter Regional climatology 70 5.

-

18 Base Camp Lago rgentino

16

14

Figure 5-5: 10 day moving average of daily temperatures measured at Lago Ar­ gentino and Moreno Base Camp.

5.3 Precipitation

No systematic precipitation measurements have been carried out on the Patagonian Iceields [Escobar et al., 1992]. Very high precipitation and no uniform trends can be derived from long term records of Chilean meteorological stations [RosenblUth et al. , 1995]. Statistical anal­ yses of precipitation data revealed high annual variation in precipitation, but small seasonal variations. The typical annual precipitation at Bahia Felix (table 5.1, ig­ ure 5-·41) amounts to 4000 the maximum of 7000 was measured in 1929. The Chilean coastal zone experiences a weakly pronounced seasonal variation in precipitation; the maximum monthly precipitation at the stations of igure 5-7 occurs typically between early autumn and winter. No clear seasonal pattern has been observed at the meteorological station Torres del Paine (igure 5-1) which is located 15 km to the south end of the SPI [Pena and Gutierrez, 1992]. Signiicantly diferent monthly sums shown in igure 5-7 indicate the extraordinary steep west­ east precipitation gradients. Escobar et al. [1992] used precipitation maps based on data records from meteorological stations around the Iceields, in order to estimate annual precipitation of 6400 to 7400 mm on the SPI. The precipitation can fall as snow during any month of the year on the Iceield. The analysis of precipitation records at four Argentinian stations by Ibarzabal [1996] shows no homogeneous trends. The yearly precipitation sum at Rio Gallegos -1 1 amounted to about 250 mmmma for the, period rom 1927 unmmatil 1990.- The mean an- Chapter Regional climatology 71 5.

period temperature (°C) summer (Dec-Feb) 95/96 9.7 autumn (Mar-May) 96 7.5 winter (Jun-Aug) 96 3.5 spring ( Sep-Nov) 96 7.1 summer (Dec-Feb) 96/97 9.1 autumn (Mar-May) 97 6.2 winter (Jun-Aug) 97 2.1 spring (Sep-Nov) 97 6.4 summer (Dec-Feb) 97/98 9.7 autumn (Mar-May) 98 6.4 winter (Jun-Aug) 98 4.7 spring ( Sep-Nov) 98 7.3 summer (Dec-Feb) 98/99 10.3 Table 5.3: Seasonal mean temperatures at Moreno Base Station.

Lago Argentino Base Camp mean 8.3°C 6.9°C maximum 27.8°C (11 Dec 96) 24.3°C (21 Feb 98) minimum -11.4°C (5 Jul 97) -9.5°C (6 Jul 97) Table 5.4: Temperature statistics for the stations at Lago Argentino and Moreno Base. The values refer to the period from 16 November 1995 until September 1998.

period correlation coeicient total 0.91 autumn (Mar-May) 0.90 winter (Jun-Aug) 0.90 spring ( Sep-Nov) 0.84 summer (Dec-Feb) 0.79 Table 5.5: Correlation coeicients for the 2 daily mean temperature records of Lago Argentino and Moreno Base Camp. The coeicients refer to the period from 16 November 1995 until lO April 1998. Chapter Regional climatology 72 5.

16 Lower Station L.A.

······--- · .... . ···-··· ... ··································-······················-·············· ··· ········-·······-·········· -· 14 , � ------· ----- 12 ------,------· 10 ______; ______

8

6 __.: ---· ...... -·--··············-···-· .. ····-··· ...... ····-··--··-···· . ·-· · ...... - / \, -\.:.� �:� , � 4

------:: ------:� ------2 .

: - 0 ------S_i:-�---�-: ./------2

Dec- Jan- Feb- Mar- Apr- May- Jun- Ju1- Aug- Sep- Oct- Nov­ � w w w w w w w w w w w

; . •• Figure 5-6: Monthly. mean air-- temperatures at the Ja---•pa---nese lower station Moreno .. terminuus , ""200 m a.�---s.l.., and at Lago Argentino Station L.A. from December( 1993 / ) ' ( ) to November 1994. ··"<::�

0 .J nual pre) cipitation sum at Lago Argentino was 210 mm during the period of 1937 . E:J - 1996.- Figure 5-8 shows extreme values of 364 mm in 1963, when extremely high J . precipitation) occurred from April to July, and of 60 mm in 1988. The decade of .. \ ' the 1980's is considered as very dry at the\ Lago Argentino station, -annual sums less than 100 mm characterized the years 1984, 1986 and 1988, in 1987 the sum was 103 mm. Measrements at Lago Argentino, at the Ventisquero Moreno and at Moreno Base Station reveal a strong precipitation increase towards the SPI. A mean annual precipitation of 1330 mm was measured by means of a totalizator at Ventisquero Moreno for the period from 1990 until 1997 table 5.6 . A short term record measured with an electronic raingauge at Moreno (Base Station) is available from 10 December 1996 until 4 April 1997 igure 5-9 . Although Moreno Base Sta­ tion is located in a distance of only 3 km to( the totaliz) ator of Ventisquero Moreno, precipitation data are diferent. The monthly sums at Moreno Base were 127, 54 and 39 mm for January, February and March 1997. Whereas in January and February the measured precipitation at Ventisquero Moreno was 220 % of the precipitation at loreno Base, the March precipitation was only 31 % of Moreno Base. The difer­ ence is certainly partly caused by wind efects on the precipitation measurements, but other error sources cannot be excluded. 73 Chapter 5. Regional climatology

------

Cabo Raper Isla San Pedro l:o (1931-60) (1911-47) j j

...... _ ' ' ......

.. . .

......

, ' . , ' . ' . . " . . . " c c � ' > c � c � > 8 i 8 i �

Isloe Eangelisas Bahia Fel ix (1911-59) (1931-60) j � ...... ' ' ......

.. . .

......

, ' . ' . " " . . c � � c � ' > 8 i

500 Puna Arenas Lago Argenti no 400 (1911-59) 400 (1937-96) c c 0 0 j 300 j 300 .. 200 .. 200 ... .. ! ... .. ' � � ' � � ! ..u 100 ..u 100 .. . .

.. .. 0 ...... � .. � 0 < 0 < Q 0 < 0 <

, ' . , ' . ' ; " " . . . " c � c � ' > c � c 8 i � � 8 �

500 - 500 Ve ntisquero oreno Torres del Paine 400 (1990-97) - 400 (1983-90) c c 0 0 300 300 ! j

... 200 . . 200 ...... ' ' � � ..u 100 ..u ...... 100

.. ... 0 0 ...... � .. .. 0 < 0 < Q

, ' . ' , ' . ; " " . . ; . " . c c ' c c ' � � 8 � 8 � �

500 500

400 400 c c 0 0 Figure 3005- 7: Sesonal variation of precipitatio300n at diferent stations. The data

records200 refer to diferent periods. Source of 200Chilean data: [Meteorolgica, 1966) , . . .. � � � � ! [Penau and100 Gutierrez, 1992] . u 100 z

0 0 . . . . � � � 0 < 0 < 0 < 0 <

500 500

400 400 c c 0 0 300 300

200 200 . . � � � .. � u 100 z u 100

0 0 . � . � � 0 0 < Q 0 < 0 < Chapter Regional climatology 5. 74

400

350

300 E E 250

= 200 ·. ·o 150

100

50

0 ------1935 1945 1955 1965 1975 1985 1995

Figure 5-8: Annual precipitation sums at Lago Argentino. Breaks in the line are due to data gaps.

....

....� ::: 0 � . 30 ..-

25

., 20

·.. 15 ·:;

10

5

0 Dec 96 Jan 97 Feb 97 Mar 97

FigureE 5-9: Daily precipitation at Moreno Base for the period from 10 December 1996 unE til April 1997. 4 - =

c 0

J -

� ... .. Chapter Regional climatology 75 5.

1990 1991 1992 1993 1994 1995 1996 1997 1998 Jan 89 65 25 59 67 46 126 280 272 Feb 32 114 47 163 64 66 145 122 254 Mar 206 191 44 391 173 85 150 12 136 Apr 50 296 158 57 99 170 165 166 May 148 87 215 243 330 190 Jun 292 14 42 59 27 144 88 Jul 231 44 33 68 19 329 72 Aug 222 108 98 101 86 26 154 218 Sep 86 39 84 80 14 185 74 62 Oct 197 10 173 148 21 32 172 131 Nov 172 64 15 59 53 71 166 63 Dec 38 94 87 79 103 157 206 107 Sum 1760 1038 805 1348 938 1551 1845 1351 Table 5.6: Totalizator measurements of precipitation at Ventisquero Moreno. The annual sums are partly inaccurate due to data gaps.

5.4 Wind

The wind is one of the most important climate elements of the Patagonian region. Measurements of the wind direction reveal dominant western and southwestern di­ rections at most stations [Endlicher and Santana, 1988] . The wind at the Chilean climate station of Evangelistas at the Paciic coast (igure 5-1) blows mainly from northwest. The wind speed at Punta Arenas is high during spring and summer, during 87.5% of the time the speed is higher than 2 (observed period: 1977- 80) [Endlicher and Santana, 1988] . The maximum wind speed at Punta Arenas was registered in 1956 with 42.7 speeds exceeding 25 may be observed all over the year. The 1996, 1997 and 1998 wind roses at Moreno Base (igure 5-10) show winds from the west and southwest during 85% of the time. The very persistent direction of winds is related to the orientation of the glacier terminus. The annual average of the wind speed ranged from 4.4 in 1996 to 3.6 in 1997 ( 4.2 in 1998). The monthly average wind speed is low during the winter in June or July (igure 5-11); the lowest monthly average wind speed of 1.8 occrred in June 1997. Though windspeed increases in the warm season, the highest measured wind speed of 31.4 occurred at M.B. on 18 July 1996. The highest monthly average windspeed was 5.9 in January 1999. Figure 5-12 shows the mean diurnal 1 cycle of wind speed and the frequency of westerly windsms- . The mean hourly wind speed ranges in summer from about 3.7 in the morning to a maximum of 7 I 1 in the afternoon. The msdiu-rnal, cycles of the lowerms wind- speeds in winter of the years 1996 and 1997 show a mean daily amplitude of only 0.8 In winter 1998 higher speeds occurred varying between 3.3 and 4.8 Wind direction measurements reveal also a diurnal cycle; with increasing wind speed the wind comes

1 1 1 ms- ms- ms-

1 ms-

1 ms- 1 ms-

1 ms- 1 ms- 1 ms- . 1 ms- . ( :hapter Regional climatology 5. 76

N 60 %

sw SE

s

Figure 5-10: Wind frequency measured at Moreno Base. Three almost identical lines show the wind direction of the years 1996, 1997 and 1998. almost exclusively from the western direction as a topographically induced efect.

5.5 Pressure, humidity and radiation at Moreno Base Station

15, 20 and 30 minutes mean values of air pressure, humidity and global solar radi­ ation at B.C. are available for the period from 15 November 1995 until 22 March 1999 (appendix, section C.0.2) . The mean diurnal cycles of air pressre, humidity and radiance for 4 summer and 3 winter seasons are presented in igures 5-13 and 5-14. Figure 5-14 shows additionally the hourly maximum values of the measured global solar radiation during the diferent seasons. Hourly luctuations of the sea­ sonal mean pressure are of the range of 1 hPa, the 2 daily maxima occur in the night and in the late morning hours. The colder air temperatres cause higher pressures during the winter months, though also exceptionally high pressure was observed in summer 1997/98 due to a persistent high pressure situation in February (table 5.7 and 5.8) . Monthly luctuations of the mean relative humidity shows minimum values in December and higher values in the winter months. Very high relative humidity occurred from May 1997 until July 1997; the diference of relative humidity during the wet winter season of 1997 and the winter of 1996 was in the average 6%. Typical maximum values of the relative humidity during the summer seasons are between 60% and 65% in the moning (igure 5-13). Chapter Regional climatology 77 5.

32 . . : : A "

. . .• . " A, : .··-. 28 . \ . . •• .. \ .. ' ...... : . ·. •· \ ' ' : • , .• : . ' 24 .. ········•· · ···· · ·. ,· • A ...... \ .. . . . ,.· , . . • ... ' . . . ., . . .. , . 116 � 12 8

' '

--,' ...._.-- ' ,.' I ' I ' ·- I ' r-··-r '-- ,- ., . ,---,--,---- T I ·-r- ·r -----,--,- - r ' ' ' ' ' ,- ' ' ' ' ' ' ' ' - ' ' ' ' ' ' ' . " . " . " :: :: :: � l . . � l � l � l � ; : ! Figure 5-11: Monthly . average.. ... and maximum wind speed at Moreno Base Station.

.

...... II ..� ...... :� .. . :; .. >" ' . The monthly su': ms.· of the global solar radiation are included. in tables and � 5. 7 . �, .• � : .. . . � . ', . 5.8. Figure 5-15 shows a comparison of the monthly sum-"·.s of global radiation at B.C. .. � .. . with the multi-annual...... period...... of the} Alpin. . e station at the University of Innsbruck .. (l1.4°E, 47.3°N, 577 m a.s.l.) . A higher mean annual radiation amplitude has been ' observed0 at B.C.. The global radiation shows a lattened annual maximum at the :: station Innsbruck. The maximum summer sums are around 600 M Jm-2month-1 in Innsbruck, a value of 643 M Jm-2month-1 has been obtained for B.C. in the average0 of the 3 years period 1996 to 1998...... � : : :

0 0 0 0 0 0 ...... , :J . :J . :J Q .. < < 0 Q .. < < 0 Q .. < < 0 Q .. Chapter Regional climatology 78 5.

7,0

6,5

6,0

5,5

! 5,0

4,5 � 4,0 S 3,5 3,0

2,5

2, 0 ·-·· --·------·------· ------·-· ··------·-· · ---- ·· · -- ·-·---··T ·· 0 3 6 9 12 15 18 21 0

-sumer 95/96 (dec-feb) -winter 96 ( j un-aug )

..- sumer 96/97 (dec-feb) •--winter 97 ( j un-aug )

-sumer 97 /98 (dec-feb) -winter 98 ( j un-aug ) -sumer 98/99 (dec-feb)

100% S llD 90% ! ...D 1 80% -l 70% :& � 60% i -j 50% •l �

30% • � · . / I •··-· · ··· •·· ...... ·· ·- · ··-·· 20% r , , r ------,------1

� 10% I 1 0%

0 3 6 9 12 15 18 21 0

-sumer 95/96 (dec-feb) -winter 96 ( j un-aug )

-slumer 96/97 (dec-feb) •-- winter 97 ( j un-aug )

sumer 97 /98 (dec-feb) -winter 98 ( j un-aug ) . -sumer 98/99 (dec-feb) A D i

D u1 .. Figure 5-12: SMean diu! rnal cycle of wind speed top and diurnal frequency of west­ A ( ) erly winds direction between 247.5 and 292.5 deg., below for summer and winter ( : ) seasons at B..C .. The time scale corresponds to Argentinian Time. Hourly values of 1 wind speed correspond•lD to the average of all measurements within the previous hour.

i

J ll j

··

- Chapter Regional climatology 5. 79

989

_ -· 988 _

987

986

985

984

0 3 6 9 12 15 18 21 0

95/96 96

96/97 97

97/98 98

•--- 98/99 . ... -•... •• ··-•-.... 11 '• .. . ·• ·- . · - . 80 ····· ··-···· ......

75

70

65

60

55 -sumer (dec- f eb) winter (jun-aug ) sumer (dec-f eb) winter (jun-aug ) 50su mer (dec-f eb) winter (jun-aug ) ... -. -smm0 er 3 6(d ec-f9eb) 12 15 18 21 0 .. •

95/96 _ 96

96/97 97

97/98 98

98/99

Figure 5-13: Mean diurnal cycle of air pressure (top) and relative humidity (below) for summer and 3 winter sesons. Mean values correspond to all measurements 4 within the running hour (10 sec. measring interval) ending at the hour of the day written in the x-axis. The time scale refers to Argentinian Time (AT=UTC-3 hours).

--- sumer (dec- f eb ) winter (jun-aug) s mmer (dec- f eb) winter (jun-aug) _sumer (dec- f eb ) winter (jun-aug) • sumer (dec- f eb) • -

. Chapter Regional climatology 80 5.

.

.

···Q···O· Q

-sumer (dec-feb) winter (jun-aug )

700sumer (de c-feb) ···+··· winter (jun-aug ) -sumer (dec-feb) --- winter (jun-aug )

···· <· ···600sum er (dec-feb)

! 500

Ne .. 400 :: 0 .. , 300 ' .. 0 ' : 200 < ' .. 0 ) 100 .. ' 0 0 0 .. ) 0 3 6 9 12 15 18 21 0

95/96 Q 96 96/97 97 97/98 98 98/99 • -s• umer (dec-feb) -- • --- winter (jun-aug )

sumer (dec- feb) -winter (jun-aug )

-s1200umer (dec- feb) �winter (jun-aug )

--.-- sumer (dec-feb) 1000

Figure 5-14: Mean diurnal800 cycle of global solar radiation for 4 summer and 3 winter seasons top . The maximum radiation recorded at the hour of the day for the ( ) 600 respective seasons is shown below. Mean values correspond to all measurements hour sec. measuring interval ending at the hour of the within the running 400 (10 ) day written in the x-axis. The time scale refers to Argentinian Time (AT=UTC-3 200 hors) .

0 0 3 6 9 12 15 18 21 0

95/96 96 96/97 97 97/98 98 98/99 • Chapter Regional climatology 81 5.

Month Pressure Humidity Radiation (hPa) (%) (M Jm-2month-1) Dec-95 985 55 682 Jan-96 983 56 659 Feb-96 981 58 489 Mar-96 980 57 371 Apr-96 982 67 196 May-96 982 69 93 Jun-96 989 68 77 Jul-96 993 70 104 Aug-96 983 69 183 Sep-96 987 59 325 Oct-96 985 60 505 Nov-96 978 58 543 Dec-96 984 55 687 Jan-97 978 63 573 Feb-97 985 60 488 Mar-97 990 59 447 Apr-97 980 70 193 May-97 990 78 123 Jun-97 986 85 68 Jul-97 984 75 94 Aug-97 986 64 191 Sep-97 990 66 319 Oct-97 985 63 520 Nov-97 984 63 564 Dec-97 985 57 685 Table 5. 7: Monthly mean values of pressure and relative humidity and monthly sums of global solar radiation. Chapter Regional climatology 82 5.

700

600

500

400 s s 300 � 200

100

0

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Dec

Figure 5-15: Monthly sums of global solar radiation at Moreno Base Station and

::at the Alpine station Universitat Innsbruck (Austria) in the average of the years J 19: 96-1998 (Moreno Base) and 1958-1997 (Innsbruck) . 0

N

.... Month Pressure Humidity Radiation (hPa) (%) (M Jm-2month-1 ) Jan-98 981 58 589 Feb-98 994 61 548 Mar-98 985 64 382 Apr-98 993 75 235 Nov May-98 992 86 123 Jun-98 990 77 81 Jul-98 977 72 63 Aug-98 982 67 152 Sep-98 991 58 333 Oct-98 982 62 372 Nov-98 985 59 580 Dec-98 984 60 555 Jan-99 987 54 692 Feb-99 983 59 450 Table 5.8: Monthly mean values of pressure and relative humidity and monthly sums of global solar radiation. Chapter 6

Surface motion by means of SAR

This chapter deals with the generation of the glacier velocity ield using repeat pass SAR images. To derive the velocity ield of a glacier a processing scheme was developed, which combines SAR interferometry and amplitude cross-correlation. The algorithm was applied on SIR-C data acquired during SRL-2 in October 1994 (section 2.2.3) in order to obtain the velocity ield of Moreno Glacier. The resulting velocity map represents the surface velocity over the short time period between the acquisition of the SAR images. The seasonal variation of velocity is discussed by comparing the motion derived from SAR with the motion of ablation stakes measured with DGPS (section 6.2.1).

6.1 Methods

SIR-C/X-SAR repeat pass images were acquired on 7, 9 and 10 October 1994 at L-band (. 24.2 em) and C-hand (. 5.7 em) with the nominal resolution Ra 8.2 min azimuth {along-track) and Rr 3.8 min slant range {direction of the radar beam), and at X-band (. 3.1 em). The X-band spatial resolution for single look is between 8 and12 min azimuth and 8.5 min slant range (ine mode). A combination of SAR interferometry (INSAR) and amplitude cross-correlation was used for the estimation of the velocity over large parts of the ablation area. Figre 6-1 shows the three main branches of the algorithm. The branch of the lowline analysis was combined with the interferometry branch in order to obtain velocity magnitudes (section 6.1.1). The algorithm of the amplitude cross-correlation shown as the third branch in igre 6- 1 yielded the velocity in regions of the glacier terminus where interferometry failed (section 6.1.3).

6.1.1 Interferometric motion analysis INSAR utilizes the interference pattern caused by the phase diferences between two repeat =pass SAR images [Massonet= and Feigl, 1998] . The phase diferences= result from the combined efects of top= ographic height diferences, the displace-

= 83 Chapter Surface motion by means of SAR 84 6.

- Optical Satellite Data SAR SLC SAR-Amplitude ( DEM - SAR Data ( Image Data ( ( I

Phase Calculation with Coherence Sensor and Analysis Orbit Geocoding [ J Parameters Cross - Correlation Map of I Interferogram ] Topographic ( Phase \ Manual Analysis of Flowlines Motion ( Interferogram � Velocity Field in I I Range and Azimuth

Phase \ \ Unwrapping Flow Direction J Interpolation

Range Component I 1 Geocoding of Velocity Magnitude ( � \ [

Digital Map of I Flow Direction � ( I Geocoding \ Velocity ] Magnitude V2

Velocity Magnitude Vl [ --( \ Frontal Areas with Phase Coherence Vl >=0.4: Phase Coherence V2 0. 4: 1 Map of Velocity Field

I

Figure\ 6-1 : Flow chart of ice motion analysis by means of repeat pass SAR images.

< Chapter Surface motion by means of SAR 85 6. ment of the scatterers and changes in the atmospheric propagation path length · [Zebker et al., 1997] . The basic principles of INSAR were described by Zebker and Goldstein [Zebker and Goldstein, 1986] , Goldstein and others [Goldstein et al., 1988] , by Geudtner [Geudtner, 1995] , and Schwabisch [Schwabisch, 1995]. The interferom­ etry software used for this study was developed at the Institut fr Meteorologie und Geophysik at the University of Innsbruck (IMGI) by A. Siegel [1996]; Siegel processed the SIR-C data for this analysis.

