Water balance studies in two catchments on Spitsbergen, Svalbard

Size: px

Start display at page:

Download "Water balance studies in two catchments on Spitsbergen, Svalbard"

Dylan Walsh
6 years ago
Views:

1 120 Northern Research Basins Water Balance (Proceedings of a workshop held at Victoria, Canada, March 2004). IAHS Publ. 290, 2004 Water balance studies in two catchments on Spitsbergen, Svalbard ÀNUND KILLINGTVEIT Department of Hydraulic and Environmental Engineering, Norwegian University of Science and Technology, N-7491 Trondheim, Norway aanund.killingtveit(5intnu.no Abstract Hydrological studies at Svalbard have been concentrated in two catchments in particular, the Bayelva catchment near Ny Alesund and De Geerdalen near Longyearbyen. Hydrological processes and water balance in these and some other catchments were monitored and studied in several projects during the 1990s and a summary of the main results were presented in a series of papers in the journal Polar Research in This paper contains a summary of some of these results, supplemented with a description of the catchments. The runoff in Svalbard is dominated by snowmelt and glacial melt. The runoff is usually much higher than observed precipitation, since the precipitation gauges only can catch around 50% of the precipitation due to strong winds, low temperatures, and snow precipitation. Evaporation is low, less than 100 mm year" 1 from glacier-free areas, and probably close to zero from glaciers. Key words Arctic hydrology; precipitation correction; Spitsbergen; Svalbard; water balance BACKGROUND The Intergovernmental Panel on Climate Change concludes that the North Atlantic region is one of the most sensitive on Earth with respect to climate change (IPCC, 2001). There is reason, therefore, to study climate dynamics and climate change and their effects on Arctic ecosystems. The hydrological processes constitute an important link between climate and the effect on ecosystems, and therefore hydrological studies have become increasingly important in studies of climate change, particularly in the Arctic. In Norway, a research programme on Arctic hydrology was initiated around 1990, with its main focus on Svalbard and Spitsbergen (the largest island in Svalbard) specifically. In 2001, about 10 years after the start of the Arctic Hydrology Programme, the Norwegian Hydrological Council arranged a workshop in Longyearbyen, Svalbard, in order to discuss and summarize the status of hydrological research in Svalbard. Here it was decided to write a series of overview papers assembling the state-of-the-art of hydrological knowledge in Svalbard, as a contribution to the Arctic Climate Impact Assessment (ACIA). A total of six papers, covering climate, snow, glaciers, water balance, sediment transport and permafrost, were published in the journal Polar Research in This paper contains a summary of some of the most important findings in these six papers (Bogen & Bonsnes, 2003; Forland etal, 2003; Hagen etal, 2003; Humlum etal, 2003; Killingtveit etal, 2003; Winther etal, 2003). In addition some new data for runoff and climate have been included here, in order to present an updated report on hydrological data and water balance calculations for two

2 Water balance studies in two catchments on Spitsbergen, Svalbard 121 catchments in Svalbard. These two, Bayelva and De Geerdalen, are the two catchments with the longest records and most complete hydrological observations in Svalbard. The Bayelva catchment is located close to the western coast; De Geerdalen is more inland, in the central part of Spitsbergen. All water balance calculations are done for hydrological years. In Svalbard the hydrological year starts in September and goes to August of the next year. By starting with 1 September, the difference in the storage in the basin from year to year is minimized, since snowmelt from the previous winter is finished, and new snow has not yet started to accumulate. ABOUT SVALBARD The Svalbard archipelago is a group of islands located north of Norway and east of Greenland, between the Arctic Ocean, Barents Sea, Greenland Sea, and Norwegian Sea, between 76 and 81 N. Spitsbergen is the name of the largest island in Svalbard. The Svalbard archipelago covers an area of km 2, of which about 60% or km 2 is covered by glaciers (Hagen etal, 2003). The rest, about km 2, has seasonal snow cover, all of it with permafrost. In Svalbard, permafrost depth is typically from 100 m in major valley bottoms and up to m in the high mountains (Humlum et al, 2003). Today, there are four main settlements on Spitsbergen (Fig. 1): Ny Alesund, Longyearbyen, Barentsburg and Sveagreuva. All these settlements were founded as coal mining towns, but today only Sveagruva and Barentsburg are solely dependent on the mining industry. Ny-Alesund has developed into a large-scale research facility and DE GEERDALEN Fig. 1 Svalbard with location of research catchments Bayelva and De Geerdalen.

