Last modified: 2011-02-14
Abstract
Predicting hydrological variables in ungauged catchments has been singled out as one of the major issues in the hydrological sciences. Mathematical methods which also include hydrological modelling are widely used in hydrology nowadays. Hydrological models can be used as useful tools to expand existing or generate completely new data series of the river streamflow. This kind of simulation of hydrological processes can be worthy solution to generate hydrological data in ungauged basins where actual observations are not made. Simulated data series after can be used for operational and planning purposes, climate change impact on hydrological regime estimation and land use change impact on hydrological regime as well. During last years the study of hydrological responses to global climate change and the assessment of water resources have placed importance in Nordic countries and in Latvia as well. Results obtained from rainfall-runoff model applied to gauged basin can be regionalized to ungauged basins.
The proxy-basin method was used for regionalization. This method involves cross-checking parameters transferability over two gauged basins in the interested region and then directly applying them to ungauged basins in the same region.
In this research conceptual HBV model is applied to simulate hydrological processes within the watershed of Gauja River. HBV model is created by Sten Bergström and was first applied to assist the hydropower stations in Sweden (IHMS manual, 2008).
The Gauja river is located in north eastern part of Latvia and its basin covers nearly 8900 km2. In this study the River Gauja watershed is divided in six sub-basins where three different stations of the hydrological observations are on Gauja River: Gauja-Sigulda, Gauja-Valmiera and Gauja-Tilderi and three hydrological stations are located on Gauja tributaries Amata-Melturi, Tirza-Lejasciems and Vaidava-Ape.
The research executes the calibration of the HBV model for all six subbbasins and aims to reach the highest possible correlation between observed and simulated river discharge. Statistical criterion R2 is used as a measure of the correlation (Nash and Sutcliffe, 1970). The HBV model parameter set obtained in the process of calibration are used to simulate the stream flow in other subbasins considered as “ungauged”. This will aid to evaluate the use of these parameters for simulating hydrological processes in other sub-basins of the Gauja watershed.
The necessary meteorological and hydrological input data for the model were obtained from the Latvian Environment, Geology and Meteorology Centre. These input data are: average daily temperature of air (obtained from five stations of the meteorological observations), the sum of daily precipitation (obtained from eight stations of the precipitation observations), and the values of potential monthly evaporation.
A ten year period from January 1st, 1980 till December 31st, 1989 is used to calibrate the HBV. The validation of the model was made for the following five year period from January 1st, 1990 till December 31st, 1994.