Phase coherence 6.1.1.1 Phase coherence is a precondition for the calculation of interferograms. Phase coher­ ence is reduced by the imaging and processing noise, by atmospheric efects and by surface efects. Surface efects result from changes of backscattering properties, for example due to melting, ice deformation at sub-pixel scale, the motion of scattering elements or other changes in the electrical and structural properties. The sources of system efects are thermal noise, and geometric decorrelation [Rignot et al., 1996] . The signal-t-noise ratio (SNR) depends on thermal noise and the sensor gain. Ge­ ometric decorrelation arises from the baseline separation between two antennas and from changes of the squint angle of the radar antenna. The phase coherence images of the image pair from 9 and 10 October 1994 are shown in igures 6-2 and 6-3. The C- and X-band interferograms of Moreno Glacier were not suitable for motion analysis, because the signal on the glaciers decorrelated almost completely within 24 hours. The surface of extended parts of the terminus was inluenced by melting due to a signiicant temperature increase at the beginning of October 1994 (igure 6-8). The temporal decorrelation for C- and X-band is quite pronounced on the glaciers and over forest, where volume-scattering from the upper canopy dominates the signal and the scatterers are randomly displaced due to wind. The coherence is better preserved along a small stripe of the lake-shore and the mountain slopes with bare rock or sparse vegetation. The highest coherence is observed at all three bands in the alluvial plane between Laguna Ameghino and Lago Argentino (section 9.1.1). The 1-band data from 9-10 October showed resonable coherence over the forests and over extended parts of the terminus [Rott et al. , 1998] . Figure 6-4 shows a coherence map of Moreno Glacier. A signiicant part of the glacier shows low coherence ( <0.4) . Low coherence is observed over glacier areas with high shear-strain and with tilting crevasses. In these areas it is hardly possible to derive ice velocity by means of SAR interferometry.

Motion analysis 6.1.1.2 The measured phase diferences, ¢12, between two repeat-pass images may be split up into 4 terms, Chapter Surface motion by means of SAR 86 6.

Figure 6-2: Phase-coherence of 9-10 October 1994 SIR-C SAR images at (1) 1- and (2) C-hand frequency. Low coherence is dark, high coherence is bright. Chapter Surfa ce motion by means of SAR 6. 87

Figure 6-3: Phase coherence of 9-10 October 1994 SIR-C X-SAR images at X-band frequency. Low coherence is dark, high coherence is bright. Chapter Surface motion by means of SA R 6. 88

� I

N

A 4m

>=0.8 0.8 0.7 Figure 6-4: Phse-coherence of Nioreno Glacier at 1-band frequency0.6 from SIR-C < 0.5

images of 9/0 October 1994. < 0.4 '

< 0.3

< 0.2

<

<

< Chapter Surface motion by means of SA R 89 6.

(6.1)

¢1e is a term describing the phase shift due to the lat earth, which can be calcu­ lated based on the SAR imaging geometry and the earth ellipsoid. denotes the Pat phase shift due to diferences in the atmospheric path length [Zebker et al., 1997] . The topographic phase shift >t is due to a stereoscopic efect, as the radar observes the topography from two slightly diferent points of view Mssonet and Feigl, 1998] . [ The topographic phase diferences >t can be removed by subtraction of a synthetic interferogram derived rom a DEM and by diferential methods using two interfer­ ograms. The simulation of the topographic interferogram requires very accurately known antenna positions. Phase diferences result also from target motion ( ¢m) in range. To derive the surface velocity we ssumed a coordinate system with and y axes pointing in ground range and azimuth direction, and a vertical (igure 6-5). By neglecting the measured diferential interferometric phase, ¢ , is related to the Pat 12 lat earth (¢1e) , to topography (>t) and target motion in range (¢m) · It is given by Siegel, 1996] : [

The irst term describing ¢1e includes the pixel spacing in slant range 6R. B121 denotes the component of the baseline between the antenns perpendicular to the radar illumination. B121 is calculated according to:

x (6.3)

z The inclination of the baseline is described by :12 . i is the local incidence angle of the radar illumination with the vertical. . denotes the wavelength, and the slant range distance between an observed point at the surface and the center of the synthetic aperture. h is the topographic height. Vx is the horizontal velocity in range, Vz correspond to vertical ice surface dis­ placements. Surface displacements parallel to the light direction cannot be mea­ sured by INSAR. The time lag between the two image acquisitions is (t2 - t1 ). Figure 6-6 shows an interferogram of the lower part of Moreno Glacier derived from L-band SIR-C data of 9/10 October 1994. Because of the short perpendicular baseline length of 29 m the sensitivity to topography is small. One fringe repre­ senting a ¢ of corresponds to an altitude diference of m, and to a surface 12 2r 620 displacement Dx r in ground range of 21.5 em (8x r . Narrow fringes on the 2 2 = 2 s i glacier indicate shear zones. � ) Based on the DEM (section 2.1) a topographic interferogram (¢ representing >t) r0

vy Chapter Surfa ce motion by means of SA R 90 6.

I ro I

- - .. --- - - .. Po - -- .. h - - .. -- _ _. :.lpl_ _ - _ I - +x ,.'' P \ , rt , \ . .. , ' , ...... -/ z

Figure 6-5: Geometry of IN8AR: 81 and 82 denote the positions of the radar anten­ nas, the observed points at the surface are P0 and P1 . Pr1 is the range displacement of Po by a distance .R projected on a horizontal surface. urther explanations of variables are included in the text (equation 6.2) .

was calculated (igure 6-7). The mximum altitude range of Moreno Glacier corre­ sponds to about 4.4 ringes. The surface motion in range is obtained by subtracting the synthetic from the real interferogram. The vertical motion compon�ent results from� the emergence velocity ( tan 3+ and the glacier thinning due to ablation b, is given by [Paterson, 1994]: aR - ) - - . --� -- -�-�- ---- �.. . - - - - \ b- tan 3 + (6.4)

\

< For a glacier in a steady state .. � 0), the emergence velocity levels ablation. ..( The angle 3 denotes the surface inclination in range, and is the vertical velocity at the surface due to mass continuity assuming incompressibility. At Moreno terminus 3 is about 2° to 3°. The slope component of the emergence velocity, ( tan 3), counteracts to ablation b. The magnitude of is estimated for the B-proile for steady state ( 0 in equation 6.4) . An ablation b between 3.0 em (central zones) and 4.0 em (marginal glacier) was estimated for the period between the 8AR data acquisitions of 9/10 October 1994 (23.617 hours) with the degree day factors of table 4.13 (section 4.2) . The same mean temperatures are assumed at the B-proile Vz -vx equal as those measured in m a.s.l. close to the glacier margin by Takeuchi w8 330 Vz [1996], (igure 6-8). Based on from the mean stake velocities (igure 4-15, section

4.3), the vertical estimated veVzlocit ies Vx along theW8 ,B-proile are between -5.6 and 4.0 cmd-1 (igure 6-9). For steady state the high horizontal velocities in the central region of the proile would cause = vzto change the sign in order to level surface lifts w8

=

vx

w8

vz

=

Vx

w8

w8 Chapter Surface motion by means of SA R 91 6.

,- 5" r + 7 , � n + "

Look Direction

Figure 6-6: Interferogram of the Moreno Glacier region, rom 1-band SIR-C data of 9/10 October 1994. Each color cycle represents an altitude diference of 620 m or 12.1 em of horizontal motion in the direction of the radar beam. The dotted :s line across the terminus corresponds to transect B. The rrow shows the velocity mximum. BS: Brazo Sur, CT: Canal de los Tempanos. Chapter Surface motion by means of SA R 92 6.

Look Direct ion

Figure Synthetic interferogram showing the topographic phase shift, based on 6-7: the system parameters for the SIR-C 1-band images of 9/10 October 1994. The look direction of the radar beam is indicated. BS: Brazo Sur, CT: Canal de los Tempanos. Chapter Surface motion by means of SAR 93 6.

12 Logo Argentino Upper Camp 10 GRPP � 8 � 6 4

� 2 < 0 -2 I I I I I I I I I I

I I I I I I I I I I

Figure 6-8: Daily mean air temperatres at the beginning of October 1994 measured at Lago Argentino and in the forest in about 330 m a.s.l. close to the orographi­ cally right margin of Moreno Glacier (GRPP Upper Camp), [Takeuchi et al. , 1996] , section 5.2.

-

up to about 8 ...cm d-1 . A st-eady surface forces an increase of towards the glacier margins due toE a decreasing velocity and an increasing ablation b. I .. a The observ..at ion of no striking surface elevation changes during the ield mea­ l surements (cha.pter 4) support the assumption, that the surface of the terminus is close to a stea.. dy state. Skvarca and N aruse mesured the surface elevation

at 11 points in the ablat0 ion 0 ar0ea of 0 Moreno 0 0Gla cier 0 [Skv0 ar0ca and0 Naruse, 1997] , t t t t t t t t t t [Naruse and Aniya, 19\92] . \ They \ fo und \ only \ small\ and\ no \ unif \orm changes \ in eleva­ ------tion, which could be mainlyu u addressed u u to u the movementu u uof th ue undulating u srface. .. N ) t 0 a ._ 0 \ 0 For the calculation of the interferometric phase (equation 6.2), the.. vertical motion was neglected. By eliminating the modulo-2r ambiguity (phase unwrapping), an image of the motion-related absolute phse was generated.

6.1.2 Analysis of low direction w8 Vx Because SAR interferometry provides the velocity component only in slant range, additional information is needed. Assuming ice low parallel to the surface, and a known horizontal low direction, the velocity vector can be calculated from The velocity was subsequently converted to two-dimensional horizontal displacements by analyzing the low direction, which is indicated by thin moraines visible along the glacier terminus. Flowlines were derived rom the analysis of Landsat and SPOT images, SAR images and aerial photography. The lowlines were manually digitized

andVz subsequently the orientation along the individual lines was calculated. A map of low directions covering the whole glacier terminus was derived by interpolating between the discrete azimuth angles.

Vx .

Vx Chapter Surface motion by means of SA R 94 6.

0,10 z(vx tan� 0,08 � ...... 0,06 � E 0,04 z(ws) � � 0,02 -

' 0,00 '

] -0,02 _,I 4000 0 z(-b) � -0,04 -0,06 I 4 .. -0,08 l- Ditance ( m)

Figure 6-9: Verticai l surface displacements along B-proile for steady state. The displacements refer to a period of 23.617 hours between 9 and 10 October 1994 i SRL-2 dataE acquisitions . ( I )

6.1.3 Amplitude cross-correlation ! .!! In contrstQ to INSAR amplitude cross-correlation provides both magnitude and ) direction ofu the velocity vector and enables the derivation of the glacier velocity ield. .. The accuracy depends on the pixel size and is therefore lower than for interferometry, but the amplitudes may correlate also in regions where INSAR fails due to low phase - coherence. L NNW SSE The cross-correlation matrix of image chips is calculated to determine the local shift. To increase the resolution the cross-correlation matrix is interpolated using a function, enabling the detection of shifts down to pixels. For matching sine 1 the method uses glacial surface featres, such as crevsses 1or0 moraines, and image speckle. In the case of SIR-C L-band data of Moreno Glacier, image speckle decor­ relates signiicantly over periods/ exceeding a few days [Michel and Rignot, 1999 . Michel and Rignot 1999 used the SIR-C L-band amplitude images from and 10) [ 9 October 1994 to apply a) cross-correlation method; to enhance the accuracy of the correlation method they utilized also image speckle, which was preserved only over periods of about one day for the SIR-C L-band amplitude images. The correlation matrix C between two template windows in a reference image and a time-lapsed image is quantiied by means of the cross-correlation matrix C ( m, n) calculated according to [Wu , 1995 : )

::�= * ' 1 :�=1 (r(m, n) - Lr) (s(m,n;u, v) - L8 (u, v)) C(m n) [ [ ::�=1 :: = ( n) - Lr) 112 * ::�=1 ::�=1 (s(m, n; u, v) - L8(u, v)) ) 112 � 1 r(m, 2) 6.52 ( )

= Chapter Surface motion by means of SAR 95 6.

The variables m and describe the center positions of the reference window with the n digital value r. A square template of pixels was used. Sub-pixel registration p * p precision is required to determine u and The average digital numbers of the templates master image) and slave image) are denoted by and - (u, ) ( ( -r 5 Uncertainties are involved in template matching with cross-correlation, because the cross-correlation matrix is considered to be generically multi-modal. The ambi­ guity is reduced by enforcing a smooth vector ield and a threshold in the correla­ tion matrix C m, ) Solutions were proposed with the maximum cross-correlation ( n . method, which were operationally used for cloud tracking [Wu, 1995]. Recent inves­ tigations show improvements by iteratively updating the likelihood of a correlation result with the relaxation labelling method [Wu et al., 1997]. With increasing win­ dow size the number of correlation peaks decreases. The optimum template size depends on low and surface characteristics. In areas with high shear strain only small windows can be used, resulting in a decrease of the accuracy. Wrong correla­ tion peaks may occur with small templates. Possible resons are shear distortion of the surface features and changes in the relav.ti ve positions of the features. The precisionr of the amplitudes cross-correlation increases with a greaterv ti. me interval between the scene acquisition dates for stable srface features. However, the surface features of fast lowing glacier ice may change rapidly with time intervals exceeding a few days. The motion ield of Moreno Glacier ws derived from the SIR­ 1-band amplitude images from 7 and 10 October 1994, with a repeat time interval e of 2.952 days. Amplitude cross-correlation provided a srface velocity ield close to the calving front, where ice deformation results in reduced phase coherence (igure 6-10). In this area crevasses act as the main matching features. The template window size was selected empirically. A 71 x 71 pixel window with 419 min ground range and 370 m in azimuth was used. Smaller templates showed an increased number of correlation peaks, and the velocity vectors partly deviated signiicantly from the main low direction. Mximum velocities of 3.5 m d-1, observed at the upper terminus of Moreno Glacier, correspond to a range shift of 1.75 pixels. At some sections of the upper part of the terminus velocities could not be derived by cross-correlation, because local features are less distinct than near the front.

6.1.4 Comparison of the velocity ield from INSAR and am­ plitude cross-correlation In the central part of the terminus both interferometry and cross-correlation matched well; a standard error of estimate 0.09 m d-1 in ground range results from the linear regression indicated in igure 6-11. A map of the range component of velocity vectors was derived in order to complement the interferometric map of v in regions with coherence below 0.4.

a est

=

x Chapter Surface motion by means of SAR 96 6.

+ 2km 2m/day

Figure 6-10: Radar amplitude image of Moreno Glacier terminus acquired on 7 October 1994 with ice low vectors derived by amplitude cross-correlation. Chapter Surface motion by means of SAR 97 6.

2,5

y 1, 0287x 0,0171 = 0, 9001 = 2,0 number of samp l es : 3298

1,0 �

0,5

R2

0 , 0 ------.+

0,0 0,5 1,0 1,5 2,0 2,5

Cross-Correlation (m/day )

. � u ) Figure l Comparison of cross-correlation with interferometry. Ground range 6-11:0 velocities from� the central terminus of Moreno Glacier are shown for areas with � phase coherence) > 0.5 image pair 9 and 10 October 1994, red and blue areas of u ( igure 6-4). ::

H Chapter Surface motion by means of SAR 98 6.

6.1.5 Measurement uncertainties The accuracy of interferometry and cross-correlation is diferent. The factors, which inluence the accuracy of interferometric velocities, include baseline uncertainties, phase noise and unknown range delays caused by system clock-timing, and changes of the atmospheric path length. Uncertainties in interferometrically derived velocities depend on the radar wavelength and the statistical phase noise, which is determined by the temporal coherence of the phase and the number of looks. The interferometric baselines had to be estimated from the SAR images, because of insuicient accuracy of the shuttle ephemeris data. Taking these error sources into account, the achievable accuracy of interferometry for areas with a coherence >0.4 is estimated at about 0.1 ringes. This corresponds to a velocity of 2 em d-1 , which is two orders of magnitude below the surface velocity in the center of Moreno Glacier (2 m d-1 ). Cross-correlation has lower accuracy, but the method can be applied also for low coherence areas. Cross-correlation is limited in precision by the spatial resolution of the SAR and the pixel size. The interpolation of the correlation matrix provides a matching accuracy of about 0.1 pixel corresponding to a shift of 20 em d-1 in ground range and 17.5 em d-1 in azimuth for the 1-band data. The estimated accuracy of cross-correlation is 9 cmd-1 in ground range motion resulting from comparison with interferometry by assuming statistically independent areas (igure 6-11). Michel and Rignot [1999] speciied an average diference of 14 cmd-1 in ground range between the velocity calculated with the phase correlation and the interferometry velocity.

6.2 The motion ield

The motion ield from the terminus of Moreno Glacier was previously published by Rott et al. [1998], Michel and Rignot [1999] repeated the calculation of velocities for the last 4 km towards the front. Figure 6-13 shows the motion ield from the combined interferometric and cross­ correlation analysis; slight changes in comparison to the previously published veloc­ ity ield [Rott et al., 1998] are due to by a factor of two enlarged template matching window (section 6.1 .3) and an improved topographic phase. Velocities up to 3.5 m d-1 are reached in the crevasse zone close to the equilibrium line, which is not included in the igure. The orographically right glacier stream, which drains an area of 100 km2 and is only 1.4 km wide, provides the main mss inlow to the ablation area. Comparatively low velocities ( <1.5 md-1) above proile A at the orographi­ cally right glacier side indicate an obstacle at the bed. Due to a rising glacier bed and a partly decresing glacier width the converging low accelerates the ice further glacier downward. The horizontal surface velocity along a central line ending in the Canal de los Tempanos is shown in igure 6-12. The maximum velocity close to the front is 2.4 d-1, which corresponds to 2.2 md-1 in ground range, slightly higher than the velocities obtained by Michel and Rignot [1999]. They show ground range

m Chapter Surface motion by means of SA R 6. 99

Interferety 2,5 . 10 Cross-Correlation

2,0 " 8

· 7 : ] �

6 � 5 ·· · • · • • / � --- ···•- +- --0 ·· ___ . � 4 � . • 3 . 8

. 0,5 ; 2

. 1

0,0 ; 0 0 1 2 3 4 5 6 7 8 9 10 11

. . Distnce frm Front .

Figure 6-12: Surface velocity along a central low line of Moreno Glacier obtained from interferometry (dark line) and amplitude correlation . The lower curve shows

the width of the glacier. .. ., � .. :

:: . . velocities from 2 md-1 near. the front to about 1.1 md-1 at a distance of 3.6 km . / . .. ·I above the ront along a line from' ... Canal de los Tempanos towards A-proile.L .. • ·" .... • ··o. o .. . / ·1 The motion analysis, based on cross-. . correlation, shows strong variations of ve­ - ...... �.. locity close to the calving front. This.. complex motion ield is probably caused by the bottom topography, diverting the export of mass between Canal de los Tempanos and Brazo Sur, and resulting• in a complex strain pattern.

.. (m) 6.2.1 Comparison to stake velocities

The interferometric motion at. the proiles A, B and C is 3.6 % higher than the mean annual motion mesured by DGPS (igure 6-14). Ice motion at stakes of the A-proile (section 4.3) agree well with interferometric results. At B- and C-proiles interferometrically derived velocities are higher except for 2 stakes (B02, B03) at the orographically right margin, where the intercomparison is diicult due to the high velocity gradient. Interferometric velocities are not obtained at the marginal stakes BOl and A05, as the ice low direction deviates more than 65 degrees from the radar look direction. Angles of about 60 degrees between radar look direction and ice low direction may enhance possible interferometric errors in the shear zone of stake B02 (igure 6-14). Velocities at stake sites resulting from cross-correlation are also shown in igure 6-14. The velocities are within the estimated accuracy of 0.2 m d-1 except for stake A03. Marginal stakes may be afected by the large size of the template matching window (section 6.1.3), which covers a wide range Chapter Surfa ce motion by means of SA R 6. 100

N A

[em/day]

4 km

Figure 6-13: 'lagnitude of the ice velocity vector on the terminus of Vloreno Glacier derived from SIR-C data color coded in steps of em/day. The broken lines indicate 50 the direction of ice low.

>250 >200 >150 >100 > 50 <=50 Chapter Surfa ce motion by means of SAR 101 6.

.GPS 2.5 Ointerferometry

Ocross-Correlation 2.0

1.5

1.0

0.5

0.0 I,_ .Ill ;I. .� ' l� ____ Figure 6-14: Comparison between the mean annual stake motion, the interferometric motion and the motion derived rom amplitude cross-correlation at transects A, B and C. The mean annual stake motion measured with DGPS refers to the mean values of two summer periods (1995/96, 1996/97) and the motion of one winter period 1996.

, ofJ velocities, because the stakes are located in zones of strong shear. Therefore 0 ... the: cross-correlation results at the marginal stakes in igure 6-14 are unfavorable for comparison. No cross-correlation results are obtained at C-proile. The mean diference between cross-correlation results and DGPS measured annual mean stake motion is 4.1%, almost the same as for the interferometric mean.

;

' l

1 ... N ., .. ... N ., .. 1 ' ... 0 ' 0 ... N ., 0 0 0 0 0 0 0 0 0 0 0 ... 0 0 ... 0 0 0 ! ! ! ! ! 0 0 0 0 0 0 0 0 0 0 l l l Chapter 7

On the dynamic behaviour of Moreno Glacier

The velocity ield in a glacier depends on several interrelated factors, which are: basal riction geometry of the bed hydraulic force of water shear-stress distribution internal ice deformation deformation of the bed itself glacier depth slope of the glacier surface. For Moreno Glacier information on the bed geometry and on the ice depth is available at the transects A and B. The surface slope can be derived from the DEM. The surface motion, u5 , (chapter 6) results from plastic deformation of the ice, ud, the basal sliding, ub, and the deformation of the bed, udb:

(7.1)

For basal ice at melting point, as assumed for the terminus of Moreno Glacier, • sliding is of crucial importance for the glacier's dynamic behaviour. High basal • water pressures are assumed to be responsible for the high glacier low velocities. •

• 7.1• Ve locity vriation with depth • :

The• variation of low velocity with depth can be calculated using a "parallel-sided slab"• model [Paterson, 1994] or, for a valley glacier more: realistic, by introducing a convenient shape factor. The lowlines are assumed to be parallel to the surface as the ice deforms with depth in simple shear. The z-component of the velocity is omitted taking a coordinate system as illustrated in igre 7-1 with the velocity u in direction of the x-axis. The approach of Paterson based on Glen's low-law reduces

102 Chapter On the dynamic behaviour of Moreno Glacier 103 7. the low relation to: du � A (7.2) 2 d n where is the low law exponent, and A is the low parameter depending on n temperature. For the temperate glacier ice of ooc, as observed at the terminus of Moreno Glacier troughout the year (Rott et al., 1998] , a wide range of values have been reported for A. Nye derived a value of 5.3 x 10-24 Pa-3s-1 by defor­ mation measurements of ice tunnels (Nye, 1953]. Paterson quotes values reach­ ing from 5.5 x 10-24 to 9.3 x 10-24 Pa-3s-1 , he recommends a mean value of - 1 6.8 x 10-24 Pa 3s- (Paterson, 1994] . Gudmundson derived an A 2.37 x 10-24 Pa-3s-1 by experimental and theoretical work at Unteraargletscher in the Swiss Alps (Gudmundsson, 1994] . The deformation velocity ud describes the velocity de­ crease from the glacier surface to the bed (igure 7-2), ud is derived by integrating the low relation from 0 to h (Paterson, 1994] : - - T z xz (7.3)

The shear stress on the centre-line of a valley glacier at depth ( h- ) depends on the surface slope a and the weight of ice above ( h - )

p f g ( h - ) sin a (7.4) where is the density of ice, is the acceleration of gravity, and f is a shape factor p g = of a cross section which accounts for the drag of the valley walls. A relation between the shape factor f and the shape index W (half - width/depth) for parabolic cross z z sections was given by Nye (1965]. For the deformation velocity udc at a central lowline follows: = =

2A Udc n+ (7.5) T - (fpg sin ath l z xz n+l z : The integration of the low law from the surface to the bed leads to the mean velocity at the centre-line in a vertical column of ice Umc:

T z xz

= (7.6)

The velocity Umc is described in terms of the central surface motion Usc and ice deformation idc:

= Chapter On the dynamic behaviour of Moreno Glacier 104 7.