3 122 Ânund Killingtveit has between 40 and 100 inhabitants, depending on the season. The Bayelva catchment is located close to Ny Alesund, and the operation of gauging stations is based on infrastructure in Ny Âlesund. With a population of more than 1700, Longyearbyen depends on tourism, research and education, in addition to mining. The De Geerdalen catchment is located approximately 20 km from Longyearbyen, and fieldwork and operation of gauging stations can be based on support from Longyearbyen. Ny- Âlesund, Longyearbyen, and Sveagruva are Norwegian settlements, while Barentsburg, with approximately 1000 inhabitants, is Russian (Humlum et al, 2003). CLIMATE Svalbard is located in the main transport pathway for air masses into the Arctic Basin, and the climate is characterized by cool summers and cold winters. The North Atlantic current flows along the west and north coasts of Spitsbergen, keeping water open and navigable most of the year. A major climatic control, especially for winter conditions, is the Siberian High, an intense, cold anticyclone that forms over eastern Siberia in winter. When the Siberian high extends far to the west, it covers Russia and part of Europe, creating a strong southerly airflow over the Nordic seas, causing advection of warm air to the Svalbard region. During such events, heavy snowfall and even snowmelt may occur in Svalbard in the middle of the winter (Humlum et al, 2003). Air temperature can vary from less than -40 C to more than +20 C, with an average annual temperature at sea level of around -5 C in Ny Âlesund and slightly colder at Longyearbyen (Svalbard Airport). There are pronounced long-term fluctuations in Arctic climate and air temperature, but no significant long-term trends. From the start of observations in 1910 there was a positive trend up to the late 1930s, a decrease to the 1960s, and a new increase to present temperatures. Despite strong warming in the decades since the mid-1960s, the air temperature today (2000) is still below the warmest two decades in the 1930s and 1940s (Forland & Hanssen-Bauer, 2003). HYDROLOGICAL RESEARCH CATCHMENTS The first runoff measurements in Svalbard were started in Bayelva in These measurements were stopped in 1978, but started again in 1990 as part of an initiative taken by the Norwegian National Hydrological Committee in In 1990 a second runoff station was started in De Geerdalen, about 20 km northeast of Longyearbyen. Since then, both these stations have been in continuous operation, and the two catchments have since been the main hydrological research catchments in Svalbard. The data in Tables 1 and 2 display long term consideration of all water balance components. WATER BALANCE ELEMENTS The main elements of the water balance in an Arctic catchment are precipitation, runoff, évapotranspiration, and storage change. In order to compute the water balance, all elements should be measured or estimated by independent methods.

4 Water balance studies in two catchments on Spitsbergen, Svalbard 123 Table 1 Main data for the Bayelva catchment. Name Location Bayelva, Spitsbergen, Svalbard 'N, 11 56'E Area 30.9 km 2 Permafrost extent Continuous Soils description Moraines, riverbed, tundra, rock Vegetation Uniform lichen cover with patches of rock sedge (Carex rupestris) and mountain avens (Dryas octopetela). There are no trees or tall shrubs. At higher elevations mostly gravel, stones and rock. Climate High Arctic, mean annual temperature -6.3 C Topography It consists of a flat river plain in the centre, and steep and tall mountains from southeast to southwest, the Zeppelin and the Schetelig mountains. Elevation ranges from 4 up to 742 m a.s.l., average elevation is 265 m a.s.l. Glaciers 55% of the catchment is covered by glaciers Period of record Runoff and 1990 to date. Precipitation and climate from Sediment transport and 1989 to date. Other A permanent research station is located in Ny Alesund, and an airport with regular flights. Excellent conditions for field oriented research. Table 2 Main data for the De Geerdalen catchment. Name Location De Geerdalen (The De Geer valley) Spitsbergen, Svalbard ^, 11 19'E Area 79.1 km 2 Permafrost extent Continuous Soils description Moraines, riverbed, tundra, rock Vegetation Uniform lichen cover with patches of rock sedge (Carex rupestris) and mountain avens (Dryas octopetela). There are no trees or tall shrubs. At higher elevations mostly gravel, stones and rock. Climate High Arctic, mean annual temperature -6 C Average measured precipitation 182 mm year" 1 Topography It consists of a river valley in the centre, with mountains on both sides. Elevation ranges from 40 to 987 m.a.s.l., average is 410 m.a.s.l. Glaciers 10% of the catchment is covered by glaciers Period of record Runoff from 1990 to date. Precipitation and climate from 1911 (Svalbard Airport 20 km southeast). Seasonal snow measured since Other Approximately 20 km from Longyearbyen with airport. Runoff In Svalbard, almost all river runoff occurs during the four months from June to September. In the autumn, all rivers freeze up completely, except short reaches of rivers fed by springs or in front of some glaciers. Runoff measurements are difficult to collect, due to ice and snow blocking the river channel at gauging stations, and due to unstable river beds in braided rivers with high rates of sediment transport. The runoff gauging station in Bayelva is shown in Fig. 2. The station is located near a narrow gorge, close to the river outlet into Adventfjorden. The station is equipped with an automatic water level recorder, and sediment sampling equipment for both suspended load and bedload measurements. Some results from the sediment measurements are presented by Bogen & Bonsnes (2003).

