Methodology:

Adventdalen is a medium-sized (513 km2) catchment in continuous permafrost zone of central Spitsbergen. It is an important study area, but the river is highly unsuitable for monitoring runoff on account of the high sediment yield and extremely unstable channel sections in the large delta. This necessitates the estimation of runoff volume through the application of a suitable model. Daily runoff was therefore simulated using daily precipitation and temperature data series (1991-2016) for Svalbard Lufthavn (Longyearbyen airport) from the Norwegian Meteorological Institute's eklima.met.no website. Approximately 37% of the precipitation series had missing values, which were treated as zeros. After Killingtveit [2004], the precipitation record required an undercatch adjustment of 1.65 for snowfall and 1.15 for rainfall. An alternative approach involving modelling precipitation using a regional climate model and statistical downscaling yielded no improvement and so is not described here.

Runoff observations were simulated at a daily time step using the well-known HBV package (https://www.geo.uzh.ch/en/units/h2k/Services/HBV-Model.html). Calibration and validation was undertaken using input data from De Geerdalen, spanning 1st August 1991-31st December 2001. These data were combined with observed temperature and precipitation (as from above) data for the same period, in an input .txt file. De Geerdalen is a long-term monitoring site adjacent to Adventdalen, where data are collected from a semi-permanent installation near the coast by the Norwegian Water and Energy Administration (see Sund, 2008).

Although monitoring runoff volume was not successful, basic water quality parameters: electrical conductivity (in µS cm-1) and water temperature (in oC) were successfully monitored using a Campbell Scientific Cr800 data logger with a Cr247 sensor. Daily averages of hourly values are included in this file.

HBV calibration using De Geerdalen is explained in further detail in the README file.

References:

Killingtveit A., 2004, Water balance studies in two catchments in Spitsbergen, Svalbard, Northern Research Basins Water Balance, Proceedings of a workshop held at Victoria, Canada, March 2004), IAHS Publication Number: 290, 2004.

Sund, M. 2008. Polar Hydrology. NVE Report No. 2, 2008. Oslo, Norway. http://publikasjoner.nve.no/report/2008/report2008_02.pdf

Data quality:

The calibration of De Geerdalen resulted in a Reff = 0.8454(Nash-Sutcliffe efficiency equation) with a mean difference between the observed and simulated discharge of -17 mm/year. The results of the De Geerdalen GAP optimisation can be found in the README file.

Note: there is more uncertainty over the winter rainfall events, whose impact were very marked in 2016. Therefore runoff data are included as two series, one with and one without runoff during the winter season.