� gh

Figure 7-1: Coordinate system for laminar low [Paterson, 1994] .

1 Umc Usc -Udc (7.7) -n+ 2 The equation considers only ice deformation and gives no information about the sliding velocity ub[Paterson, 1994]. If a low law exponent 3 is assumed, the n velocity decreases strongly with depth in the layers near the bed (igure 7-2). At the center line of the cross section B, udc is calculated using numerical values speciied in table 7.1. The ice depth h is known from seismic measurements (igure 8-5), and the surface slope is obtained from the DEM as a mean value over a distance of 5 km normal to the proile. The central surface velocity Usc normal to the proile represents the average velocity over 3 seasons measured at stake B06 (section 4.3, igure 4-15). The resulting deformation velocity udc ranges from 0.54 to 1.55 md-1 depending on the diferent low parameters A (igure 7-2). The mean deformation velocity of a cross section ud results rom the deformation velocity at a central lowline Udc according to [Nye, 1965]:

= (7.8)

where f is a factor of proportionality depending on the shape of the bed. = The possible range of me between zero (ub 0) and full sliding (ub us) is: 0.8us U Us [Rott et al., 1998]. :; m :; :

w

= = Chapter On the dynamic behaviour of Moreno Glacier 105 7.

he W f fw m deg. 900 684 3.2 0.75 0.61 1.76 2.5 Table 7.1: Numerical values of variables used for the calculation of ice deformation at B-proile. [he] maximum glacier depth in the center; [usc] central surface velocity normal to the proile.

:

700 Surface

600 � A1 =2 . 37E-24 500 A2=5.30E-24

AJ =6 . 80E-24 400

300 I I 200 �

100 l

.--- 0,0i 0,2 0,4 0, 6 0,8 1,0 1,2 1,4 1,6 1,8 i Ve locity (m/day)

!

Figure 7-2: Variation1 of horizontal velocity with depth at the center of B-proile for 3 diferent low parameters A assuming n=3; surface velocity Usc 1.76 md-1.

�

N

= ( :hapter On the dynamic behaviour of Moreno Glacier 1. 106

7. 2 Glacier sliding

�asal velocities have been measured on some glaciers in subglacial cavities and I t.nnnels or by drilling boreholes down to the glacier bed. The measured sliding values reveal partly a strong variation with time and space [Paterson, 1994] . These observations support an interpretation of the sliding mechanism as jerky movement. As sliding has not been measured in situ on any of the Patagonian glaciers, an empirical theory is used. Paterson [1994] derived a relation between basal sliding, shear stress, water pressure, and the characteristics of the glacier bed by formulating sliding law. Figure 7-3 shows a longitudinal proile along a central lowline of Moreno Glacier from the maximum glacier depth at B-proile (684 m) to the calving front at Canal de los Tempanos. The elevation of the bed at A-proile is assumed to be 200 m below sea level by extrapolating the seismic results (section 4.4, igure 4-24) to the central lowline. rom A-proile (xd 4730) to the calving front (xd 0) the shape of the bed is estimated by the analytic form Zb 7 10-6x� - 0.09xd + 77[m] , zb -200 is used between A- and B-proile. The rising glacier bed towards the calving front afects the subglacial drainage of the terminus of Moreno Glacier. Measurements of both horizontal surface velocities and water pressures at Findelen­ gletscher [Iken and Bindschadler, 1986] and at Storglaciaren [Jansson, 1995] yielded a high correlation between them. The relation between surface velocities and efec­ tive pressures for the center proile of Moreno Glacier is shown in igure 7-5. The efective pressre exerted by the bed on the base of the ice along this proile is a Pe given by [Paterson, 1994]:

(7.9)

where Pi is the ice overburden pressure and Pw is the subglacial water pressre (igure 7-4). The subglacial water pressure of Moreno Glacier is estimated from Pw * the depth d of the glacier bed below the lake level of Lago Argentino (d 180 - zb : = = )

= (7.10) = where g is the acceleration of gravity. Decreasing basal water pressures appear glacier downward towards the calving front due to a decreasing depth d. The efective from: pressure Pe results directly

Pe Pigh* (7.11) with P h* h- w d. (7.12) Pi

Where h* is the height of the glacier surface above buoyancy, and Pi the ice density.

=

=

= Chapter On the dynamic behaviour of Moreno Glacier 107 7.

500

------400 --- 300

------200 Lake Level (180 n) 100

0 .. --- -- }ff(!fl ... · -100 _ _ ... 7-·· ;f/!Bed - . --- -- . - 777!1171777 . 777ff/77 -200 /11////711/111/ll/lllllll/l!l71777 -3 00

8000 7000 6000 5000 4000 3000 2000 1000 0

B-Profile A-Profile

Figure 7-3: Longitudinal cross sectional proile of Moreno Glacier from the region of the maximum depth at B-proile to the calving front at Canal de los Tempanos; triangles mark the measured depths of the bed. The basis for estimating the bedrock proile is described in the text. Vertical exaggeration: 3.2.

-- The efective pressure, shows in igure 7-4 a decrease towards the front indicat­< Pe , ll ing increasing contributions of sliding or ploughing to the basal motion. Measure­ll ments of the efective pressure in boreholes on Ice Stream B of the West Antarc­; :: tic ice sheet have shown much lower efective pressures than predicted with 7.110 ' [Engelhardt et al., Equation implies a free connection of the subglacialll 1990] . 7.10 > ) < water to the lake, and is therefore limited to the glacier terminus [Paterson, 1994] l. For this reason the water pressure shown in igure 7-4 represents a minimum value with increasing distance to the glacier front. High water pressures are accompanied by a reduction in the coupling of ice with the bed. Shear-strain rates of the bed are assumed to depend on the efective pressure Pe [Iverson et al., 1999]. The basal shear stress is calculated according to 7.4 with 0. The shape factor f depends on the shape index W [Nye, 1965]; the values of f are between 0. 75 and 0.82 along the upper central lowline assuming a parabolic proile shape. Along the last kilometer to the front f decreases to values around 0.6. increases in general from the calving front glacier upward due to the increasing glacier depth (igure 7-6). Fluctuations are due to changes in the surface slope between 1.7 and 2.8 degrees. has been derived with a moving average of 5 km length along the lowline. The resulting range of values exceeds in some parts the range of 50 to 150 kPa reported by Paterson [Paterson, 1994] . Funk and Haeberli derived a basal shear stress between 60 and 120 kPa along a central lowline of Nordboglacier [Funk and Haeberli, 1990] . By neglecting the bed deformation ( udb) in equation 7.1 the central sliding veloc­ ity ub at the terminus is estimar ted by substracting the deformation velocity rom the interferometric surface motion (chapter 6). The deformationr velocity ud, calculated according toz 7.3, can only bexz es timated with considerable uncertainty due to the

xz

=

r

a xz a

r

r xz

xz Chapter On the dynamic behaviour of Moreno Glacier 108 1.

60

· · ·-··· ·· ····· - ...... ····· ···- .

. · · · --- - ... 50 I ce Overburden Pressure . ··- . .

_ 40 .: Water Pressure ···- ..

...... ·-··· ········· · · -· - · ···-· . · · · ····. · -·· · . . - - ... · . . · · · · . - . , . . 0 . _ 30

- ·. ·......

...... Effective Pressure ··. . . . . 20 ...... � - - _ . ------. . ------. ------__ - . ... ' 10 -- . . ... _ _ -

···� ------� 0

8000 7000 6000 5000 4000 3000 2000 1000 0

Distance

Figure 7-4: Pressure distribution along a longitudinal proile from the calving front to transect B.

ll '

0 0 ..

2, 2 J < =0 . 6848 l ll 2,1 . . . ··- ... < ·- ' J . .. .· . · .. . 2,0 .. :: (m) 1,9

·

1,8

1,7

1, 6

1, 5 - 0 5 10 15 20 25 R2 Effective Pressure (100 kPa)

. ll 0 0 Figure 7-5: Relation between0 horizontal surface velocity derived by interferometric

analysis .and the estimated efective pressure along a central lowline with a polyno­ mial best' it to the data. The correlation coeicient R2 0.6848. u 0 i ) >i ) u ll � 1 )

l

= Chapter On the dynamic behaviour of Moreno Glacier 109 7.

250

200

150

100

50

0

8000 7000 6000 5000 4000 3000 2000 1000 0

Distance (m)

Figure 7-6: Estimated basal shear stress exerted at the glacier bed along the longi­ tudinal proile. ll

, wide range of possible low parameters A (igures 7-2 and 7-7). For the calculation� of the basal sliding a low law exponent = 3 and a mean value of A= 3.8 x w-24 n l l1 Pa-3s-1 [Rott et al., 1998] is used in igure 7-8. With these assumptions high values)ll 1) ::J of ub are derived, only in the irst 1.5 km from the B-proile glacier downward) is the deformation contribution signiicant (ud 0.5 md-1). Down to the A-proilell ud ll decreases to Almost full sliding can be assumed (with a) high 0.25 md-1. (ub 'us) ll low parameter, igure 7-7) for the last 2 km to the calving front. -t l The lower sliding velocity at B-proile in igure 7-8 and possible sliding trends along a central lowline can be discussed taking into account the efective pressure. Empirical relations between the sliding velocity ub, the basal shear stress and the efective pressure Pe are given by [Paterson, 1994] :

(7.13)

where and are integers and the parameter k depends on the thermal and p q mechanical properties of the ice and on roughness of the bed. k is treated constant in the calculations for the sliding velocity ub. Proposed values are = 1 and 1 � � 3 > q p [Paterson, 1994]. For the central lowline of Moreno Glacier the product ub Pe has been correlated with the power of the basal shear stress The basal velocity p ( ) ub calculated with 7.1 and diferent low parameters A (within the range from 2.37 w-24 to 6.8 w-24 Pa-3 s-1) has been related to the basal velocity calculated with the equation 7.13. The best correlation (r= 0.86) was obtained with = 1 T p and the lower limit of low parameters A (2.37 10-24Pa-3s-1). Figure 7-9 shows the comparison of the sliding velocities; k = 16 was used in order to adjust the mean values of the 2 curves (ub = 16 ;e Ub = Us - ud) . Velocities calculated

*

rP .

X X

*

? Chapter On the dynamic behaviour of Moreno Glacier 110 7.

1, 8

1, 6

1,4

1,2

1, 0

0, 8

· 0, 6

0, 4 A=2 . 37E-24 0, 2

0, 0

8000 7000 6000 5000 4000 3000 2000 1000 0

Distance (m)

Figure 7-7: Deformation velocity ud along the longitudinal proile for 2 diferent low parameters A. . l 0 :

. ...

2,5 .) >i u 0 t ) 2,0 I 1,5 I 1,0 ·

0,5

0,0

8000 7000 6000 5000 4000 3000 2000 1000 0

Distance (m)

Figure 7-8: Interferometric surface velocity u8, deformation velocity ud and bsal velocity ub along the longitudinal proile. For the deformation velocity a mean .low l 2 0 parameter A=3 .8 w- 4 Pa-3s-l been used. ! has :

. ...

.) >i u 0 t )

L

X Chapter On the dynamic behaviour of oreno Glacier 111 7.

�

2,5

Distance (m)

Figure 7-9: Basal velocity along a central lowline of Moreno Glacier derived with 2 diferent relations.

with 7.13 do not show a general increase towards the glacier front, the maximum. 1, 0 sliding velocity is derived at a distance of about 3 km above the front (igure 7- 9). At B-proile higher sliding velocities are derived with the empirical approach J ·! 7.13. A sliding decrease towards the front would result taking into account0, 5a loweru 0 efective pressure glacier upward than calculated with 7.9. A power 1 would1 p also support this inverse trend. The calculation of the basal shear stress with ------r 0,0 a shorter slope averaging distance (< 5 km) would cause enhanced luctuations of the 8000sliding ve7000locit y in6000 equa tion5000 7. 13, however4000 30no00 changes 2000 in the1000 general sliding0 trend can be obtained.

ub

>

T

xz Chapter 8

Mass balance

The various components of the mass balance of Moreno Glacier were estimated for a multi-year period. For the lower terminus of Moreno Glacier melting may occur throughout the year (section 4.2) . The parameters determining the net mass balance Bn of a calving glacier are the total net accumulation Ba, the total net ablation due to melting, Bb, and the ice export due to calving Be:

where is the density of ice and is the direction parallel to the ice front p y [Rott et al. , 1998]. The glaciological method for determining Bn is based on mea­ srements of the net balance bn at representative points in the accumulation area Sa and in the ablation area Sb for a time period, typically one year [Paterson, 1994] . The spatial distribution of bn is inferred from the point measurements. The calving lux is calculated from the velocity normal to the front and the ice thickness h.

8.1 Ablation

The mass loss due to melting is derived directly from ield measurements. The energy available for melting processes can be calculated with an energy balance model [Hock, 1998] . At a glacier surface with a temperature of 0°C, the terms describing the energy lux at the surface-air interface are assumed to be:

(8.2)

where denotes the net radiation, is the sensible heat input, is the latent R H V heat input or loss due to condensation or to evaporationuc , Qa is the transport of heat from a vertical column from the surface to the ice depth at which vertical heat transfer is negligible, Q is the sensible heat lux supplied by rain and Q is the

112

R 111 Chapter 8. Mass balance 113 energy for melt. In terms of energy the speciic mass balance referred to a speciic bn place on the glacier can be expressed by neglecting and [Kuhn, 1981]: Qc QR

(8.3)

where tm is the length of the melt season, and is the latent heat of fusion, L 0.334 MJkg-1 . The parameterization of the transfer of sensible and latent heat to the glacier surface by means of the temperature diference between the air (Ta ) and the surface (Ts ) leads to the equation:

(8.4)

A relatively wide range of values between 0.5 and 2.7 Jm-2d-1oc-1 has been vi derived for the constant of proportionality [Kuhn, 1989]. This variability of im­ 1 1 plies other parameters than temperature which inluence the energy transfer. Other efects are for example wind speed and surface roughness. By neglecting the internal transport of heat and the inluence of rain in Qc Q R, a combination of equations 8.2, 8.3 and 8.4 is given by:

(8.5)

According to equation 8.5 the dependence of the net balance on elevation can bn be expressed in terms of 8R/8z and 8Ta/8z [Kuhn, 1989].

8.1.1 Areal extrapolation of stake measurements The mean annual ablation due to melting is estimated taking into account the mea­ sured ablation at proile-B during two summer and two winter seasons (section 4.2, table 4.10), and assuming an ice density of 900 kg/m2 . By neglecting the eleva­ p tion dependence of the net radiation (8R/8z '0) and the length of the melt season (8tm/8z ' 0), a constant altitude gradient of ablation can be assumed according to equation 8.5 with 8Ta/8z c t .. This linear relation of ablation with elevation n is in accordance with simple degree-day models, which are based on an exclusive relationship between ablation and air temperature [Hock, 1998]. The annual values (b8,1, b8,II) in table 8.1 are used as base values for the linear extrapolation of ablation in dependence of elevation. The total net ablation Bb is given by:

o s

= Chapter 8. Mass balance 114

mean altitude Center (I) Margin (II) (m a.s.l.) 500 Table 8.1: Mean annual ablation of B-proile stakes for the two roughness zones.

Center (I) Margin (II) .z (kg m-2a-1m-1 ) (kg m-2a-1 m-1 ) B ELA (670 m) 3I =11,47 3u =14,39 Table 8.2: Altitude gradients of ablation between B-proile and the ELA.

{ELA {ELA Bb = (bi + 3I (z - ZB))Sb,I(z)dz + (bu + 3u (z - zB))Sb,II(z)dz , o Zn o J J (8.6)

with 8b 3 = (8.7) az ' where ELA is the equilibrium line altitude and is the reference altitude of the ::? z8 B-proile. The altitude gradients of ablation are calculated separately for the two roughness zones along the central- smooth surface (3 I) and the crevassed- marginal surface (3 II). Haefeli suggests mass budget gradients 30 at the ELA between 10 kg m-2a-1m-1for glaciers in maritime settings and 3 kg m-2a-1m-1 for glaciers in continental settings [Haefeli, 1962]. The fact, that the derived values for 3 (table 8.2) are above Haefeli's range for 30, indicates high mass transport, or a high glacier activity index (Meier, 1961). Sb,I and Sb,II denote the areas of the two roughness zones with Sb,I + Sb,II = Sb , -the total ablation area (section 3.4) . The derived maximum ablation (b8,u ) at the glacier front is 14245 kgm-2a-1 , corresponding to an ice ablation of 15.8 ma-1 . Diferences between values in table 8.3 and the values presented in Rott et al. [1998] are due to the longer measurement period used here for calculating the base values (table 8.1); the mean annual ablation of -544*109 kg a-1 is within the pre­ viously estimated error of 20% for Bb· The representativeness of the ablation Bb (table 8.3) for a multi-year period can be checked using a degree-day method.

Bb,I Bb,l I Bb -192 -352 -544

Table 8.3: The total mean annual ablation (109 kg a-1) obtained for the two rough­ ness zones and the total ablation area, based on the ablation values and the reference elevation of B-proile.

max Chapter 8. Mass balance 115

8.1.2 Estimation of multi-year ablation with degree-days The empirical degree-day method is another possibility to compute the total ablation of a glacier for periods of a day to a year [Hock, 1998). Ablation and air-temperature measurements covering diferent periods show that degree-day ratios vary seasonally as opposed to being constant [Rango and Martinec, 1995). The degree-day method calculates the ablation at a given elevation, Bb ( z), by multiplying the degree-day ratio k with the cumulative daily mean air temperatures above 0°C, :T+(z):

(8.8)

where Sb (z) results from the hypsometric curve of the total ablation area Sb. The net ablation for Moreno Glacier is calculated according to:

(8.9)

The degree-day factors obtained from ablation stake measurements (section 4.2, table 4.12) are used as base input, though a signiicant spatial and daily variability of degree-day factors may appear [Hock, 1998). Mean values are 0.65 and 0.76 d-1c-1 for the center and marginal zones during the summer periods, and 0.29 and 0,39 d-1c-1 dring the winter periods. Summer and winter periods are assumed to last from November to March and from April to October to match the ield measurement periods (section 4.2, table 4.13). The mean monthly air temperatures measured at Moreno Bse Camp were below 7°C during the winter periods (igure 5-3). Figure 8-1 shows the total monthly ablation calculated with the mean daily air temperatures of Moreno Base. The degree-days are calculated for elevation zones of 10 m starting from 200 m a.s.l. to the mean ELA in 1170 m a.s.l., assuming a constant temperature-altitude gradient of 0.8°C/10 0 [Takeuchi et al., 1996) , (section 4.2) . According to 8.8 and 8.9 the total annual ablation from 1 November until 31 October amounts to -537* 109 kg a-1 for 1996/97 and -604* 109 kg a-1 for 1997/98. The resulting mean value of -570* 109 kg a-1 is 4.8% higher than the total mean annual ablation Bb in table 8.3. Increased ablation in 1997/98 occurred dueem to the high temperatures in 1998. The mean annual temperatures of 1998 at the nearest synopticem station of Lago Argentino reached a maximum value in the long term record since 1941 (section 5.2, igure 5-2), whereas the temperatures in 1997 were 0.5°C colder than the climatological mean. A maximum monthly ablation of -128* 109 kg a-1 results for January 1999 (igure 8-1). In order to estimate the ablation for a longer period than 3 years the daily mean temperatures of Lago Argentino station are related to the daily mean temperatures of the Base Camp using a regression given by: m Chapter 8. Mass balance 116

140

120

.._ 100 � � i 80 � l i J I + 60 � A �II � I I \ , I I r l � 20 �joL -, ..·v

Figure 8-1: The total monthly ablation of Moreno Glacier calculated with a degree­ day model using daily temperatures of Base Camp (triangles, period from December ! 19E 95 to February 1999), and of Lago Argentino station (diamonds, period from ....: Ja- nuary 1990 to April 1999). :: 0 0 ...... I 1 � I \.2 ! <:: -0.0097 Ti A + 0.8037TL A + 1.2011 (8.10) TB c = - The0 relation describes the temperatures at the Bse Camp with R2 0.84; igure 0 0 ...... � � 0 0 o o ... = ... 0 0 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ 8-1 shows the., resultin.,g estima.,ted ablation since., 1990. Whereas... the annual... ablation., ; ; ,,,, m·T·r., ., ' sums derived., ., from ., tempera., ture., regression., ., ., are .,higher ., than., .,the ablation., ., .,sums ., deriv., ed ., N N ) ) from BaseI Camp I temperaI tures (.BI b I 16I 109 kgI I ) forI the yearsI 1996/97I andI :: :: :: :: :: :: :: :: :: :: :J :J :J :J :J :J :J 1997/98, lthe meanl monthlyc ablaltion sumsl are similal r forl the overc lappingl periodl from December 1995 onward. A mean ablation value of -588 109 kg results from the ablation calculated with Lago Argentino temperatures for the period from 1 November 1990 to 31 October 1998; the estimated minimum and maximum ablation values were in 1996/97 ( -553 109 kg ) and in 1997/98 ( -620 109 kg ).