5 124 Ânund Killingtveit Fig. 2 Runoff gauging station in Bayelva (Photo 5 September 2000 by Â. Killingtveit). A summary of runoff data for hydrological years (September-August) for the two catchments for the period 1990/1991 to 2001/2002 can be found in Tables 3 and 4. Runoff has been converted to mm per year. The average monthly specific runoff for the two catchments is shown in Fig. 3. Average seasonal runoff distribution ( ) Fig. 3 Average specific runoff distribution in Bayelva and De Geerdalen

6 Water balance studies in two catchments on Spitsbergen, Svalbard 125 Table 3 Water balance (mm year" 1 ) for the Bayelva catchment. Pa A G Q E (mm) (mm) (mm) (mm) (mm) 1990/ / / / / / / / / / / Average SD P/. areal precipitation; A G : glacier melt; Q: runoff; E: evaporation; e: error term. Table 4 Water balance (mm year" 1 ) for the De Geerdalen catchment. Pa A 0 0 E s (mm) (mm) (mm) (mm) (mm) 1990/ / / / / / / / / / / Average SD P A : areal precipitation; A 0 : glacier melt; Q: runoff; E: evaporation; s: error term. Almost all runoff occurs between June and September. The specific runoff is almost the same in both catchments during June, when snowmelt is the dominating process. After snowmelt, the glacial melt, together with precipitation, produces a much higher specific runoff in Bayelva than in De Geerdalen, due to a larger percentage of glaciers in Bayelva. Precipitation Precipitation is currently measured at six manual weather stations in Svalbard. One of these is located in Ny Alesund, close to the Bayelva catchment; another is at Svalbard

7 126 Ânund Killingtveit Airport, about 20 km southwest of De Geerdalen. The combination of dry snow, high wind speed and open tundra increases measuring errors for precipitation at most stations in the Arctic, and true precipitation for all stations at Svalbard is probably 50% higher than measured (Forland & Hanssen-Bauer, 2003). In addition, there is the problem of non-representative location for the precipitation stations. All precipitation stations are located in settlements close to the sea, and at elevations close to sea level. It is well known that precipitation usually increases with increasing elevation, and this has also been verified at Svalbard. Studies using precipitation gauges (Forland & Hanssen-Bauer, 2003), snow measurements (Humlum etal, 2003; Killingtveit etal, 2003; Winther etal, 2003) and mass balance of glaciers (Hagen et al, 2003), all confirm that the precipitation gradient is significant, and often of the order of 15-20%>, or even higher at some locations. The gradient is highest along the coast and lower inland (Humlum et al, 2003). Before precipitation data can be used in water balance calculations, it is therefore necessary to make corrections both for catch errors and for elevation gradients. These corrections are explained in detail in Killingtveit et al. (2003). An average correction factor of 1.15 for rainfall and 1.65 for snow precipitation was used for the Ny-Alesund data and slightly higher correction factors, 1.15 and 1.75, were used for the Svalbard Airport data. An average precipitation gradient of 15% per 100 m increase in elevation from sea level was used for the Bayelva catchments, and 20% per 100 m for De Geerdalen. The results (areal precipitation) are shown in Tables 5 and 6. Evaporation Evaporation measurements have been (and are still) very scarce in Svalbard. There are still no regular measurements of evaporation, and estimates must be based on correlation to air temperature or data from other catchments. In Killingtveit et al. (2003) the average annual evaporation from glacier-free catchments close to sea level Table 5 Areal precipitation calculation for Bayelva based on precipitation data from Ny Âlesund (Killingtveit etal, 2003). Hydrological P àp year (mm year" 1 ) (mm year ) E (mm year" ) (mm year ) Average P: observed precipitation; AP C : catch correction; AP E : elevation correction; P A : areal precipitation.