* 8.2 Calving rate

8.2.1 Introduction Calving is as a complex, not continual mechanism [Hughes, 1992], containing stochas­ 1 tic components which are not predictable. Fac* tors inluencina- g calving are for exam-

1 = * a-

- 1 -1 * a * a Chapter 8. Mass balance 117 ple the size and numbers of crevasses and precipitation. Particularly the geometry of the glacier bed at the calving terminus and the rontal glacier velocity inluence the calving. Other important factors arc the water depth, the water density and the temporal change of the water level in the surrounding of the water-ice interface [Funk and Rothlisberger, 1989] . The calving speed is deined as the ice volume discharge from the terminus divided by the cross-sectional area of the terminus. Calving speed is the sum of the frontal glacier speed and the rate of terminus change. With ice rising at the front 50 to 70 m above the water surface Moreno Glacier is grounded everywhere. Icebergs calve into the two diferent channels of Lago Argentino and partly on land. Fluctuations of the terminus are not directly related to climatic variations. A critical factor is the water depth at the terminus. Although the changes in the height of the water surface in front of the glacier are smal compared to tidewater glaciers, unevenness of the glacierbed topography at the terminus may lead to water depth changes during advances or retreats.

8.2.2 Determination of calving speed The calving speed of Moreno Glacier was calculated separately for three sections of the front (igure 8-2) . The calving speed can be described as a function of surface speed and frontal change

(8.11)

For Moreno Glacier the value of is negligeable (section 3.3) , and the surface speed was assumed to be constant for a longer time period. The ice velocity vectors were derived from the SAR motion map (section 6.2) . Due to ice deformation and tilting of pinnacles, the motion immediately at the calving front was diicult to determine. The ice motion at transects about 400m behind the front was used for calculating the ice export (igure 8-2) , assuming the ratio of the short-term to Be annual motion is 104% according to the results derived at the stake proiles (section Uc 6.2.1) , [Rott et al., 1998] . The velocity magnitude at the orographically left margin Uc at C1-proile towards Canal de los Tempanos ws estimated with 0.5 md-1 due v x: to the observation of lateral calving; the amplitude cross-correlation results at the margins were afected by the large size of the template matching window (section

6.1.3) . Uc V-X At the calving front perfect sliding was assumed (section 7.2) .

x =

v 8.2.3 Ice export The frontal section towards the Canal de los Tempanos (C1) yields 66% of the total calving rate (table 8.4) . Figures 8-3 and 8-4 show the cross sections of C1 and Be Chapter 8. Mass balance 118

Figure 8-2: SPOT image (23 Aug.1995) showing the terminus with the calving transects Cl (Canal de los Tempanos) , C2 (front of Peninsula Magellanes) and C3 (Brazo Sur) . Chapter 8. Mass balance 119

1 m/day ! ! .. I I t I i ! t .' • • l I ! t ! l' + ' I I ' I I + I ' I I \ L ' I i Distance (m)

f Mean Ice Surface i ______

·-·- __ ··- Lak� .. ��v.e l ______-· __ __ ·- ___ _ ·I ��� ·j! L:-- ...... SE .. --��-- -- ______/ � E .. _, � • .. .. • ! ...... ! .. ! Distance (m) .. � I i Figure 8-3: Cross section C1 at the calving front towards Canal de los Tempanos. Top: velocity vectors from the SAR motion map. Bottom: lake bottom topography; 0 500 1000 2000 vertical exaggeration 2.2. 1500

-- .. L

C3 proiles.; 308; the bed elevation values were extrapolated from echo sounding (section 4.6). The 25maximum0 -_ and average lake depths the Canal de los Tempanos close to are 140 m and 105 m respectively. .... For the ca:.lculation of the calving cross-sections only few measurements of the i O:) frontal surface elevatW ion were available. In 1990 Naruse et al. have measured 50 ----- 0 � heights between 55" -- �and--- 77 m above the lake level of Brazo Sr at a few spots 0 ------· - [Naruse and Aniya,0 1992].500 The mean surf1000ace heights15 00were assumed 2000 (t able 8.4) ac­ ------� cording Mto these measurements and the observation of a rising surface along the last few hundred meters towards Peninsula Magallanes. The extrapolation of the elevation model (section 2.1) from the Peninsula Magallanes revealed a mean water depth of 21 m below lake level for the bed of C2-proile. The resulting total calving lx of Moreno Glacier is 346 109kga-1 • A min­ imum calving speed is derived for C3-proile (table 8.4) as the motion vectors deviate 25° to 35° from the proile-normal (igure 8-4). Partly diferent values in table 8.4 in comparison to Rott et al. [1998] are due to re-analysis of the velocity ield (section 6.2) and slightly diferent positions of the calving proiles. The ra­ tio of calving velocity to lake depth is signiicantly higher for C1 and C3 proiles than the average ratio for freshwater calving reported by unk and Rothlisberger [unk and Haeberli, 1990] (section 9.2).

*

Uc Chapter 8. Mass balance 120

1 m/day

\ i t + Distance

� r Mean Ice Surface a e - .. .. ····· ...... ·-··· ··· ·· -·· ····- -··· ... --- -··· ··-·· ...... --- .... ···- ·-· -··· ·-··· -···· ····· .. . ·- · - ·- -·· � � ,�.��- J 1,)\, � . -- ·3 E ------� N S 50�--�-- --�- - - � Distance �

Figre 8-4: Cross section C3 at the calving front towards Brazo Sur. Top: velocity vectors rom the SAR motion map; the normal-direction of the C3 proile (indicated by the direction arrow devi500ates signiicant10ly00 from the direction1500 of ice low2000. Bottom: ) (m) lake bottom topography; vertical exaggeration 2.0.

300 - 2�0

/00 < � - ll .------0 100 e h \ : o�c � � ----�- --�------· ------0 500 1000 1500 2000 Section Width(m) h(m) c(m/d(ml) < _� C1� 2300 165 60 2.3 670 -229 C2 500 96 75 1.8 509 -22 C3 2100 119 55 1.7 420 -95 Table 8.4: Characteristics of the three sections of the calving ront of Moreno Glacier. [h] mean ice thickness, [hd estimated mean height of the surface above lake level, [c] magnitude of the mean annual velocity in the center of the proile, [uc] normal component of the mean annual velocity averaged over the proile, Be calving lx . Chapter 8. Mass balance 121

8.3 Ice discharge through B proile

The total ablation of Moreno Glacier can be estimated from the stake measurements and the calving lux only with considerable uncertainty. Another possibility for estimating the total mass loss , assuming a stationary glacier, is: n-

{8.12)

where Bb1 denotes the total net ablation between B-proile and the ELA, and vft is the mass lux through B-proile. The assumption of a stationary glacier (B 0) is supported by the observations that n * the glacier ront has been nearly stationary since 1917; * the margins of the glacier terminus were similar during the ield measurements carried out between November 1995 and March 1999 (table 4.1); satellite images and air photographs (chapter 2) showed no signiicant margin changes; * no signiicant changes of surface height or of ice velocity have been observed since the beginning of geodetic measurements on transect A {1990) and transect B {1995); * only minor inter-annual surface velocity luctuations occurred at the stake sites; * snowline positions in satellite images and oblique photography have been sim­ ilar at the end of the summers 1986 and 1994 to 1998; * the average temperature of the investigation period at the nearest synoptic station (Lago Argentino station, 5.1) has difered only by 0.4°C from the long-term� average [Rott et al., 1998] . To calculate the mass lux across a glacier section, the mean velocity Um in a cross-section is estimated from the annual srface velocity Us {chapter 7). With the mean deformation velocity of the cross section ud calculated according to 7.5 and 7.8 the mean sliding velocity follows:

{8.13)

where the mean surface velocity Us is derived by interpolating the measured surface velocities Us at the stakes with velocities derived from interferometry (chapter 6). The deformation of the bed is neglected in equation 8.13. Values are presented in 2 3 1 table 7.1 (section 7.1), a mean value of the low parameter A=3.8 * w- 4Pa- s- and the low law exponent n=3 is assumed [Rott et al., 1998]. The comparison of ub with the mean annual sliding velocity in the center of B-proile Ubc results in almost the same value (ubc = 306m/day ub = 304m/day) emphasizing the validity of the approach of unk and Haeberli [Funk and Haeberli, 1990] , who have equalized the sliding velocity at a central lowline with the sliding velocity averaged over the bed of a cross section (ub Ubc)· The decrease of the sliding velocity towards the

:}

� Chapter 8. Mass balance 122

Width h ( ) ( ) 4380 401 1.76 193 497 -786 -707

Table 8.5: Characteristics of B-proile. [h] mean ice thickness, [u ] deformation velocity, [us] normal component of the mean annual surface velocity averaged over the proile, Mt max the upper limit of the mass lx assuming full sliding, estimated Mt mass lx through B-proile.

-104 -139 -243 Table 8.6: Estimated annual ablation above B-proile. glacier margins has been assumed negligible. The mean velocity at the center is derived according to equation 7.7 as Umc m m = 0.90 * Usc; assuming the same factor of proportionality for the total cross section leads to Um 0.90 * Us and Mt 0.90Mt max · The upper limit of the mass lx = = Mt max follows for full sliding across the proile ls)(table 8.5). d (um = The mass lx through B-proile, Mt , is equal to the net mass balances above and below the section and to the mass balance residuals Br1 and Br2 , which result from errors in the mass balance terms Bi and/or from possible deviations from the equilibrium:

Mt [um (y) h (y)] dy Ba + Bb1 + Brl Bb2 + Be + Br2 (8.14) = -pi = = with:

(8.15)

Bb1 and Bb2 are the total net ablation above and below the section [Rott et al., 1998] ; y is the direction parallel to the cross section. Figure 8-5 shows the mean annual ve­ locity measured at the stakes (section 4.3) and the shape of the cross section (section 4.4) ; the ssociated characteristic values are given in table 8.5. A repeated analysis of the seismic measurements resulted in a cross section area 0.15 km2 smaller than used in Rott et al. [1998]. For the calculation of Bb1 ablation gradients 3 and 3 II (table 8.2) were taken for the spatial extrapolation. A total ablation of Bb1 -243 * 109kga-1 results with = the base values bB,I and bB,II (table 8.1). The total ablation estimate of Moreno Glacier with (Bb)B (table 8.3) as total ablation value reveals a mss loss residual Br2 -60 * 109kga-1 on the order of = 6% of the total ablation B-. The contribution of calving is estimated at 36% of B-; about 24% of B- amounts to the calving into Canal de los Tempanos (Bc1 ) . The estimation of the errors in the components of the mass loss components reveals

1 Chapter 8. Mass balance 123

' � \ ' \ \ \ \\ \' \ \ \

\ \ I \ \ -�- --\. -�- .\. \ ·, h•• • .� 4000 3000 2000 1000 0 Distance (m)

I /day . � 4000 3000 \ 2000 1000 0 \Dista nce (m) \ \

. \ \ . \ \ \ Figure 8-5: Cross section at transect\ B. Top: mean\ annual velocit\ y vectors at stakes, \ \ \ \ \ \ .-� ... \ \ . and velocity curve derived from stake\ measurements\ and interferometry. Bottom: \ \ \ \ \ \ i \ \ elevation of the ice surface and bedrock; vertica\l exaggeration 1.5. \ \ \ \ \ i . \ Mt Bb1 Ba Be Bb Br2 -707 -243 950 -346 -544 -60 Table 8.7: Mass balance components and mass luxes on Moreno Glacier in (109kga-1 ). comparatively high uncertainties of about 20% for Bb due to errors in the ablation gradients and the ablation base values. The estimated errors for Mt and Be are 10% including uncertainties in the basal sliding, the surface velocity and errors in the bed topography [Rott et al., 1998].

8.4 Accumulation

Due to the size and inaccessibility of the accumulation area, it has not been possible to determine Ba directly. An accumulation value of 1500 mm at one point near the ice divide of Moreno Glacier has been inferred from sequences of deuterium by analyzing an ice core [Aristarain and Delmas, 1993]. Ba is estimated assuming Moreno Glacier to be close to an equilibrium state, and that the ield measurements are representative for a multi-years period. For perfect equilibrium (Bn 0) the accumulation is identical to the mass losses, 1 (Bn Ba+B- 0). A total mass loss B- Mt + Bb1 -950* 109kga- (table 8.7) corresponds to a speciic annual net accumulation, averaged over the accumulation

=

= = = = Chapter 8. Mass balance 124 area, of 5250 ± 660 kg m-2. Considering that part of the precipitation in the accumulation area is lost due to melting, runof and wind drift into diferent drainage areas, the average annual precipitation can be signiicantly higher. This assessment is in agreement with the results of Escobar et al., who have derived average annual precipitation values on the order of 7000 to 8000 kg m-2 by estimating large-scale water balance of Patagonia [Escobar et al., 1992]. Chapter 9

Comparison with other freshwater calving glaciers

Whereas Moreno Glacier has attained an almost stable state during this century, the majority of the outlet glaciers of the Patagonian Iceields have shown pronounced retreat since the last few decades. According to an estimation of Aniya [1996] , the SPI area shrunk by about 4% from 1944/45 until 1986. In this chapter the behaviour of Moreno Glacier is compared with glaciers of the SPI and in Greenland. Investigations of freshwater calving into deep lakes are sparse; the most comprehensive ield data come from Nordboglacier in South-West Greenland [Funk and Haeberli, 1990] .

9.1 Characteristics of SPI glaciers

Aniya et al. mapped a total of 48 outlet glaciers using stereo pairs of aerial pho­ tographs and a Landsat TM mosaic of the SPI [Aniya et al., 1996] . All of these glaciers except Frias- (50°45'S, 75°05'W) and Bravo Glacier (48°38'S, 73°10'W) calve into Paciic fj ords or proglacial lakes. The largest glaciers, covering areas of about 1000 km2 are Briggen-, O'Higgins-, Viedma- and Upsala Glacier [Aniya et al., 1996] , [Skvarca ct al., 1995]. In 1986 the total area of the iceields and outlet glaciers has been about 11260 km2. Small valley and cirque glaciers at the margins of the SPI cover about 1510 km2 . The altitude of the transient snow line of SPI glaciers, as deduced by Aniya from analysis of Landsat TM images acquired on 14 Jan­ uary 1986, ranges typically from 650 m a.s.l. at the southern end of the SPI (Bal­ maceda Glacier, 51°23'8) to 1500 m a.s.l. at the northeastern side (Bravo Glacier, 48°38') [Aniya et al., 1996]. The average accumulation-area ratio (AAR) of the out­ let glaciers is estimated at 0. 72. Exceptional high AAR's above 0.80 have been inferred for heavily calving glaciers oriented towards Paciic fj ords. One of the most conspicuous retreats occurred at O'Higgins Glacier; from 1945 until 1995 the glacier lost and area of 65 km2 and the frontal retreat was about 14.5 km [Casassa et al., 1997]. The frontal area of Upsala Glacier decreased by

125 Chapter 9. Comparison with other freshwater calving glaciers 126 at least 29.7 km2 between 1945 and 1996 [Aniya et al., 1992], [Rott et al., 1997] . In contrast to this general trend, special surface and bedrock topography or exceptional calving dynamics caused some glaciers to behave diferently. Examples are Briggen Glacier (or Glaciar Pio XI), which is located at a similar latitude as O'Higgins Glacier, but calves towards the west. Briggen Glacier is regarded to have reached its Neoglacial maximum [Warren et al., 1997] . The glacier advanced since 1945, blocking a Paciic fj ord, and its terminus separated into a northern and a southern tongue. After an advance of about 9.5 km between 1945 and 1982, the southern tidewater calving tongue of Briggen Glacier retreated between 1 and 2 km until 1989 [Aniya et al., 1996], [Warren et al., 1997] . Readvance of partly more than 2 km occurred in the early 1990s until 1993. In contrast, the northern tongue of Briggen Glacier calving into Lago Greve continued to advance uninterruptedly (but at diferent rates) by more than 2 km between 1976 and 1994. Briggen Glacier gained a total area of 132 km2 within the period 1945 to 1994 [Warren et al., 1997]. Only few ablation data of SPI glaciers are available. Besides extensive measure­ ments on Moreno Glacier (section 4.2) ice-thickness changes have been measured in the ablation areas of Tyndall and Upsala Glacier. Naruse summarized these mea­ surements [Naruse et al., 1995a] , pointing out the stability of Moreno Glacier and a considerable thinning rate of 11 ma-1 measured on Upsala Glacier within a 3 years period from 1991 until 1993.

9.1.1 Ameghino Glacier In March 1999 ield activities have been initiated on Ameghino Glacier to study the diferent behavior of two adj acent calving glaciers. In contrast to the stable terminus of Moreno Glacier, Ameghino Glacier retreated signiicantly during the 20th century, though the glaciers are only about 8 km apart. Ameghino Glacier is also an eastern outlet with a length of about 15 km and an area of 48 km2 , the northward oriented accumulation area adjoins Moreno Glacier (igure 9-1). Aniya et al. [1996] assumed an ELA of 1000 m a.s.l. for Ameghino Glacier. This ELA correspond to an AAR of 39%, the ablation area of Ameghino Glacier would amount to 29 km2• The terminus is between 1 and 3 km wide. At the orographically left side there are some small lanking glaciers rom independent snowields. Warren [1994] has documented the retreat from 1928 until 1993 using pho­ tographs and satellite imagery. The irst known oblique photograph of Ameghino Glacier was taken in 1928 [Agostini, 1945]; the photograph shows the terminus form­ ing a little frontal lake (unoicially named 'Laguna Ameghino'). An aerial photo­ graph, taken by the Instituto Geogniico Militar (IGM) of Argentina (9-3) in March 1970, shows only minor frontal changes since February 1967 (compare Warren [1994] , igures 5 and 6). These images show small icebergs calving into a small proglacial lake at terminal and also lateral sections. The main part of the lake extended along the orographically left margin of the glacier front (igure 9-3). The distance be­ tween the central front and the frontal moraine, which was formed during the last Chapter 9. Comparison with other freshwater calving glaciers 127

0 Skm

-Rock

Figure 9-1: Map of Ameghino and Moreno glaciers. LA: Laguna Ameghino, CT: Canal de los Tempanos, BR: Brazo Sur. Chapter 9. Comparison with other freshwater calving glaciers 128

advanced position probably between 1870 and 1880 [Nichols and Miller, 1952], was only about 100 m. A trimline indicates this maximum of the last century. The srface lowering caused a decrease of the surface slope from about 8° to partly less than 3° in 1967 [vVarren, 1994]. A rapid increase of calving resulted in a retreat of the main front of about 3 km between 1970 and 1976. The retreat slowed down until 1993. Figure 9-3 shows the retreat between 1970 and 1995. Between 1970 and 1995 Ameghino Glacier lost a frontal area of 4.23 km2 , the rontal retreat was 4.2 km. Warren measured the height of the front at several points between 30 and 40 min December 1993 [Warren, 1994] . Depth sounding in Lago Ameghino has been carried out on 18 March 1999; the maximum depth measured close to the ice front is 160 m. The tracks of the boat, measured with DGPS, and the analysis of an ERS2 image acquired on 6 March 1998 reveal no signiicant changes of the frontal position since 1995. 7 ablation stakes were established on the terminus of Ameghino Glacier in March 1999; repeated ield measurements to derive annual ablation and stake motion will be carried out in March 2000. A preliminary analysis of the surface velocity at the terminus of Ameghino Glacier is derived from SIR-C data using the same methods as described in chapter 6. The maximum surface velocity shown in the map 9-4 is slightly less than 1.2 md-1, which is signiicantly lower than velocities of Moreno Glacier. The velocities at the last 2 km were derived by amplitude cross-correlation, these velocities may be afected by the size of the template matching window covering zones of strong shear. With an estimated elevation of the calving front of 35 m above the lake level (lake level = 203 m a.s.l.), the calving cross section is derived using depth sounding results (igure 9-5). Ice motion values at a transect about 250 m behind the ront are used for calculating the mass lux Be, assuming that the ratio of the short-term to annual motion is similar as at Moreno Glacier (�104%). The frontal change is neglected (section 8.2.2) . Values characterizing Ameghino calving include a considerable uncertainty (table 9.1), as annual motion data will be available only after re-measurements of stake positions in March 2000. The total mass loss B- of Ameghino Glacier is estimated assuming for the total annual ablation amount Bb a linear ablation decrease with height from the glacier front to an elevation of 1000 m a.s.l., and an annual ablation at the front of 13 m as base value. The resulting Bb is -108 109kg -1• The ratio of calving lux to total mass loss Be/ B- 0.16 is clearly below the value derived for Moreno Glacier ( 0.36). A possible reason for the accelerated retreat of Ameghino Glacier since 1970 (igre 9-3) might have been a loating glacier tongue. Surface lowering might have caused a reduction of the efective pressure Pe (section 7.2) to a point, where the height of the glacier surface equals buoyancy (h* 0, formula 7.12). Recent inves­ tigations of the frontal position using satellite imagery reveal only minor changes duringx the nineties.

* a �

�

= Chapter 9. Compaison with other freshwater calving glaciers 129

2500

1500

1000 -� + � 500 +

0 - -- -r ---· ------r--·------.---· ··------, r ------· 0 10 20 30 40 50

Area

Figure 9-2: Area altitude distribution of Ameghino Glacier.

T r

:

l

<

t f ::

l

) -f -- -- ... ll

- , (m2 ) Width(m) h(m) d(m) 880 110 75 288 227 -20 Table 9.1: Characteristics of the calving front of Ameghino Glacier. [h] mean ice thickness, [d] mean water depth, [Vc] magnitude of the mean annual velocity in the center of the proile, [uc] normal component of the mean annual velocity averaged over the proile, Be calving lx . The velocity values are preliminary results; a more detailed analysis is expected after the planned re-measrements of stakes in March 2000. Chapter Comparison with other freshwater calving glaciers 130 9.

Figure 9-3: The front of Ameghino Glacier imaged by an aerial photograph (top) , Landsat TM (center) , and SPOT (bottom) . Chapter Compaison with other freshwater calving glaciers 131 9.

N A

[em/day] 90 2km 60 30 <=30

Figure 9-4: Magnitude of the ice velocity vector on the terminus of Ameghino Glacier derived from SIR-C data in steps of 30 em/ day.

>

>

>

. ��- ..

1 m/day

Figure 9-5: Cross section at the calving front of Ameghino Glacier. Top: velocity vectors from the SAR motion map. Bottom: lke bottom topography; vertical exaggeration 1.2. 800 600 400 200 0

250 - Mean IGe Surface -:·j ... 200 ' ' 150 .. 5 n 100 : ... ' 50 : . ...:: 0 . NNE ssw � J o: > 800 600 400 200 0 l Distance (m) u Chapter Comparison with other freshwater calving glaciers 132 9.

Width(m) h(m) d(m) 1200 175 90 244 230 -33 Table 9.2: Characteristics of the calving front of Nordbo Glacier. [h] mean glacier thickness at the front, [d] mean water depth, [Vc] magnitude of the mean velocity in the center of the proile, [uc] normal component of the mean annual velocity averaged over the proile, Be calving lx [Funk and Haeberli, 1990] .

9.2 reshwater calving glaciers in other regions

Investigations of freshwater calving glaciers are sparse. Funk et al. have investigated the efects of a planned reservoir on the calving of Unteraargletscher in Switzer­ land [unk and Rothlisberger, 1989] , [unk and Haeberli, 1990]. Data exist from the Alaskan Portage Glacier (60°45'N), which had a similar mean proglacial lake depth as Moreno Glacier. A calving velocity 220 ma-1 was derived with surface velocity measurements during one summer period in 1972, taking into account the average rate of the frontal retreat for the period 1960-72 [Funk and Rothlisberger, 1989] . Recently Portage Glacier has retreated into shallow water. In order to investigate calving into deep freshwater, ield work was carried out on Nordboglacier.

9.2.1 Nordboglacier Nordboglacier is a southern outlet glacier of the Eqalorutsit Kangigdlit Sermiat (EKS) Iceield in South-West Greenland (61°27'N, 45°24'W) . In 1989 and 1990 Funk et al. carried out 3 ield campaigns to complement previous ield data from the Greenland Geological Survey [Funk and Haeberli, 1990]. Field work focused on the following main topics: mesurements of 5 glacier cross-sections with radio-echosounding, Uc repeated measurements of the calving front by means of terrestrial photogram­ metry, = observation of the calving front dring the summer 1989 by means of 2 auto- matic cameras, repeated measurements of the surface velocity, measurement of the water- temperatre in Nordboso lake. Aerial photographs have shown a rontal advance between 1952 and 1982. Al­ though a lowering of the surface was observed from 1982 to 1987, the front continued to advance. From August 1987 until August 1990 an over-all retreat of 37 m was observed with photogrammetric methods. The calculated mean calving lx of Nord­ boglacier (table 9.2) is signiicantly lower than the calving lx of Moreno Glacier. In analogy to the linear calving relation for tidewater glaciers [Brown and Post, 1982], an almost linear increase of the calving speed with the lake depth for fresh-water calving glaciers has been formulated [unk and Rothlisberger, 1989] , (igure 9-6). Formula• 9.1 is based on results from Nordboglacier, on investigations of glaciers

•

•

•

•

Uc Chapter 9. Comparison with other freshwater calving glaciers 133 in the Italian and Swiss Alps calving into more shallow water, and on data from Portage Glacier [unk and Haeberli, 1990] .

2.5 d, (9.1) where and d denote the calving velocity and the water depth averaged over the calving cross-section. There is a diference of a factor of proportionality of 11 between tidewater glaciers ( 27 -l) and glaciers calving into freshwater. Enhanced water circulation at the front of tidal glaciers may explain high calving rates. Melting causes density diferences, which are about 200 times higher in sea water than in fresh water at 4°C [Funk and Rothlisberger, 1989]. Discrepancies between the relation derived by Funk et al. and high calving rates observed at Moreno Glacier may be due to other important factors inluencing calv­ ing (igure 9-6). The water circulation at the ice-water interface performs the major­ ity of the heat exchange at the calving front. The water circulation at the Moreno front should be more intense than at Nordbo Glacier due to higher water tempera­ tures in Lago Argentino than in NordUcbo so. Depth* proiles of water in Nordboso show temperatres below 3°C in spring and autumn [Funk and Haeberli, 1990], whereas measuremUc ents on 1 April 1997 in Lago =Ar gentino reveal water temperatures between 7.5°C (at 10 m depth) and 8°C (at 2mdepth) . The Lago Argentino temperatures close to Moreno Gla�cier area in accordance with Hauthal's water temperatures of 8°C in March 1900 [Hauthal, 1904] and with the measurements of Warren in December 1993, who has found a cooling rom 7.6°C at the surface to 5.5°C at a depth of 100 m [Warren, 1994]. The calving characteristics of Ameghino Glacier, which are similar to Portage Glacier (igure 9-6), it to the factor 2.5 in equation 9.1. 2mwat er temperatures of 2.6°C have been measured at the surface of Laguna Ameghino close to the glacier front and in the middle of the lake on 18 March 1999. The observation of a thin ice layer at the top of the lake on 17 March 1999 indicates a stable stratiication. Freshwater calving characteristics seem to be consistent with the linear formula 9.1 for glaciers calving in lakes with temperatures below 4°C troughout the year. Chapter 9. Comparison with other freshwater calving glaciers 134

! I j i �I I Nordbo ---- .--- � � no - Po ta;ge� 1I -- - . ------: ------� ----- __.. -- --1 --

Water Depth (m) 700 ,

Figre 9-6:600 Rela tion between calving velocity and lake depth for fresh-water calving glaciers reported by Funk and Rothlisberger [1989 and for Ameghino and Moreno 500 ) Glacier; C1 - Canal de los Tempanos, C3 - Brazo Sur. C 3 400

- 300 ! •

2 0 0

1 0 0 --

-- 0

0 20 40 60 80 100 120 Chapter 10

Summary and conclusion

Although the Southern Patagonian Iceield (SPI) has been subject to major ice retreat for about ive decades, Moreno Glacier has been stable. This contrasting behaviour relects the complex nature of calving glaciers with a diferent response to climate. Detailed studies on mass luxes and low dynamics were carried out for Moreno Glacier, resulting in the most comprehensive data set of any glaciers in Patagonia. The applied synergism of remote sensing and extensive ield measurements is a use­ ful approach to investigate large, inaccessible glaciers.

Field mesrements The activities carried out during seven ield campaigns on Moreno Glacier be­ tween November 1995 and March 1999 are described in chapter 4. A network consist­ ing of two transverse (A-, B-proiles) and one longitudinal proiles (C-proile) with a total number of up to 19 ablation stakes was established at the terminus of Moreno Glacier in November 1995. The upper B-proile was spanning the whole glacier width of 4.4 kilometers, its distance upstream from the calving glacier front was 7.5 kilometers. The stakes were redrilled at the original positions during each ield campaign to obtain comparable ablation values for diferent years, and to avoid the propagation of the stakes into crevassed zones. The stake measurements of ablation revealed distinct diferences between the central-, smooth parts of the glacier and the crevasse and boundary zones. Melting was also observed during the winter pe­ riods, as temperatures above 0°C may occur at the glacier terminus throughout the year. The average daily ice ablation at B-proile during the periods from November until March of the years 1995/96 and 1996/97 was 5.0 cmd-1 for the central proile and 6.1 cmd-1 for the marginal proile. Degree-day factors were calculated using temperatures measured close to the terminus at Moreno Base Station. The mean degree-day factors were 0.65 cmrc day and 0. 76 cm;oc day during the summer months from November to March averaged over central and marginal zones. The surface velocities of the stakes were measured by means of GPS. The maxi­ mum stake velocity was 2.39 md-1 in the average of the period from 25 November

135 Chapter Summary and conclusion 10. 136

1996 to 26 March 1997 at the longitudinal C-proile. Ice thicknesses of Moreno Glacier were determined by using the seismic-relection method along A- and B-proiles in November 1996. 100 kg of explosive charges were placed at 47 points in little crevasses or in water puddles. The P-wave velocity is assumed to amount to 3600 for the temperate ice of Moreno Glacier. The relected seismic waves at A-proile revealed a steady lowering of the glacier bed from the margin towards the centerline; 1650 m from the glacier margin an ice depth of 540 m was obtained. The measurements along B-proile revealed a subglacial trough with an approximately parabolic shape; 684 m was the maximum ice depth at a proile location slightly shifted to the north of the centre of the glacier. The maximum ice thicknesses at the two seismic proiles indicated a glacier bed going down to about 200 m below sea level and rising for the last 5 km towards the calving terminus. Bathymetry was carried out close to the calving front. The deepest signals were received at 164 m in Canal de los Tempanos and in 110 m at Brazo Sur. The ield activities included the measurements of several geodetic points and lines reference for satellite images. Only minor changes of parts of the orographically 1 right glacier margin were derivedms be-tween March 1996 and April 1998 from repeated GPS lines. An automatic climate station was installed in November 1995 at the shore of Brazo Sur at a distance of 360 m from the orographically right terminus of Moreno Glacier. lVIeteorological records rom 16 November 1995 until March 1999 of six diferent sensors were analyzed for this study.

egionl climate Patagonian glaciers are inluenced by Antarctic and mid-latitude atmospheric circulation patterns. Very persistent westerly winds, wet air rom the Paciic and the high frequency of subpolar depressions [Warren and Sugden, 1993] characterize the wet climate of the Chilean coastal zones and the main parts of the SPI. The foasrmida ble orographic barrier of the Andes result in sharp precipitation gradients in the region around the eastern slopes of the Patagonian Andes. The mean annual precipitation sums at Bahia Felix of 4000 and at Lago Argentino of 210 are examples for the large gradients. Abundant precipitation and high melt rates a.t the glacier termini result in steep mss balance gradients. Homogenous long-term climate records are sparse in Patagonia. A warming trend since about 1940 in the region east of the Iceields was reported by several authors [Hofmann et al., 1997] , [Rosenbluth et al., 1995] , [Ibarzabal et al., 1996] . The mean annual temperatures measured at Moreno Base Station were 6.9°C in 1996, 5.9°C in 1997 and 7.1°C in 1998; the amplitude between the warmest and the coldest months of 11.3°C in January 1999 and 0.7°C in July 1997 show the maritime inluence. The wind mesurements at Moreno Base Station show persistent western and southwestern directions in accordance with the orientation of the glacier terminus.

1 mma- 1 mma- Chapter Summary and conclusion 137 10.

Though the maximum wind speed at Moreno Base Station of 31.4 ms-1 occurred in winter on 18 July 1996, the monthly average wind speed is comparatively low during the winter in June or July. The highest monthly average was 5.9 ms-1 in January 1999.

Glacier motion Methods and results of ice motion analysis by means of repeat pass SAR images are described in chapter 6. A processing scheme, which combines SAR interferometry and amplitude cross-correlation, was applied on SIR-C data acquired on 7, 9 and 10 October 1994. The resulting velocity map, covering the main parts of the glacier terminus, represents the surface velocity over the period between the acquisition of the SAR images. The motion of the terminus of Moreno Glacier is characterized by: high surface velocities. At the central parts of the terminus the velocities are higher than 1.5 md-1. Velocities up to 3.5 m d-1 are reached at the orographically right glacier stream close to the equilibrium line. The mean calving velocity towards the main calving section at Canal de los Tempanos amounts to about 670 ma-1; strong variations of velocity close to the calving front. This complex motion ield is probably caused by the bottom topography, diverting the export of mass between Canal de los Tempanos and Brazo Sur, and resulting in a complex strain pattern; little temporal variations. The average diference in ice motion measured at all stakes between the two summer periods November 1995 to March 1996 and November 1996 to March 1997 and the winter period March 1996 to November 1996 was 7%. The mean diference between velocities derived from SAR images and the measured annual mean stake motion is about 4%; high sliding velocities, which have been estimated for various assumptions (chapter 7). The basal water pressure is assumed to rise glacier upward from about 10 • 105 Pa at the calving front to values higher than 40 105 Pa at a distance of 5 * * km to the front.

Mss balance For• the mass balance of Moreno Glacier a steady state of the glacier is assumed. This assumption is supported by ield observations, various satellite images and aerial photographs, which reveal similar margins of the glacier and minor luctuations of the calving terminus since 1917. No signiicant surface changes and changes of the snowl• ine positions were analysed. For the mean equilibrium line an altitude (ELA) of 1170 m a.s.l. was derived. The motion ield, ablation measurements at stake sites, and the measurements of glacier cross sections were the basis for estimating mass balance components. The total ablation of Moreno Glacier is calculated, assuming a stationary glacier, (i) as the •sum of the mass lx through B-proile and the estimated total ablation between the B-proile and the ELA, and (ii) from the estimated total ablation due to melting Chapter Summary and conclusion 138 10. and the total ice export due to calving (chapter 8). (ii) For estimating the ablation an areal extrapolation method with stake mea­ surements as base values was applied. A mean annual ablation of -544* 109 kg a-1 and a possible error of 20% was estimated. The representativeness of this ablation was checked for a multi-year period using a degree-day method. The calving speed ie was calculated separately for tree sections of the front by neglecting the small frontal changes. The total calving lux of Moreno Glacier is estimated at 346 * 109kga-1• The frontal section towards the Canal de los Tempanos yields about 66% of the total calving lux. (i) To calculate the mass lx across B-proile, the mean velocity um in the cross-section was estimated from the mean annual surface velocity Us derived from stake velocities and SAR interferometry. With a mean velocity um = 0.90 * Us, with Us 497ma-1 and the cross-sectional area of 1.76 km2 a total mass lux through B-proile of -707 * 109 kga-1 was derived. A total ablation of Moreno Glacier of -950 * 109kga-1 results as the sum of the lx through B-proile and the estimated total ablation between B-proile and the ELA. Although the errors in the components of the mass loss components were assumed to be comparatively high for the areal extrapolation of ablation, the diferences between the methods (i, ii) is -60 * 109kga-1 on the order of 6% of the total loss. In case of perfect equilibrium, a speciic annual net accumulation = 5250 660 ± kg m-2 is deduced. This is among the highest accumulation values worldwide. Tak­ ing into account that not all of the snow and rain deposited in the accumulation area remains stored in the irn, the average annual precipitation is certainly higher than 5250 kg m-2 • This is in agreement with the estimations of large-scale water balances of Patagonia [Escobar et al., 1992] and points out that the annual accumulation of 1500 kg m-2 , inferred from a single ice core near the ice divide of Moreno Glacier, is not representative [Aristarain and Delmas, 1993].

A high= ratio of calving lux to net accumulation Be/ Ba 0.36 points out the importance of calving for the mass balance of Moreno Glacier. According to the magnitude of the mass lxes and the rapidity of the icelow, the climatic response of Moreno Glacier should be comparatively fast, as for other Patag­ onian Glaciers. The stability of Moreno Glacier throughout this century is probably an efect of the particular calving conditions and the surface topography. The glacier surface is comparatively steep near the equilibrium line. Elevation changes of the equilibrium line due to temperature changes cause only small changes of the AAR. For this reason, and due to the fact that the mean altitude of the accumulation area is 700 m higher than the equilibrium line, changes in precipitation should be at least of similar importance for the mass balance than changes of temperature. Concluding rom the estimated mass balance components and mass lxes, and considering the special surface and bedrock topography, a rapid retreat of Moreno Glacier in the near future is not expected. � Bibliography

[Agostini, 1945] Agostini, A. (1945). Andes Patagonicos. Buenos Aires: Guillemo kraft. second edition.

[Aniya et al., 1992] Aniya, M., Naruse, R., Shizukuishi, M., Skvarca, P. , and Casassa, G. (1992). Monitoring recent glacier variations in the Southern Patag­ onia Iceield utilizing remote sensing data. Int Archives of Photogrammetry and Remote Sensing, 29(B7):87-94.

[Aniya et al., 1996] Aniya, M., Sato, H., Naruse, R., Skvarca, P., and Casassa, G. (1996). The use of satellite and airborne imagery to inventory outlet glaciers of the Southern Patagonia Iceield, South America. Photogammetric Engineering and Remote Sensing, 62(12):1361-1369.

[Aniya and Skvarca, 1992] Aniya, M. and Skvarca, P. (1992). Characteristics and variations of Upsala and Moreno glaciers, southern Patagonia. Bulletin of Glacier Research, (10):39-53. Japanese Society of Snow and Ice.

[Aristarain and Delms, 1993] Aristarain, A. and Delmas, R. (1993). Firn-core study from the southern Patagonia Ice Cap, South America. Jounal of Glaciol­ ogy, 39(132):249-254.

[Braithwaite, 1977] Braithwaite, R. (1977). Air Te mpeature and Glacier Ablation - a Paametic Appoach. PhD thesis, McGill University.

[Braithwaite, 1995] Braithwaite, R. (1995). Positive degree-day factors for abla­ tion on the Greenland ice sheet studied by energy-balance modelling. Jounal of Glaciology, 41(137):153-160.

[Braithwaite and Olesen, 1993] Braithwaite, R. and Olesen, 0. (1993). Seasonal variation of ice ablation at the margin of the Greenland ice sheet and its sensitivity to climate change, Qamanarssup Sermia, West Greenland. Jounal of Glaciology, 39(132):267-274.

[Brown and Post, 1982] Brown, B.C., M. and Post, A. (1982). Calving speed of M. Alaska tidewater glaciers, with application to Columbia Glacier. S. Geological . Survey Professional Paper, pages 1258-C.

139 140 BIBLIOGAPY

[Burgos et al. , 1991] Burgos, J., Ponce, H., and Molion, L. (1991). Climate change predictions for South America. Climatic Change, 18:223-239.

[Casssa et al., 1997] Casassa, G., Brecher, H., Rivera, A., and Aniya, M. (1997). A century long recession record of Glaciar o' Higgins, Chilean Patagonia. Annals of Glaciology, 24:106-110.

[Clarke and Echelmeyer, 1996] Clarke, T. and Echelmeyer, K. (1996). Seismic­ relection evidence for a deep subglacial trough beneath Jakobshavns Isbrae, West Greenland. Jounal of Glaciology, 43(141):219-232.

[Deichmann et al., ress] Deichmann, N., Ansorge, J., Scherbaum, F., Aschwanden, A., Bernardi, F., and Gudmundsson, G. (in press). Evidence for deep icequakes in an alpine glacier. submitted to Annals of Glaciology, 31. proceedings of the International Symposium on the Veriication of Cryospheric Models, 16-20 August 1999 in Zrich, Switzerland.

[Endlicher and Santana, 1988] Endlicher, W. and Santana, A. (1988). El clima del sur de la Patagonia y sus aspectos ecologicos. un siglo de mediciones climatologicas en Punta Arenas. Anales Instituto de la Patagonia, 18:57-86.

[Engelhardt et al. , 1990] Engelhardt, H., Humphrey, N., Kamb, B., and Fahnestock, M. (1990). Physical conditions at the base of a fast moving Antarctic ice stream. Science, 248:57-59.

[Escobar et al., 1992] Escobar, F., Vidal, F., Garin, C., and Naruse, R. (1992). Water balance in the Patagonia Iceield. Glaciological Researches in Patagonia, Inst. of Low Te mpeature Science, Hokkaido Univ., Japan, pages 109-120.

[et al., 1995] et al., E. S. (1995). Overview of results of spaceborne imaging radar­ C, X-band synthetic aperture radar (SIR-C/X-SAR) . IEEE rans. Geosc. Rem. Sens., 33:817-828.

[unk and Haeberli, 1990] Funk, M. and Haeberli, W. (1990). Gletscher Kalbungs­ geschwindigkeit im Susswasser; eine Studie am Nordbogletscher im Johan Dahl Land, Sud- west Gronland. VAW Studie im Autrag der Kraftwerke Oberhasli AG 20.8, Versuchsanstalt fir Wasserbau, Hydrologie und Glaziologie der Eidgenossis­ chen Technischen Hochschule Zurich, Kraftwerke Oberhasli AG, Innertkirchen.

[unk and Rothlisberger, 1989] Funk, M. and Rothlisberger, H. (1989). Forecasting the efects of a planned reservoir which will partially lood the tongue of Unter­ aargletscher in Switzerland. Annals of Glaciology pages 76-81. 13, [Geudtner, 1995] Geudtner, D. (1995). Die interferometrische Verarbeitung von SAR Daten. Forschungsbericht 95-28, Deutsche Forschungsanstalt fur Luft- und Raumfahrt e.V. 141 BIBLIOGAPHY

[Goldstein et al., 1988] Goldstein, R., Zebker, H., and Werner, C. (1988). Satel­ lite radar interferometry: two dimensional phase unwrapping. Radio Science, 23(4):713-720.

[Gudmundsson, 1994] Gudmundsson, H. (1994). Converging glacier low - a case study: the Unteraarglacier. Mitteilungen der Ve rsuchsanstalt fur Wasserbau, Hydrologie und Glaziologie der Eidgenossischen Technischen Hochschule (E TH) Zurich, 131:120.

[Haefeli, 1962] Haefeli, R. (1962). The ablation gradient and the retreat of a glacier tongue. Symposium of Obergurgl, Intenational Association of Hydrological Sci­ ences Publication, 58:49-59.

[Hauthal, 1904] Hauthal, R. (1904). Gletscherbilder aus der Argentinischen Cordillere. Zeitschrift des deutschen und osterreichischen Alpenvereins, 35:30- 56.

[Hauthal, 1911] Hauthal, R. (1911). Der Bismarck-Gletscher, ein vorrickender Gletscher in der Patagonischen Cordillere. Zeitschrit fur Gletscherkunde, 5:133- 143.

[Heinsheimer, 1958] Heinsheimer, G. (1958). Zur Hydrologie und Glaziologie des Lago Argentino und Ventisquero Moreno. Zeitschrit fur Gletscherkunde und Glazialgeologie, 4(1-2):61-72.

[Heucke, ress] Heucke, E. (in press). A light portable steam-driven ice drill suitable for drilling holes in ice and irn. Proceedings to the workshop on methods of mass balance measurements and modelling, Tarfala, Sweden, August 1 gg8.

[Hock, 1998] Hock, R. (1998). Modelling of glacier melt and discharge. Zurcher Geogaphische Schriten, 70:126.

[Hofmann et al., 1997] Hofmann, J., Nunez, S., and Vargas, W. (1997). Temper­ ature, humidity and precipitation variations in Argentina and the adj acent sub­ antarctic region during the present century. Meteorologische Zeitschrit, N. F. , 6:3-11.

[Hoinkes and Steinacker, 1975] Hoinkes, H. and Steinacker, R. (1975). Hydromete­ orological implications of the mass balance of Hintereisferner, 1952-53 to 1968-69. IA HS-AISH Publ., 104:144-149. Proceedings of the Snow and Ice- Symposium in Moscow, August 1971.

[Hughes, 1992] Hughes, T. (1992). Theoretical calving rates rom glaciers along ice walls grounded in water of variable depths. J. Glaciol. 38{128}, pages 282-294.

[Ibarzabal et al., 1996] lbarzabal, T., Hofmann, J., and Naruse, R. (1996). Recent climate changes in Southern Patagonia. Bulletin of Glacier Research, 14:29-36. 142 BIBLIOGAPHY

[Iken and Bindschadler, 1986] Iken, A. and Bindschadler, R. (1986). Combined measurements of subglacial water pressure and surface velocity of Findelen­ gletscher, Switzerland: conclusions about drainage system and sliding mechanism. Jounal of Glaciology, 32(110): 101-119.

[Iverson et al., 1999] Iverson, N., Baker, R., Hooke, R. L., Hanson, B., and Jansson, P. (1999). Coupling between a glacier and a soft bed: I. a relation between efective pressure and local shear stress determined from till elasticity. Jounal of Glaciology, 45(149):31-40.

[Jansson, 1995] Jansson, P. (1995). Water pressure and basal sliding on Stor­ glaciaren, northern Sweden. Jounal of Glaciology, 41(138):232-240.