8 Water balance studies in two catchments on Spitsbergen, Svalbard 127 Table 6 Areal precipitation calculation for De Geerdalen based on precipitation data from Ny Alesund (Killingtveit et al, 2003). Hydrological P àpc àp E Pa year (mm year" 1 ) (mm year ) (mm year" ) (mm year ) 1990/ / / / / / / / / / / Average P: observed precipitation; AP C : catch correction; AP E : elevation correction; P A : areal precipitation. was estimated at approximately 100 mm year". Due to the negative temperature lapse rate, both the air temperature and evaporation are reduced at higher elevations, and the average evaporation from non-glaciated areas in the two catchments were estimated at 80 mm year" 1 in Bayelva and 82 mm year" 1 in De Geerdalen. The average annual evaporation of glaciers is assumed to be nil. This assumption may be questioned, but to date no studies have quantified the annual evaporation from glaciers on Svalbard. There are probably some sublimation losses during winter, and some condensation during summer, of about the same magnitude. On average, evaporation from the total catchment was computed to be 35 mm year" 1 in Bayelva and 72 mm year" 1 in De Geerdalen. Evaporation for each hydrological year is shown in Tables 3 and 4, computed from air temperature data, using a method similar to that described in Killingtveit et al (1994). Storage changes If the water balance is computed on an annual basis and for hydrological years, most of the storage terms can be neglected. One remaining storage term of great importance is the change in glacier storage. Time series of terrestrial and aerial photographs show that most Svalbard glaciers have been retreating and thinning since about Small glaciers (<5 km 2 ) and glaciers below 500 m a.m.s.l. seem to have a negative mass balance, while some larger glaciers and glaciers covering higher accumulation areas seem to be closer to equilibrium (Hagen et al, 2003). In Bayelva and De Geerdalen the glaciers are small and located at low elevation, and the average net balance has been negative during most years. The average net balance for the glaciers in Bayelva during the hydrological year 1990/91 to 2000/01 was estimated to be 458 mm year" 1. In De Geerdalen the net balance was estimated to be 550 mm year" 1 in the same period. This makes a very significant contribution to the annual runoff; in Bayelva 24% and in De Geerdalen almost 10% of annual runoff was

9 128 Âmtnd Killingtveit generated from glacial melt. In Tables 3 and 4 the annual change in glacier storage has been converted to mm year" 1 for the whole catchment, by considering the percentage of area covered by glaciers (Killingtveit et ai, 2003). WATER BALANCE The water balance equation for a catchment can be written: PA-QS-QG-E A ±&M=Z where P A is areal precipitation input (mm), Qs is surface (river) runoff from the catchment (mm), Q G is groundwater runoff from the catchment (mm), EA is evaporation from the catchment (mm), AM is changes in water storage within the catchment (mm) and s is an error term (mm). The water balance for each of the two catchments is summarized in Tables 3 and 4. The water balance is based on data for runoff, areal precipitation (Tables 5 and 6), evaporation, and glacier storage change. Groundwater runoff was assumed to be nonexistent, because of the deep permafrost layer. SUMMARY AND CONCLUSIONS The water balance for the two catchments has an average residual term close to zero, so in this sense the water balance measurements are good. But errors in individual years are still large, with positive and negative deviations. This indicates that the individual terms in the water balance are still not known well enough; the largest errors are probably due to insufficient knowledge of precipitation corrections and glacial balance. Even if evaporation is not a dominant factor in the water balance, it would be useful to improve both data collection and computational methods. Both evaporation from bare ground and snow should be studied (Killingtveit et al., 2003). REFERENCES Bogen, J. & Bonsnes, T. E. (2003) Erosion and sediment transport in High Arctic rivers, Svalbard. Polar Res. 22, Forland, E. J. & Hanssen-Bauer, I. (2003) Past and future climate variations in the Norwegian Arctic: overview and novel analyses. Polar Res. 22, Hagen, J. O., Kohler, J., Melvold, K. & Winther, J-G. (2003) Glaciers in Svalbard: mass balance, runoff and freshwater flux. Polar Res. 22, Humlum, O., Instanes, A. & Sollid, J. L. (2003) Permafrost in Svalbard: a review of research history, climatic background and engineering challenges. Polar Res. 22, IPPC (Intergovernmental Panel on Climate Change) (2001) Observed climate variability and change. In: Climate Change 2001: The Scientific Basis: Contribution of Working Group I to the third assessment report of the Intergovernmental Panel on Climate Change, (ed. by J. T. Houghton et al.), Cambridge University Press, Cambridge, UK. Killingtveit, Â., Pettersson, L-E. & Sand, K. (1994) Water balance studies in Spitsbergen, Svalbard. In: Proc. 10th International Research Basins Symposium and Workshop (Spitsbergen, Norway) (ed. by K. Sand & Â. Killingtveit) SINTEF Report 22 A96415, Norwegian Institute of Technology, Trondheim, Norway. Killingtveit, Â., Pettersson, L.-E. & Sand, K. (2003) Water balance investigations in Svalbard. Polar Res. 22, Winther, J-G., Bruland, O., Sand, K., Gerland, S., Maréchal, D., Ivanov, B., Glowacki, P. & Konig, M. (2003) Snow research in Svalbard. Polar Res. 22,

The elevations on the interior plateau generally vary between 300 and 650 meters with

The elevations on the interior plateau generally vary between 300 and 650 meters with 11 2. HYDROLOGICAL SETTING 2.1 Physical Features and Relief Labrador is bounded in the east by the Labrador Sea (Atlantic Ocean), in the west by the watershed divide, and in the south, for the most part,