[Johannesson et al., 1995] Johannesson, T., Sigurdsson, 0., Laumann, T., and Ken­ net, M. (1995). Degree-day mass-balance modelling with applications to glaciers in Iceland, Norway and Greenland. Jounal of Glaciology, 41(138):345-358.

[Kohnen, 1974] Kohnen, H. (1974). The temperature dependence of seismic waves in ice. Jounal of Glaciology, 13(67):144-147.

[Kuhn, 1981] Kuhn, M. (1981). Climate and glaciers. Intenational Association of Scientiic Hydology Publication, 131:3-20.

[Kuhn, 1989] Kuhn, M. (1989). The response of the equilibrium line altitude to climate luctuations: Theory and observations. J. Oerlemans (ed}, Glacier Fluc­ tuations and Climatic Change. Dortrecht: Kluwer Academic Publishers, pages 407-417.

[Liss, 1970] Liss, C. (1970). Der Morenogletscher in der Patagonischen Kordillere. Zeitschrit fur Gletscherkunde und Glazialgeologie, V1(1-2): 161-180.

[Lliboutry, 1956] Lliboutry, L. (1956). Nieves y glaciares de Chile: undamentos de glaciologia. Ediciones de la Universidad de Chile, page 4 71. Santiago.

[Massonet and Feigl, 1998] Massonet, D. and Feigl, K. (1998). Radar interferometry and its application to changes in the earth's surface. Rev. Geophys., 36(4):441- 500.

[Meteorolgica, 1966] Meteorolgica, 0. (1966). Pluviometria de Chile. Ministerio de Defensa Nacional, Santiago, Chile.

[Michel and Rignot, 1999] Michel, R. and Rignot, E. (1999). Flow of Glaciar Moreno, Argentina, from repeat-pass shuttle imaging radar images: comparison of the phase correlation method with radar interferometry. Jounal of Glaciology, 45(149):93-100.

[Miller, 1976] Miller, A. (1976). The climate of Chile. Wo rld Survey of Climatology, 12:1 13-145. Climates of Central and South America. 143 BIBLIOGAPHY

[Miller, 1982] Miller, H. (1982). Elastic properties. Landolt-Bonstein, Numeri­ cal Data and Fu nctional Relationships in Science and Technology, 1, Subvolume b:496-. Group V: Geophysics and Space Research.

[Nagler, 1996] Nagler, T. (1996). Methods and Analysis of Synthetic Aperture Radar Data from ERS-1 and X-SAR for Snow and Glacier Applications. PhD thesis, Leopold-ranzens-U niversi tat, Inns bruck.

[Naruse and Aniya, 1992] Naruse, R. and Aniya, M. (1992). Outline of glacier re­ search project in Patagonia. Bulletin of Glacier Research, 10:31-38.

[Naruse and Aniya, 1995] Naruse, R. and Aniya, M. (1995). Synopsis of glacier researches in Patagonia, 1993. Bulletin of Glacier Research, 13:1-10.

[Naruse et al., 1995a] Naruse, R., Aniya, M., Skvarca, P. , and Casassa, G. (1995a). Recent variations of calving glaciers in Patagonia, South America, revealed by ground surveys, satellite-data analyses and numerical experiments. Annals of Glaciology, 21:297-303.

[Naruse et al., 1992] Naruse, R., Skvarca, P. , Kadota, T., and Koizumi, K. (1992). Flow of Upsala and Moreno glaciers, Southern Patagonia. Bulletin of Glacier Research, 10:55-62.

[Naruse et al., 1995b] Naruse, R., Skvarca, P. , Satow, K., Takeuchi, Y., and Nishida, K. (1995b). Thickness change and short-term low variation of Moreno Glacier, Patagonia. Bulletin of Glacier Research, 13:21-28.

[Nichols and Miller, 1952] Nichols, R. and Miller, M. (1952). The Moreno Glacier, Lago Argentino, Patagonia. advancing glaciers and nearby simultaneously retreat­ ing glaciers. Jounal of Glaciology, 2(11):41-50.

[Nye, 1953] Nye, J. (1953). The low law of ice from measurements in glacier tunnels, laboratory experiments and the Jungfrauirn borehole experiment. Poceedings of the Royal Society London, Ser. A(219):477-489.

[Nye, 1965] Nye, J. (1965). The low of a glacier in a channel of rectangular, elliptic or parabolic cross-section. Jounal of Glaciology, 5(41):661-690.

[Paterson, 1994] Paterson, W. (1994). The physics of glaciers. Pergamon Press, 3rd edition.

[Pena and Gutierrez, 1992] Pena, H. and Gutierrez, R. (1992). Statistical analysis of precipitation and air temperature in the Southern Patagonia Iceield. Glacio­ logical Researches in Patagonia, 1990. Report of the Japan-Argentina-Chile Joint Poject., pages 95-108. Japanese Society of Snow and Ice. 144 BIBLIOGRAPHY

[Prohaska, 1976] Prohaska, F. (1976). The climate of Argentina, Paraguay and Uruguay. World Survey of Climatology, 12:13-112. Climates of Central and South America.

[Rafo et al., 1953] Rafo, J., Colqui, B., and Madejski, M. (1953). Glaciar Moreno. Revista Meteoos, 3(4):293-341. Buenos Aires. Direccion General del Servicio Meteorologico Nacional. Publicacion 9.

[Rango and Martinec, 1995] Ra.ngo, A. and Martinec, J. (1995). Revisiting the degree-day method for snowmelt computations. Water Resources Bulletin, 31(4):657-669.

[Reichert, 1917] Reichert, F. (1917). Vorlauiger Bericht iber das Ergebnis der Expedition nach dem sog. Patagonischen Inlandeis und dem Bismarckgletscher. Zeitschrit fur Gletscherkunde, 10:225-230.

[Rignot et al., 1996] Rignot, E., Forster, R., and !sacks, B. (1996). Interfero­ metric radar observations of Glaciar San Rafael, Chile. Jounal of Glaciology, 42(141) :279-291.

[Rosenbluth et al., 1995] Rosenbluth, B., Casassa, G., and Fuenzalida, H. (1995). Recent climatic changes in Western Patagonia. Bulletin of Glacier Research, 13:127-132.

[Rott et al., 1997] Rott, H., Skvarca, P., Rack, W., and Stuefer, M. (1997). Signif­ icant ice retreat in the region Patagonia - Antarctic Peninsula observed by ERS SAR. ESA SP-4 14, in press. Proc. of 3rd ERS Symposium, Space at the Service of our Environment.

[Rott et al. , 1998] Rott, H., Stuefer, M., Siegel, A., Skvarca, P., and Eckstaller, A. (1998). Mass luxes and dynamics of Moreno Glacier, Southern Patagonia Iceield. Geophysical Research Letters., 25(9): 1407-1410.

[Schwabisch, 1995] Schwabisch, M. (1995). Die SAR-Interferometrie zr Erzeugung digitaler Gelandemodelle. Forschungsbericht 95-25, Deutsche Forschungsanstalt fr Luft- und Raumfart e.V.

[Schwerdtfeger, 1958] Schwerdtfeger, W. (1958). Ein Beitrag zur Kenntnis des Kli­ mas im Gebiet der Patagonischen Eisfelder. Zeitschrit fur Gletscherkunde und Glazialgeologie, 4(1-2):73-86.

[Siegel, 1996] Siegel, A. (1996). Methoden der Radarinterferometrie fur glaziologis­ che ragestellungen. Diplomarbeit, Institut fur Meteorologie und Geophysik.

[Skvarca and Naruse, 1997] Skvarca, P. and Naruse, R. (1997). Dynamic behavior of Glaciar Perito Moreno, Southern Patagonia. Annals of Glaciology, 24:268-271. 145 BIBLIOGAPHY

[Skvarca et al., 1995] Skvarca, P. , Rott, H., and Stuefer, M., editors (1995). Syn­ ergy of ERS-1 SA R, X-SAR, Landsat TM imagery and aerial photography for glaciological studies of Vi edma Glacier, southen Patagonia. VII Simposio Lati­ noamericano de Percepcion Remota, SELPER, Puerto Vallarta, Mexico. [Takeuchi et al. , 1995a] Takeuchi, Y., Naruse, R., and Satow, K. (1995a). Charac­ teristics of heat balance and ablation on Moreno and Tyndall glaciers, Patagonia, in the summer 1993/94. Bulletin of Glacier Research, 13:45-56. [Takeuchi et al., 1996] Takeuchi, Y., Naruse, R., and Skvarca, P. {1996). Annual air­ temperature measurement and ablation estimate at Moreno Glacier, Patagonia. Bulletin of Glacier Research, 14:23-28. [Takeuchi et al., 1995b] Takeuchi, Y., Satow, K., Naruse, R., lbarzabal, T., Nishida, K., and Matsuoka, K. (1995b). Meteorological features at Moreno and Tyndall glaciers, Patagonia, in the summer 1993/94. Bulletin of Glacier Research, 13:35- 44. [Thyssen, 1967] Thyssen, F. (1967). Die Temperaturabhangigkeit der P- Wellengeschwindigkeit in Gletschern und Inlandeisen. Zeitschrit fu er Geophysik, 33(1):65-79. [Warren, 1994] Warren, C. (1994). reshwater calving and anomalous glacier os­ cillations: recent behaviour of Moreno and Ameghino glaciers, Patagonia. The Holocene 4, pages 422-429. [Warren et al., 1997] Warren, C., Rivera, A., and Post, A. (1997). Greatest holocene advance of Glaciar Pio XI, Chilean Patagonia: possible causes. Annals of Glaciol­ ogy, 24:11-15. [Warren and Sugden, 1993] Warren, C. and Sugden, D. {1993). The Patagonian Iceields: A glaciological review. Arctic and Alpine Research, 25:316-331. [Wu, 1995] Wu , Q. {1995). A correlation-relaxation-labelling framework for comput­ ing optical low - template matching from a new perspective. IEEE ransactions on Patten Analysis and Machine Intelligence, 17(8):843-853. [Wu et al., 1997] Wu , Q., Neill, S. M., and Pairman, D. (1997). Correlation and re­ laxation labelling: an experimental investigation on fast algorithms. Intenational Jounal of Remote Sensing, 18(3):651-662. [Zebker and Goldstein, 1986] Zebker, H. and Goldstein, R. (1986). Topographic mapping from interferometric synthetic aperture radar observations. Jounal of Geophysical Research, 91(B5):4993-4999. [Zebker et al. , 1997] Zebker, H., Rosen, P., and Hensley, S. (1997). Atmospheric efects in interferometric synthetic radar surface deformation and topographic maps. Journal Geophysical Research, 102:7547-7563. 146 BIBLIOGAPHY

[Zingg, 1951] Zingg, T. (1951). Beziehung zwischen Temperatr und Schmelzwasser und ihre Bedeutung Ur Niederschlags- und Abiussfragen. IAHS Publication, 1:267-269. IUGG Gen. Ass. Brussels. Appendix A

Chronological description of ield activities

A.l Field measurements in November /December 1995

Pa.rticipants : Helmut Rott 1 , Pedro Skv.rc.2 , Teodoro Toconas2 , Thomas N .gler1 , Martin Stuefer1 1: Institut fir Meteorologie und Geophysik, Universitat Innsbruck (IMGI) 2: Instituto Antartico Argentino (IAA) The ield work was carried out between 14 November and 10 December 1995. Two camp sites next to the right glacier margin were used in order to reach the locations of the ield activities in comparatively short time: the Moreno Base camp and the Busc.ini camp at . distance of 8 km up-glacier (igure 4-1). On 14 and 15 November 1995 an automatic climate station was installed at the shore of the Br.zo Sur at . distance of 360 m to the orogr.phica.lly right terminus of the glacier (igure 4-2). The meteorological data gained during the irst ield campaign were transferred to . PC on 8 December. rom 16 to 25 November 1995 19 ablation stakes were drilled along two transverse and one longitudinal proiles (igures 4-1, -3). On 1 December 1995 P. Skva.rc. installed stake D. The positions of the 19 stakes were measured from 21 to 24 November. Repeated measurements of the stake heights above the ice surface during the geodetic and drilling work and additional measurements of the stake heights on 9 December provided information on the ablation rate during the ield campaign. The highest ablation was observed at stake A05 (table A.1). The ablation rates of the stakes A01 to A04 varied due to diferent surface roughness; stake A02 was placed at the smooth top of . little plateau. For . period of 3 weeks until 9 December 1995 the ablation rates at the B-proile

147 Appendix A. Chmnological desciption of ield activities 148

stake date I date II ablation A05 25-Nov-95 9-Dec-95 7.1 A01 21-Nov-95 9-Dec-95 5.9 A02 21-Nov-95 9-Dec-95 5.3 A03 24-Nov-95 9-Dec-95 5.7 A04 21-Nov-95 9-Dec-95 6.2 Table A.1: Mean daily ablation rates of A-proile stakes, measurement periods are from date I to date II. stake date I date II ablation B01 16-Nov-95 9-Dec-95 6.5 B02 16-Nov-95 9-Dec-95 5.1 B03 16-Nov-95 9-Dec-95 5.2 B04 17-Nov-95 9-Dec-95 4.4 B05 17-Nov-95 9-Dec-95 4.5 B06 17-Nov-95 9-Dec-95 5.1 Bll 19-Nov-95 9-Dec-95 4.5 B07 17-Nov-95 9-Dec-95 3.2 B08 17-Nov-95 9-Dec-95 5.4 B09 18-Nov-95 9-Dec-95 5.5 BlO 18-Nov-95 9-Dec-95 6.0 101 18-Nov-95 9-Dec-95 4.9 102 18-Nov-95 9-Dec-95 4.6 103 19-Nov-95 9-Dec-95 4.4 Table A.2: Mean daily ablation rates of B and C-proile stakes, measurement periods are from date I to date II. varied between 3.2 and 6.5 em/day (table A.2),(igure A-1). Whereas the ablation was lower at the stakes in smooth areas of the central parts, the highest ablation rates were measured in the strongly crevassed zones of the margin. Enhanced abla­ tion at certain stakes of the central zone (B06) was observed in an area of surface undulations. Three weeks of ablation measurements at the longitudinal C-proile revealed only minor diferences. The ablation rates varied rom 4.4 em/day (103) to 4.9 em/day (101)(table A.2).

A.2 Field measurements in March/ April 1996

Participants: Helmut Rott1 , Pedro Skvarca2 , Wolfgang Rack1 , Martin Stuefer1 1: Institut fir Meteorologie und Geophysik, Universitat Innsbruck (IMGI) 2: Instituto Antartico Argentino (IAA) The measurements at the Moreno Glacier were carried out from 16 March to 6 April 1996. The automatic climate station was checked and the stored data were Appendix A. Chonological description of ield activities 149

7.0 .. 6.5

6.0

5.0

4.0 � E 3.0

2.0

1.0

0.0 BOl B02 B03 B04 B05 B06 Bll B07 BOB B09 BlO

Figure A-1: Mean daily ablation rates of B-proile stakes averaged over 3 weeks until 9 December 1995.

downloaded. to the notebook. In order to store the meteorological data for more than 6 months a second storage module was connected to the datalogger. Additionally, the stora- ge interval was changed from 15 to 20 minutes. Two solar modules were $ installed besides the meteorological station as additional power supply.

u Middle' format photographs of the terminus were taken on 18 March and 31 March from about 700 m a.s.l. of the northeastern slope of Cerro Moreno; the photographs revealed no short-term changes of the calving glacier terminus. rom 19 March until 1 April the three proiles were redrilled at the original positions of November 1995, the maximal drilling depth was 9.9 m. Stake B07 of the B-proile was eliminated as the position was very close to the adjacent stakes Bll and B08 (igure 4-3). 10 lines were measured with DGPS along the coast of Brazo Sur and along parts of the glacier margin, and close to the calving front the lake depth of Brazo Sur was measured at 132 points by means of an echo sounding instrument.

A.3 Field measurements in November /December 1996

Participants: Helmut Rott1, Pedro Skvarca2, Alfons Eckstaler3, Juan Carlos Quinteros, Nico- las Benedetti, Wolfgang Rack1 , Martin Stuefer1 1: Institut fiir Meteorologic und Geophysik, Universitat Innsbruck (IMGI) 2: Instituto Anthtico Argentino (IAA) 3: Alfred-Wegeuer-Institut ir Polar- und Meeresforschung, Bremer haven (l) Appendix A. Chronological description of ield activities 150

The ield campaign was carried out from 16 November to 10 December 1996. An electronic raingauge was added to the climate station. In order to get com­ parative temperatre values during the ield campaign, a 2nd temperature sensor was installed at the A-proile near stake A03, which was set up at the previous position of November 1995. Seismic- relection measurements were carried out to determine the ice thickness. All stakes drilled during the previous campaign were found and the positions were measured by means of DGPS. The 18 stakes of the three proiles were redrilled at the original positions. The DGPS measurements of the summits of Cerro Perito Moreno, Cerro Buenos Aires, and of a geodetic line along the orographic left glacier border in a distance of 8 km to the glacier terminus completed the campaign.

A.4 Field measurements in March/ April 1997

Participants: Helmut Rott1 , Pedro Skvarca2, Christine Miller3, Martin Stuefer1 1: Institut fir Meteorologie und Geophysik, Universitat Innsbruck (IMGI) 2: Institute Antartico Argentino (IAA) 3: Institut fur Mineralogie und Petrographie, Universitat Innsbruck The campaign was carried out from 24 March to 3 April 1997. The climate station worked without interruption, the data from the two storage modules was transferred to a notebook. The raingauge was removed and the storage interval was changed from 20 to 30 minutes. From 25 to 27 March all of the 18 stakes were found and the positions were measured by means of DGPS. The stake heights were between 5.24 m (101) and 8.75 m (A01, BOl) . The ablation information at B09 was lost and the position measurement was inaccurate because the stake was lying at the bottom of a crevasse. The stake net was reduced to 3 stakes at A proile and 5 stakes at B proile to maintain a baseline for multi-year ablation. The maximum drilling depth was 15,96 m (stake AOl, 2 April 1997) . On 1 April 1997 weak winds and a smooth lake surface allowed echo sounding in the Canal de los Tempanos.

A.5 Field measurements in November 1997

Wolfgang Rack (IMGI) visited Patagonia on the retrn trip from the Antarctic Peninsula on 13 and 14 November 1997. The data from the climate station were downloaded and the ablation at the 8 stakes was measred. Appendix A. Chronological description of ield activities 151

A.6 Field measurements in March/ April 1998

Participants: Martin Stuefer1 , Pedro Skvarca2 1: Institut fir Meteorologie und Geophysik, Universitat Innsbruck (IMGI) 2: Instituto Antartico Argentino (IAA) The ield campaign was carried out from 19 March until 11 April 1998. The position and the ablation of 8 stakes set up one year before were measured. The position of the lower orographically right glacier margin was measured by means of a GPS line, which reached a length of 4.26 km from the DGPS base station (section A.1) along the shore of Brazo Sur and the glacier margin (igure 4-26) . Whereas the main part of the repeated line showed only minor changes since March 1996 (igure 4-25, lines 6 and 7) , a glacier advance of 45 m occurred where the right margin extends into Brazo Sur. On 20 and 21 March 1998 the A-proile was measured and redrilled. The 3 stakes were reset to the 1997 positions, the drilling depths were almost 14 m. The 5 stakes of the B proile were measured and reset from 23 March 1998 until 29 March. 4 stakes were redrilled into the ice to depths of almost 14 m, one stake (B03) was drilled to 8.92 m due to problems with the gas supply of the steam drill. The glacier observations during the ield campaign were supplemented by oblique photographs from the summit of Cerro Buenos Aires taken on 31 March 1998 and by photographs from the region around the snowline in the middle part of the Moreno Glacier. A check of the climate station and the download of the climate data collected until 11 April 1998 inished this austral autumn campaign.

A. 7 Field measurements in March/ April 1999

Helmut Rott1 , Pedro Skvarca2 , Stephan Hoinger3, Martin Stuefer1 1: Institut fir Meteorologie und Geophysik, Universitat Innsbruck (IMGI) 2: Instituto Antartico Argentino (IAA) 3: Zentralanstalt fir Meteorologie und Geodynamik, Zweigstelle Salzburg Field work was carried out on Moreno Glacier and on Ameghino Glacier from 3 March until 31 April 1998. rom 6 March 1999 until lO March 1999 the positions and the ablation of 8 ablation stakes drilled in 1998 on Moreno Glacier were measured; the 8 stakes were redrilled. The ablation value of stake B03 could not be measured, as the stake was melted out. The stake lying on the surface besides a hollow marked the position of B03. Base Camp climate data were downloaded and the climate station checked. rom 6 March 1999 until 22 March 1999 a second Kipp Zonen & radiometer was ixed in order to re-calibrate radiation measurements (igure C-1) . Appendix B

Positions measured with GPS

B.l Stakes

The stakes have been measured by means of diferential GPS. The tables B.1 to B.13 include the height above the ellipsoid (HAE) and the distance to the GPS base station (baseline length) in meter units. The coordinates are referred to the vV GS 84 geographic datum. Additional to the stake positions the position of a large stone with a length of about 8 m in the region of the centerline of Moreno Glacier was measred (table B.14) .

B.2 Seismic stations

The coordinates of the seismic stations were derived by DGPS measurement with the exception of two points at the lower A-proile. The position of the seismic station SHPA3 was measured with a Garmin GPS, SHPA4 was estimated to be 110 meter from the original A03 ablation stake downward; both positions are less accurate

Stake Date Latitude Longitude HAE Baseline L. A05 21 Nov 95 50d 30' 08.3576" 73d 05' 34. 6073" 380.27 3660.59 s w A05 17 Mar 96 50d 30' 08.3668" 73d 05' 33.4182" 376.69 3638.92 s w A05 17 Mar 96 50d 30' 08.3708" 73d 05' 33.4157" 376.64 3638.92 s w A01 21 Nov 95 50d 29' 51.2341" 73d 05' 33.2153" 379.50 3465.78 s w AOl 17 Mar 96 50d 29' 50.5241" 73d 05' 30.6047" 375.50 3410.34 s w A02 21 Nov 95 50d 29' 39.7976" 73d 05' 42.4714" 381.34 3574.38 s w A02 17 Mar 96 50d 29' 38.4482" 73d 05' 36.5292" 372.91 3452.25 s w A03 21 Nov 95 50d 29' 31.6894" 73d 05' 49.1787" 386.64 3678.62 s w A03 17 Mar 96 50d 29' 30.3175" 73d 05' 40.8353" 376.50 3511.33 s w A04 21 Nov 95 50d 29' 24.4149" 73d 05' 56.0823" 385.60 3804.97 s w A04 17 Mar 96 50d 29' 23.2073" 73d 05' 46.5875" 370.76 3617.07 s w Table B.1: Coordinates of lower A-proile stakes drilled in November 1995.

152 Appendix B. Positions measured with GPS 153

Stake Date Latitude Longitude HAE Baseline L. B01 23 Nov 95 50d 30' 53.6598" 73d 07' 49.3319" 502.66 6657.43 s w B01 28 Nov 95 50d 30' 53.6607" 73d o7' 49.2514" 502.81 6656.01 s vv B01 19 Mar 96 50d 30' 53.9084" 73d 07' 47.8065" 495.80 6633.15 s w B02 23 Nov 95 50d 30' 41.0304" 73d 07' 58.7465" 520.63 6676.27 s w B02 28 Nov 95 50d 30' 40.9492" 73d 07' 58.4901" 520.26 6670.64 s w B02 19 Mar 96 50d 30' 39.5537" 73d 07' 53.5663" 509.47 6564.21 s w B03 24 Nov 95 50d 30' 29.0206" 73d 08' 05.8856" 531.25 6685.71 s w B03 28 Nov 95 50d 30' 28.8822" 73d 08' 05.5670" 531.33 6678.43 s w B03 19 Mar 96 50d 30' 25.7791" 73d 07' 57.5287" 519.44 6498.12 s w B04 24 Nov 95 50d 30' 17.2624" 73d 08' 09.9568" 536.06 6661.85 s w B04 28 Nov 95 50d 30' 17.0988" 73d 08' 09.5861" 535.67 6653.49 s w B04 19 Mar 96 50d 30' 13.3739" 73d 08' 01.0928" 525.66 6462.67 s w B05 23 Nov 95 50d 30' 08.0726" 73d 08' 18.2820" 524.25 6755.94 s w B05 28 Nov 95 50d 30' 07.8906" 73d 08' 17.8844" 523.63 6747.10 s w B05 19 Mar 96 50d 30' 03.9793" 73d 08' 09.4982" 512.00 6560.72 s w B06 23 Nov 95 50d 29' 54.5647" 73d 08' 25.4697" 501.43 6821.60 s w B06 28 Nov 95 50d 29' 54.3787" 73d 08' 25.0374" 501.01 6812.34 s vv B06 19 Mar 96 50d 29' 50.5934" 73d 08' 16.6151" 489.69 6631.73 s w B11 23 Nov 95 50d 29' 45.2856" 73d 08' 31.8531" 509.02 s w B11 19 Mar 96 50d 29' 41.8273" 73d 08' 23.2913" 495.92 6733.73 s w B07 23 Nov 95 50d 29' 40.3508" 73d 08' 35.4818" 504.70 6969.69 s w B07 28 Nov 95 50d 29' 40.2154" 73d 08' 35.1646" 504.36 6963.13 s w B07 19 Mar 96 50d 29' 37.2222" 73d 08' 27.3210" 491.12 6802.00 s w B08 23 Nov 95 50d 29' 32.6456" 73d 08' 31.7541" 500.01 6881.99 s w B08 23 Mar 96 50d 29' 29.5878" 73d 08' 23.8453" 487.40 6722.37 s w B09 23 Nov 95 50d 29' 11.4167" 73d 08' 39.1224" 493.63 7029.85 s w B09 20 Mar 96 50d 29' 09.6835" 73d 08' 35.3912" 486.59 6959.15 s w B10 23 Nov 95 50d 28' 59.1852" 73d 08' 42.2662" 480.46 7120.47 s w B10 20 Mar 96 50d 28' 58.3305" 73d 08' 41.3535" 475.00 7105.18 s w Table B.2: Coordinates of upper B-proile stakes drilled in November 1995.

Stake Date Latitude Longitude HAE Baseline L. , 101 23 Nov 95 50d 31 08. 7973" 73d 09' 44.3002" 676.67 8930.68 s w 101 24 Mar 96 50d 31' 01.0743" 73d 09' 36.0464" 654.14 8692.67 s w 102 23 Nov 95 50d 30' 57. 7222" 73d 09' 21.0922" 640.78 8380.12 s vV L02 24 Mar 96 50d 30' 50.9350" 73d 09' 11.6974" 613.42 8133.42 s w L03 23 Nov 95 50d 30' 41.4446" 73d 08' 53.9841" 587.06 7707.35 s w L03 24 Mar 96 50d 30' 35.9687" 73d 08' 43.8702" 568.20 7464.97 s w Table B.3: Cordinates of longitudinal C-proile stakes drilled in November 1995.

Stake Date Latitude Longitude HAE Baseline L. D 17 Mar 96 50d 29' 22.1900" 73d 03' 18.5726" 236.62 696.58 s w D 8 Apr 96 50d 29' 22.2481" 73d 03' 18.4314" 235.11 693.70 s w Table B.4: Coordinates of stake D drilled in November 1995. Appendix B. Positions measured with GPS 154

Stake Date Latitude Longitude RAE Baseline L. A05 29 Mar 96 50d 30' 09.2337" 73d 05' 36.9278" 374.03 3712.88 s w A05 19 Nov 96 50d 30' 09.1964" 73d 05' 35.1087" 377.62 3679.57 s w A01 29 Mar 96 50d 29' 52.2688" 73d 05' 32.5416" 374.80 3460.94 s , w A01 19 Nov 96 50d 29' 51.0827" 73d 05 28.1482" 379.76 3368.15 s w A02 29 Mar 96 50d 29' 38.6330" 73d 05' 40.6405" 373.82 3533.26 s w A02 19 Nov 96 50d 29' 35.9505" 73d 05' 29.2137" 373.54 3299.44 s w A03 29 Mar 96 50d 29' 31.0123" 73d 05' 48.5682" 379.59 3664.83 s w A03 20 Nov 96 50d 29' 28.3161" 73d 05' 32.7799" 377.71 3349.52 s w A04 29 Mar 96 50d 29' 24.3088" 73d 05' 56.8134" 379.29 3819.05 s w A04 19 Nov 96 50d 29' 22.0540" 73d 05' 39.0418" 367.20 3468.43 s w Table B.5: Coordinates of A-proile stakes drilled in March 1996.

Stake Date Latitude Longitude RAE Baseline L. B01 27 Mar 96 50d 30' 54.2997" 73d 07' 49.4625" 498.62 6667.92 s w B01 22 Nov 96 50d 30' 54.8528" 73d 07' 46. 7918" 497.31 6627.57 s vv B02 27 Mar 96 50d 30' 41.9917" 73d 07' 56.6784" 515.65 6648.93 s w B02 22 Nov 96 50d 30' 39.8953" 73d 07' 48.0901" 510.14 6467.68 s w B02 26 Nov 96 50d 30' 39.8246" 73d 07' 47.8346" 512.20 6462.29 s w B03 19 Mar 96 50d 30' 29. 7861" 73d 08' 09.0477" 531.12 6752.19 s w B03 22 Nov 96 50d 30' 23.1269" 73d 07' 51.7593" 519.43 6365.45 s w B04 23 Mar 96 50d 30' 18.2303" 73d 08' 10.6181" 531.85 6681.80 s w B04 22 Nov 96 50d 30' 10.5271" 73d 07' 53.0589" 520.71 6287.97 s w B05 19 Mar 96 50d 30' 08.8122" 73d 08' 17.6654" 523.59 6748.78 s w B05 22 Nov 96 50d 30' 00.5119" 73d 07' 59.8078" 509.98 6353.15 s w B06 23 Mar 96 50d 29' 54.6713" 73d 08' 24.6440" 496.11 6805.76 s w B06 24 Nov 96 50d 29' 46.7168" 73d 08' 07.0202" 484.96 6429.48 s w B11 23 Mar 96 50d 29' 44.1075" 73d 08' 31.6165" 502.40 6903.64 s w B11 24 Nov 96 50d 29' 37.1816" 73d 08' 14.9834" 488.10 6559.32 s w B11 28 Nov 96 50d 29' 37.0529" 73d 08' 14.3793" 488.29 6547.20 s w B08 23 Mar 96 50d 29' 33. 7058" 73d 08' 33.3513" 498.21 6914.82 s w B08 24 Nov 96 50d 29' 27.6953" 73d 08' 17.7908" 487.14 6601.64 s w B09 20 Mar 96 50d 29' 11.2882" 73d 08' 38.2724" 488.78 7013.14 s w B09 24 Nov 96 50d 29' 07.8266" 73d 08' 30.7899" 487.04 6872.39 s w B10 20 Mar 96 50d 28' 57.8825" 73d 08' 43.3675" 470.48 7145.96 s w B10 24 Nov 96 50d 28' 56.5589" 73d 08' 41.8854" 471.96 7121.59 s w Table B.6: Coordinates of B-proile stakes drilled in March 1996. Appendix B. Pos·itions measured with GPS 155

Stake Date Latitude Longitude HAE Baseline L. L01 24 Mar 96 50d 31' 07.4987" 73d 09' 43.6146" 668.61 8903.10 s w L01 25 Nov 96 50d 30' 52.7550" 73d 09' 26.1664" 636.44 8421.84 s w L02 24 Mar 96 50d 30' 59.1941" 73d 09' 22.8525" 642.60 8428.59 s vv L02 25 Nov 96 50d 30' 46.1071" 73d 09' 03.9822" 603.14 7940.54 s w L02 28 Nov 96 50d 30' 45.9474" 73d 09' 03. 7676" 602.59 7934.93 s w L03 24 Mar 96 50d 30' 41.6235" 73d 08' 53.1819" 583.47 7693.93 s w L03 25 Nov 96 50d 30' 31.3415" 73d 08' 33.2506" 561.99 7222.66 s w Table B.7: Coordinates of longitudinal C-proile stakes drilled in March 1996.

Stake Date Latitude Longitude HAE Baseline L. A05 19 Nov 96 50d 30' 13.1978" 73d 05' 41.8921" 377.47 3851.12 s w A05 25 Mar 97 50d 30' 13.1546" 73d 05' 40.7830" 373.47 3830.43 s w A01 20 Nov 96 50d 29' 52.2503" 73d 05' 37.1875" 381.06 3549.47 s vv A01 25 Mar 97 50d 29' 51.4469" 73d 05' 34.3587" 376.39 3489.08 s w A02 20 Nov 96 50d 29' 40. 7592" 73d 05' 38.2638" 378.39 3496.84 s w A02 25 Mar 97 50d 29' 39.4062" 73d 05' 32.7051" 371.45 3381.94 s w A03 19 Nov 96 50d 29' 30.5990" 73d 05' 48.6671" 384.63 3666.17 s w A03 25 Mar 97 50d 29' 29.1681" 73d 05' 39.7460" 375.44 3487.76 s w A04 20 Nov 96 50d 29' 24.6393" 73d 05' 51.1374" 379.39 3707.95 s w A04 25 Mar 97 50d 29' 24.2900" 73d 05' 41.6092" 369.07 3519.17 s w Table B.8: Coordinates of A-proile stakes drilled in November 1996.

Stake Date Latitude Longitude HAE Baseline L. B01 22 Nov 96 50d 30' 55.0592" 73d 07' 50.0708" 500.19 6688.79 s w B01 26 Mar 97 50d 30' 55.3943" 73d 07' 48.6919" 492.80 6668.34 s w B02 22 Nov 96 50d 30' 40.6655" 73d 07' 56.7395" 519.46 6635.31 s w B02 26 Mar 97 50d 30' 39.2648" 73d 07' 51.6117" 510.69 6525.16 s w B03 22 Nov 96 50d 30' 29.5299" 73d 08' 11.5270" 537.22 6796.57 s w B03 26 Mar 97 50d 30' 25.8401" 73d 08' 02.2480" 522.55 6587.54 s w B04 22 Nov 96 50d 30' 19.3657" 73d 08' 11.1991" 536.53 6702.11 s w B04 26 Mar 97 50d 30' 15.3255" 73d 08' 01.8140" 526.21 6491.22 s w B05 22 Nov 96 50d 30' 08.0921" 73d 08' 20.1196" 522.48 6791.40 s w B05 27 Mar 97 50d 30' 03.6992" 73d 08' 10. 7636" 509.98 6583.42 s w B06 24 Nov 96 50d 29' 54.0258" 73d 08' 24.8842" 500.73 6807.81 s w B06 27 Mar 97 50d 29' 49.9550" 73d 08' 15.8065" 490.26 6613.46 s w Bll 24 Nov 96 50d 29' 41.3070" 73d 08' 32.0045" 500.41 6903.58 s w Bll 27 Mar 97 50d 29' 37.9183" 73d 08' 23.5344" 488.11 6728.88 s w B08 24 Nov 96 50d 29' 34.4167" 73d 08' 32.8475" 500.24 6906.08 s w B08 27 Mar 97 50d 29' 31.3225" 73d 08' 24.8573" 487.86 6744.12 s w B09 24 Nov 96 50d 29' 11.0458" 73d 08' 38.8490" 491.49 7024.99 s vv B09 27 Mar 97 50d 29' 09.3438" 73d 08' 35.2129" 484.46 6956.20 s w B10 24 Nov 96 50d 28' 57.3119" 73d 08' 44.5134" 474.68 7170.50 s w B10 27 Mar 97 50d 28' 56.6435" 73d 08' 43. 7606" 469.47 7157.90 s vv Table B.9: Coordinates of B-proile stakes drilled in November 1996. Appendix B. Positions measured with GPS 15G

Stake Date Latitude Longitude HAE Baseline L. , L01 25 Nov 96 50d 31 07. 7854" 73cl 09' 44.0413" 673.22 8914.38 s w L01 26 Mar 97 50d 31' 00.1201" 73cl 09' 35.6362" 653.74 8674.85 s w L02 25 Nov 96 50d 30' 59.2961" 73cl 09' 23.7914" 645.09 8447.12 s w L02 27 Mar 97 50d 30' 52.4261" 73cl 09' 14.4781" 619.16 8200.70 s w L03 25 Nov 96 50d 30' 42.8075" 73d 08' 52.7140" 589.75 7697.14 s vv L03 27 Mar 97 50d 30' 37.4010" 73l 08' 42.7617" 572.53 7457.89 s w Table B.lO: Coordinates of C-proile stakes drilled in November 1996.

Stake Date Latitude Longitude HAE Baseline L. AOl 31 Mar 97 50d 29' 48.9262" 73d 05' 42.2691" 378.06 3622.06 s w A01 20 Mar 98 50d 29' 46.2428" 73d 05' 31.7053" 376.15 3400.98 s w A02 31 Mar 97 50d 29' 38.4769" 73d 05' 49.4882" 382.92 3705.71 s w A02 20 Mar 98 50d 29' 34.4732" 73d 05' 30.0812" 374.57 3311.24 s w A04 31 Mar 97 50d 29' 28.1655" 73d 05' 58.2800" 382.31 3851.15 s w A04 20 Mar 98 50d 29' 24.4580" 73d 05' 32.9006" 370.58 3347.82 s w B03 27 Mar 97 50d 30' 29.1880" 73d 08' 13.8388" 537.85 6836.82 s w B03 23 Mar 98 50d 30' 18.7341" 73d 07' 47.7172" 515.78 6250.51 s w B04 27 Mar 97 50d 30' 23.6619" 73d 08' 21.2193" 542.14 6927.72 s w B04 23 Mar 98 50d 30' 11.755 1" 73d 07' 53.8282" 520.22 6311.60 s w B05 28 Mar 97 50d 30' 14.5529" 73d 08' 27.7433" 534.73 6981.67 s w B05 23 Mar 98 50d 30' 01.9458" 73d 08' 00.5745" 511.33 6376.25 s w B06 30 Mar 97 50d 29' 58.7887" 73d 08' 34.4447" 508.61 7016.21 s w B06 23 Mar 98 50d 29' 46.8037" 73d 08' 07.8587" 484.62 6446.17 s w B11 30 Mar 97 50d 29' 46.4388" 73d 08' 39.0628" 512.23 7057.06 s w B11 23 Mar 98 50d 29' 36.2692" 73d 08' 13.8427" 487.84 6535.07 s w Table B.11: Coordinates of A- and B-proile stakes drilled in March 1997. Appendix Positions measured with GPS 157 B.

Stake Date Latitude Longitude HAE Baseline L. A01 20 Mar 98 50d29'50.3947"S 73d05 '42. 7602" 379.40 3641.83 w A01 19 Dec 98 50d29' 48.3430" 73d05 '34. 9648" 379.34 s w A01 7 Mar 99 50d29'47.8181"S 73d05'33.0221"W 375.59 3436.83 A02 21 Mar 98 50d29'40.3044"S 73d05'48.2367"W 384.71 3689.12 A02 7 Mar 99 50d29'36.4026" 73d05'30.6491"W 372.52 3329.15 s A04 20 Mar 98 50d29'28.6832" 73d05'59.0435"W 385.37 3867.00 s A04 7 Mar 99 50d29'24.9806"S 73d05'33.8515" W 370.37 3366.79 B03 29 Mar 98 50d30'29.6248" 73d08'14.5560"W 542.25 6854.54 s B03 9 Mar 99 50d30'19.8230"S 73d07'50.0311"W 515.64 6303.57 B04 29 Mar 98 50d30'23.0261" 73d08'18.1731"W 542.54 6864.71 s B04 7 Mar 99 50d30' 12.0438" 73d07'52.9115"W 518.54 6296.13 s B05 29 Mar 98 50d30'14.6081" 73d08'23.9397"W 535.79 6909.23 s B05 7 iar 99 50d30'02.9487"S 73d07'58.7507"W 512.09 6346.97 B06 29 Mar 98 50d29'56.5241"S 73d08'32.3853"W 505.91 6965.33 B06 8 Mar 99 50d29' 45.3932" 73d08'07.4046"W 482.72 6432.39 s Bll 29 Mar 98 50d29'45.4781"S 73d08'37.5364"W 511.87 7024.16 Bll 8 Mar 99 50d29'35.8780" 73d08'13.7124"W 486.65 6531.72 s Table B.12: Coordinates of A- and B-proile stakes drilled in March 1998.

Stake Date Latitude Longitude HAE Baseline L. A01 10 Mar 99 50d29'49.9347"S 73d05 '44.4820" 381.54 3671.66 w A02 10 Mar 99 50d29'39. 7299" 73d05'50.2672"W 383.42 3726.12 s A04 10 Mar 99 50d29'28.2332" 73d05' 57.9961" 383.05 3845.68 s w B03 9 Mar 99 50d30'30.8868"S 73d08'12.7378"W 537.10 6832.02 B04 9 Mar 99 50d30' 19.4420" 73d08'19.5627"W 538.39 6861.91 s B05 9 Mar 99 50d30' 12. 9400" 73d08'24.4933" 532.08 6907.98 s w B06 9 Mar 99 50d29'57.2890"S 73d08'30.4443" 504.55 6931.03 w Bll 8 Mar 99 50d29' 46.0460" 73d08' 40.4164" 512.45 7082.36 s w Table B.13: Coordinates of A- and B-proile stakes drilled in March 1999.

Date Latitude Longitude HAE Baseline L. 25 Nov 96 50d 29 58.6530 S 73 08 44.7963 528.21 7217.80 w 30 Mar 97 50d 29 54.5879 S 73 08 35.0905 514.81 7009.85 w 23 Mar 98 50d 29 43.0702 S 73 08 09.1022 489.80 6458.67 w Table B.14: Coordinates of large stone in the region of the centerline at proile B. Appendix B. Positions measured with GPS 158

Station Date Latitude Longitude HAE Baseline L. SHPA1 18 Nov 96 73d05'35.1087"W 50d30'09.1964" 377.620 3679.574 s SHPA2 18 Nov 96 73d05'37.4160"W 50d29'52.3225"S 380.396 3554.345 SHPA3 19 Nov 96 73d05'44.00"W 50d29'42.00"S 373.000 3614.000 SHPA4 19 Nov 96 73d05 '43. 50" 50d29'30.30"S 377.000 3564.000 w Table B.15: Position of seismic stations at A-proile, the date represents the day of seismic mesurement. Station Date Latitude Longitude HAE Baseline L. SHP1 22 Nov 96 73d07' 46. 9594" 50d30'49.9737" S 504.683 6567.805 w SHP2 22 Nov 96 73d07'50.8344"W 50d30' 36. 2643" 513.418 6478.078 s SHP23 27 Nov 96 73d07' 52. 9089" 50d30'31.5096" 517.435 6467.017 w s SHP3 22 Nov 96 73d07' 56.8718" 50d30'24.4596" 522.373 6473.862 w s SHP34 27 Nov 96 73d07' 57. 9832" 50d30'21.6235" 524.692 6469.840 w s SHP4 22 Nov 96 73d08'01.6617" W 50d30' 12.8894" 527.572 6470.065 s SHP45 27 Nov 96 73d08'03.8357" 50d30'06.5120" 520.224 6467.538 w s SHP5 22 Nov 96 73d08'07.1798"W 50d30'00.0585" 503.146 6493.127 s SHP6R 27 Nov 96 73d08' 12.6583" 50d29'50.9718" 487.436 6555.891 w s SHP6 23 Nov 96 73d08'16.3404"W 50d29' 46. 9629" 493.152 6613.061 s SHP67 26 Nov 96 73d08'20.1397" 50d29'42 .5469" 497.183 6673.922 w s SHP7 23 Nov 96 73d08'22.9235"W 50d29'38.8824" 492.789 6719.171 s SHP8 26 Nov 96 73d08'28.0526"W 50d29'30.8216" 496.190 6806.793 s SHP9 23 Nov 96 73d08'39.000"W 50d29' 15.000" 468.000 7023.000 s Table B.16: Positions of seismic stations at B-proile, the date represents the day of seismic measurement.

(table B.15, B.16) . The irst seismic station SHPA1 was placed at the ablation stake A05, the position agrees with table B.5. Appendix C

Meteorological station

The automatic station near the Moreno Base Camp was installed from 14 to 15 November 1995 (section 4.1). The location of the station was measured with DGPS (section 4.5) .

C.O.l Te chnical details The data management in the ield is realized by a 21X micrologger from Campbell Scientiic. The 21X is a compact, self-contained datalogger with an own power supply, a keyboard and a little display for programming and checking purpose are built in. It has analog input, pulse count, and switched excitation output channels. The 21X's power supply consists of eight alkaline "D" cell batteries. The datalogger stays in a waterproof enclosure in a rock recess besides the scafolding pole, where most of the sensors are mounted. The measuring interval of each sensor is 10 seconds.

C.0.2 Data capture, processing and transfer Climatological measurements, made either in a short time interval or daily, con­ tain information representative of the whole period. The data sampling rate and averaging periods are set to match the atmospheric variability of the individual parameters. Means, samples, maxima and minima are obtained rom the basic values over the selected time period. Wind direction is sampled once in the time interval because of the discontinuity between 0 and 360 degrees. The measuring interval of each sensor is 10 seconds. With the exception of the winddirection mean values of several minutes of all these measurements are stored in the datalogger and in storage modules (SM 192) of Campbell Scientiic. Additionally, the daily extreme values of the temperature and the maximum windspeed are recorded. Two storage modules are diferently switched as ring stor­ age module and as ill-and-stop module in order to obtain long interval measrements for at least 6 months.

159 Appendix C. Meteorological station 160

Element Instrument Type Producer Accuracy Height (m) Temperature HMP35AC Vaisala 0.1-1.0° 2 Humidity HMP35AC Vaisala 1% 2 Winddirection W200P Windvane Vector Instrument +-20 2.3 Winds peed AlOOR Anemometer Vector Instrument 2.2 Air-pressure PTB101B Vaisala 0 Solar radiation Pyranometer Eppley 1.1 Precipitation R102 Tipping bucket Munro 0.2 mm 0 Table C.1: Meteorological sensors at Moreno climate station; the height speciies the sensor mounting height above ground.

Kipp & onen � Eppley

-� g

. 8 . 8 . 8 � . 8 . 8 . 8 . 8 - ! �

Figure C-1: Comparison of radiation measurements with 2 diferent pyranometers. The daily cycles 60of radiation refer to the average of the period from 6 March 1999 until 22 March 195099 . N� 40 " C.0.3 Sensors - 30 Up to seven sensors described in table C.1 have been installed. � 20 The HMP35ACc temperature and relative humidity probe contains a capacitive 0 relative humidity10 sensor and a precision thermistor. The W200P Windvane incorporates a 358° micro-torque potentiometer. An alu­ 0 minium in and the.. pot..entiom..eter housed" ..in a cyli.. nder.. of st..ainless steel ensures 0 i � ; 0 0 ; i ; : 0 operation in regions with heay turbulent winds. The Eppley radiometer0 0 Seria0 l Nr. 019 981F0 3 was0 calibra0 ted 0by the producer with ( ) i a sensitivity of 10.92 w-6 per wm-2; the calibration procedureN N was checked at v the site of the Moreno Base Station with a CM21 Kipp Zonen pyranometer Serial & ( Nr.: 950232) from 6 March 1999 until 22 March 1999 (igure C-1) . The period of comparison was characterized by a high mean radiation and a maximum value of 880 Wm-2• The mean deviation of the 2 instruments was 7.9 Wm-2; the Kipp & Zonen instrument showed in general slightly higher radiation values. Appendix D

Seismograph

Ice thicknesses of Moreno Glacier were determined by using the seismic-relection method at two proile lines. Seismic signals were picked up using two geophone­ strings with twelve geophones at a spacing of 10 meters. The geophones used (Sensor SM-4) were vertical geophones with a natural frequency of 10 Hz. Recording was carried out with a 24-channel "StrataView R24" , a light-weight and rugged digital seismic data acquisition system. Operating the " StrataView R24" is easy and seismic recordings can be made in a quite a short time by pre-programmed system-setups. The system's software ofers a variety of tools for controlling data quality and also for basic data processing already in the ield (automatic gain control AGC, iltering, velocity analysis). Data are stored on hard disk and optionally on loppy disk. Recorded shots can also be viewed on paper printouts. Some technical speciications of the "StrataView R24" are listed in table D .1. The StrataView is portable and has a wide bandwidth. After digital conversion of the geophone signals, data is streamed into digital signal processing chips (DSP's). The specialized CPU's perform low cut, high cut, notch iltering and correlation in a fraction of the time.

Speciications: Number of Channels 24 channels (12V portable module) A/D Conversion 18 bit at 32 kHz sample frequency Dynamic Range 104 dB measured at 0.25 msec, 2 to 2000 Hz Bandwidth 2.0 to 16 kHz rigger Accuracy 1 microsecond Sample Intervals 31.25, 62.5, 125, 250, 500, 1000 or 2000 microsec. Anti-alias ilters -3 dB corner frequency at 60% of Nyquist frequency corner frequency automatically selected Record Lengths 24000 samples per channel Table D.1: Technical details of 11STRATAVIEW11 series R seismograph.

161 Appendix E

Lake depth measurements

Bathymetric measurements were carried out in Brazo Sur on 28 March 1996 and in Canal de los Tempanos on 1 April 1997 (section 4.6) . Table E.1 shows the technical speciications of the echo sounding instrument, the measured positions and depths are included in tables E.2 to E.5.

Speciications Manufacturer Echopilot Ltd., Hampshire, United Kingdom Type Super Maxi Depth Depth Range 0.8 m-199 m Displayed Depth Steps 0.2 m, 1.0 m for depths over 20 m Accuracy of Display ±2% Voltage 12/24 v rigger Power 50 W Frequency 150 kHz ± 3% Table E.1: Technical speciications of the echo sounding instrument used to measure depths in Lago Argentino.

162 depth measurements 163 AppendJ: 8. Lake

Latitude Longitude Depth Latitude Longitude Depth 50d29'06. 13" 73d02'36.06"W 99.0 50d29'16.96"S 73d03'02.41"W 6.0 s 50d29'06.54" 73d02'35.87"W 95.0 50d29'15.55"S 73d02'58.14"W 29.0 s 50d29'08. 75" 73d02'36.42"W 96.0 50d29'13.90" S 73d02'55.00"W 35.0 s 50d29'09.00" 73d02'36.40"W 96.0 50d29'12.42"S 73d02'51.43"W 66.0 s 50d29' 11.42" 73d02'36.99"W 95.0 50d29'10.76"S 73d02'47.78"W 90.0 s 50d29'13.16"S 73d02'37.76"W 86.0 50d29'09.33" 73d02'44.15"W 94.0 s 50d29'14.38"S 73d02'38.14"W 74.0 50d29'07. 70" 73d02'41.10"W 94.0 s 50d29' 16.68" 73d02'39.13"W 57.0 50d29'05.97"S 73d02'38.03"W 91.0 s 50d29'18.57"S 73d02'40.10"W 42.0 50d29'02.98" 73d02'35.98"W 63.0 s 50d29'17.10"S 73d02'33.69"W 16.9 50d29'00.03" 73d02'35.99"W 82.0 s 50d29'12.98"S 73d02'31.58"W 45.0 50d28'58.15" S 73d02'37.77"W 100.0 50d29'08.45" 73d02'31.59"W 96.0 50d28'57.70" 73d02'39.84"W 53.0 s s 50d29'04.54" 73d02'33.14"W 90.0 50d28'57.51"S 73d02'34.94"W 97.0 s 50d29'02.12"S 73d02'34.54"W 61.0 50d28'57.16" S 73d02'32.23"W 84.0 50d29'01.45" 73d02'35.40"W 67.0 50d28' 55.92" 73d02'30.24"W 78.0 s s 50d28'58.64" 73d02'37.39"W 96.0 50d28'54.49"S 73d02'27.45"W 69.0 s 50d28'58.73"S 73d02'40.40"W 92.0 50d28'57.03" 73d02'18.72"W 95.0 s 50d28'59.45"S 73d02'43.07"W 61.0 50d28'57.96" 73d02'21.27" 92.0 s w 50d29'01.16"S 73d02'46.93"W 64.0 50d28'59.28"S 73d02'23.15"W 91.0 50d29'02.56" 73d02'50.52"W 56.0 50d29'01.15"S 73d02'20.73"W 102.0 s 50d29'05.05" 73d02'53.47"W 46.0 50d29'03.69" 73d02'20.31"W 104.0 s s 50d29'08.46" 73d02' 56.82" 61.0 50d29'05.41"S 73d02'21.35"W 105.0 s w 50d29'10.03"S 73d03'00.27"W 51.0 50d29'05.92"S 73d02'18.14" W 108.0 50d29'11.66"S 73d03'03.39"W 20.0 50d29'04.08" 73d02'15.96"W 107.0 s 50d29'13.67"S 73d03'05.86"W 18.4 50d29'01.27" 73d02'15.52"W 107.0 s 50d29'14.99" 73d03'08.07"W 14.2 50d28'58.89" 73d02'15.79"W 104.0 s s 50d29'16.47"S 73d03'06.03"W 4.5 50d28' 56 .40" 73d02' 16.29" 91.0 s w Table E.2: GPS positions and depths resulting from echosounding in the Canal de los Tempanos on 28 March 1996. Appendix E. Lake depth measurements 164

Latitude Longitude Depth Latitude Longitude Depth 50d28'54.83"S 73d02'17.34"W 72.0 50d28'35.66"S 73d02'08.38" 57.0 w 50d28' 53.34" 73d02'19.33"W 58.0 50d28'35.06" 73d02'10.88"W 81.0 s s 50d28'52.51"S 73d02'19. 75"W 55.0 50d28'34.35" 73d02' 13.66" 88.0 s w 50d28'51.43"S 73d02'18.85"W 56.0 50d28'33.66"S 73d02'16.12"W 78.0 50d28 '49. 92" 73d02' 18.46"W 54.0 50d28'33.68" 73d02'17.78"W 84.0 s s 50d28'48.65"S 73d02'18.01"W 53.0 50d28'33.35"S 73d02'18.88"W 94.0 50d28' 46. 71" 73d02'18.17"W 54.0 50d28'35.34" S 73d02'21.39"W 69.0 s 50d28 '45.59" 73d02'18.56"W 56.0 50d28'37.48"S 73d02'23.25"W 84.0 s 50d28'44.56"S 73d02'18.74"W 62.0 50d28'39.41" 73d02'22.51"vV 78.0 s 50d28'43.06"S 73d02'19.71"W 65.0 50d28' 40.82" 73d02'19.89"W 82.0 s 50d28'41.81"S 73d02'20.74"W 76.0 50d28' 41.65" 73d02' 16.57" 80.0 s w 50d28'40.21"S 73d02'21.52"W 79.0 50d28 '42 .42" 73d02' 13.65" 79.0 s w 50d28'38.97"S 73d02'22.28"W 87.0 50d28' 43.08" 73d02'10.96"W 78.0 s 50d28'37.14"S 73d02'21.51"W 88.0 50d28' 43.59" 73d02'07.51"W 80.0 s 50d28'36.06"S 73d02'21.51"W 69.0 50d28 '44.46" 73d02'03.36" 67.0 s w 50d28'34.26"S 73d02'20.92"W 83.0 50d28'46.05"S 73d02'03.04" 76.0 w 50d28'32.97"S 73d02'20.19"W 81.0 50d28'48.93"S 73d02'04.33"W 101.0 50d28'31.63"S 73d02'19.32"W 86.0 50d28'51.34" 73d02'05.57" 105.0 s w 50d28'30.04"S 73d02'17.81"W 83.0 50d28'53.91"S 73d02'06.40"W 107.0 50d28'29.17" 73d02' 16.82" 79.0 50d28'56.57" 73d02'07.53"W 109.0 s w s 50d28'27.95"S 73d02'15.42"W 72.0 50d28'59.13"S 73d02'08.69" 110.0 w 50d28'26.68" 73d02'13.98"W 56.0 50d29'01.22" 73d02'10.27"W 110.0 s s 50d28'25.84" 73d02'13.04"W 37.0 50d29'02.89" 73d02'11.12"W 110.0 s s 50d28'26.20" 73d02' 12.12" 28.0 50d29'05.26" 73d02'12.01"W 110.0 s w s 50d28'28.22" 73d02' 12.42" 67.0 50d29'07.94"S 73d02'11.72"W 100.0 s w 50d28'29.87" 73d02'11.89"W 84.0 50d29'10.32" 73d02'11.93"W 68.0 s s 50d28'32.46" 73d02'11.58"W 83.0 50d29'10.27"S 73d02'11.49"W 70.0 s 50d28'34.08" 73d02'11.38"W 83.0 50d29'10.25"S 73d02'11.28"W 72.0 s 50d28'36.06" 73d02'11.74"W 84.0 50d29'10.95"S 73d02'08.81" vv 71.0 s 50d28'38.89"S 73d02'11.84"W 77.0 50d29'16.05"S 73d02'48.37"W 61.0 50d28' 40.53" 73d02'11.97"W 75.0 50d29'16.28" S 73d02'50.30"W 47.0 s 50d28'41.93"S 73d02' 11.76" 80.0 50d29'17.33" S 73d02'51.75"W 36.0 w 50d28'43.90"S 73d02'11.70"W 80.0 50d29'19.56" S 73d02'53.55"vV 27.0 50d28'44.33"S 73d02'08.60"W 90.0 50d29'21.91"S 73d02'53.51"W 12.3 50d28' 43.69" 73d02'05.09"W 78.0 50d29'22. 77" 73d02'56. 76"W 13.1 s s 50d28' 43.31" 73d02'02.56" 41.0 50d29'20.40" 73d02'58.41"W 15.8 s w s 50d28' 41.19" 73d02'03.12"W 28.0 50d29'17.26" 73d02'59.25"W 21.0 s s 50d28'38.80" 73d02'04.76"W 17.2 50d29'14.89"S 73d02'59.49"W 15.0 s 50d28'36.85" 73d02'06.26"W 16.0 50d29'15.15"S 73d02'55.95"W 38.0 s Table E.3: GPS positions and depths resulting from echosounding in the Canal de los Tempanos on 28 March 1996. Appendix E. Lake depth measurements 165

Latitude Longitude Depth Latitude Longitude Depth 50d28'04.82" 73d02'06.85"W 13.0 50d27'31.59" 73d03'53.31"W 10.5 s s 50d28'01 .65" 73d02'08.88" 49.0 50d27'31.83" 73d03'53.50"W 9.0 s w s 50d27'59.58"S 73d02'12.14"W 77.0 50d27'35.91"S 73d03'38.11"W 73.0 50d28'00.37" 73d02' 15.35" 80.0 50d27'36.57"S 73d03'35.08"W 85.0 s w 50d27'59.69"S 73d02'18.19"W 94.0 50d27'36.83" 73d03'26.68"W 54.0 s 50d27' 58.85" 73d02'20.72"W 113.0 50d27'36.86" 73d03'23.51"W 92.0 s s 50d27' 58.19" 73d02'23.98"W 134.0 50d27' 36. 97" 73d03'19.69"W 110.0 s s 50d27'57.35"S 73d02'26.42"W 130.0 50d27'37.17"S 73d03'17.17"W 131.0 50l27'55.66"S 73d02'29.73"W 140.0 50d27'38.85" 73d03'13.71" W 138.0 s 50l27'54.36" 73d02'31.89"W 137.0 50d27'40 . 44" 73d03'11.29"W 126.0 s s 50l27' 52.86" 73d02'34.37"W 138.0 50d27'42. 07" 73d03'08.29"W 119.0 s s 50d27'51.41" 73d02'35.86"W 141.0 50d27' 43.33" 73d03'05.52"W 121.0 s s 50d27' 50.76" 73d02'27.01"W 156.0 50d27' 44.42" 73d03'03.22"W 88.0 s s 50d27'49.13"S 73d02'20.71"W 158.0 50d27' 45. 64" 73d03'00.84"W 124.0 s 50d27' 48.19" 73d02'25.03"W 161.0 50d27' 4 7. 23" 73d02'57.02"W 134.0 s s 50d27'4 7.77" 73d02'28.22"W 159.0 50d27' 48.66" 73d02'54.63"W 129.0 s s 50d27' 48.18" 73d02'30.65"W 158.0 50d27'50.04"S 73d02'51.06"W 122.0 s 50d27' 48.18" 73d02'34.16"W 156.0 50d27' 51. 23" 73d02'47.25"W 107.0 s s 50d27'48.87"S 73d02'38.21"W 155.0 50d27'52.27"S 73d02'43.64"W 107.0 50d27' 50.08" 73d02'40.42"W 144.0 50d27'50.63" 73d02'40.98"W 123.0 s s 50d27'51.38"S 73d02'43.31"W 118.0 50d27'49.36"S 73d02'38.44"W 154.0 50d27'50.15" S 73d02'46.11"W 117.0 50d27'49.07"S 73d02'35.41"W 156.0 50d27' 48. 84" 73l02'49.00"W 135.0 50d27' 48. 79" 73d02'31.49"W 156.0 s s 50d27' 4 7. 72" 73l02'51.76"W 145.0 50d27' 48.63" 73d02'27.11"W 160.0 s s 50d27'4 7 .14" 73d02'55.30"W 140.0 50d27' 48.16" 73d02'23.91"W 162.0 s s 50d27'45.98"S 73d02'58.95"W 133.0 50d27' 48.14" 73d02'19.90"W 161.0 s 50d27'44.51"S 73d03'02.74"W 127.0 50d27'47.52"S 73d02'15.32"W 164.0 50d27' 42.88" 73d03'05.26"W 86.0 50d27' 4 7. 90" 73d02'11.87" W 142.0 s s 50d27'41.04"S 73d03'07.77"W 131.0 50l27' 48.31" 73d02'08.55"W 114.0 s 50d27'39.87" 73d03'10.72"W 127.0 50d27'49.03" S 73d02'07.18"W 57.0 s 50d27'38. 73" 73d03'14.29"W 130.0 50d27'50.32" 73d02'08.29" 38.0 s s w 50d27'33.25"S 73d03'16.80"W 157.0 50d27' 51. 77" 73d02'09.65" 74.0 s w 50d27'33.95"S 73d03'20.63"W 151.0 50d27'53.60" 73d02'10.96"W 93.0 s 50d27'33.94"S 73d03'24.68"W 132.0 50d27' 55. 20" 73d02'11.85"W 95.0 s 50d27'33.52" 73d03'27.09"W 118.0 50d27'56.67"S 73d02'12.36"W 100.0 s 50d27'32. 70" 73d03'29.22"W 113.0 50l27' 58.14" 73d02'12.98"W 91.0 s s 50d27'30.95"S 73d03'30.94"W 130.0 50d27'59.52"S 73d02'13.93"W 89.0 50d27'31.93"S 73d03'51.87"W 38.0 50d28'00.82" 73d02'15.18"W 84.0 s 50d27'31.80"S 73d03'52.22"W 16.0 50d28'02.16"S 73d02'16.55"W 81.0 50d27'31.60" 73d03'52.36"W 12.5 50d28'02.57"S 73d02'17. 10"W 90.0 s Table E.4: GPS positions and depths resulting from echosounding in Brazo Sur on 1 April 1997. E. Lake depth measurements 166 Appt!nd-i::

Latitude Longitude Depth 50d28'02.64" 73d02'18.07"W 92.0 s 50d28'01.87" 73d02'20.57" W 102.0 s 50d28'01.23" 73d02'21.93"W 106.0 s 50d28'00. 77" 73d02'23.45" 110.0 s w 50d27'59.96"S 73d02'24.44"W 110.0 50d27'58.7l"S 73d02'23.23" 120.0 w 50d27'56.36"S 73d02'18.67"W 136.0 50d27'54.47" S 73d02'16.00"W 131.0 50d27'52.06" S 73d02'12.49"W 118.0 50d27'49.47" S 73d02'08. 91" 94.0 w 50d27' 4 7. 77" 73d02'05.61"W 48.0 s Table E.5: GPS positions and depths resulting from echosounding in Brazo Sr on 1 April 1997. Curriculum Vitae

Martin Stuefer

13 February 1964 in Innsbruck nom Position Project scientist at the Institut fir Meteorologie und Geophysik, Crrent Universitat Innsbruck

Education

1970 1974 Volksschule Michael-Gaismayr-Str., Innsbruck.

1974 1982 Bundesrealgymnsium, Adolf Pichler Platz, Innsbruck (mathe­ matical department). 17 Jun 1982 School leaving exam (Matua) at the Bundesrealgymnasium, Adolf Pichler Platz, Innsbruck. 1 Oct 1982 to Military service. 31 Mai 1983 6 Mar 1984 Enrollment at the Universitat Innsbruck, Studienrichtung Erdwis­ senschaften. 3 Oct 1984 Beginning of the study of Meteorology and Geophysics at the Uni­ versitat Innsbruck. 1986 to 1989 Study at the Institute of Mathematics at the Universitat Inns­ Oil bruck. 27 Jan 1995 Final examination at the institute of Meteorology and Geophysics, Title of the Diploma Thesis: Der unterschiedliche Eintuss des Kli­

to mas auf die Gletscher der Otztaler Alpen und der Silvrettagruppe. 25 Mar 1995 Graduation to Magister rerum natualium at the Universitat Inns­ to bruck.

Scientiic Experience

1990 to 1996 Wo rk contract: Alpinwetterdienst des Alpenvereins at the We t­ terdienststelle Innsbruck, ZA fur Meteorologie, Austria.

1 Jan to Wo rk contract within the fr amework of the International Geo­ 29 Feb 1995 sphere Biosphere Project on the topic: The Temperature and Pre­ cipitation Data at 15 Climate Stations in the Silvretta Region.

1 Jun to Wo rk contracts on the topic: Preparatoring analysis of ERS­ 30 Sep 1995 SAR data for the ERS- 1/ERS-2 Tandem Mission and on the preparation of ieldwork within the framework of SIR-C/ERS-SAR projects of the Austrian Scientiic Fund.

1 Aug 1995 to Research scientist (Projektassistent) within the FWF project P 31 Jul 1997 10709-GEO with the title 11lnvestigations of Glacier Behaviour on the Southern Patagonian Iceield and on the Antarctic Peninsula bsed on Space borne SAR Images and Field Mesurements at the Institut fur Meteorologie und Geophysik, Universitat Innsbruck.

1 Aug 1997 to Research scientist (Projektassistent) within the framework of the 31 Mar 1998 Hydalp project at the Institut fir Meteorologie und Geophysik, Universitat Innsbruck.

11 1 May 1998 to Research scientist (Projektassistent) within the framework of the .Tun 1998 Hydalp project at the Institut fir Meteorologie und Geophysik, 30 Universitat Innsbruck. 1 .Jul 1998 to Research scientist (Projektassistent) within the FWF project P 1999 12923-GEO with the title "Disintegration of Northern Larsen Ice 30 Sep Shelf, Antarctic Peninsula, and Ice Retreat in Patagonia" at the Institut fir Meteorologie und Geophysik, Universitat Innsbruck. since 1 Oct 1999 Research scientist (Pojektassistent) within the framework of the Hydalp project at the Institut fir Meteorologie und Geophysik, Universitat Innsbruck.

Research visits and scientiic expeditions

Apr / Okt 1994 Field-campaign at the test site Hintereisferner and Kesselwand­ ferner in the Otztal during SIR-C/ X-SAR psses. Aug 1995 to Three ield-campaigns at the test site Hintereisferner and Kessel­ 1996 wandferner in the Otztal during ERS-1 / ERS-2 Tandem Missions. Feb Nov/Dec 1995, Field-expeditions to Moreno Glacier, South Patagonian Iceield. Mar/ Apr 1996, The ield activities in Patagonia were carried out within the FWF Nov/Dec 1996, projects P 10709-GEO and P 12923-GEO. Mar/ Apr 1997 , Mar/ Apr 1998 and Mar 1999