WP3c: HYDROLOGY

Contents

  1. Introduction
  2. Meteorological data analyses
  3. Hydrological results
      3.1  Changes in mean flow and seasonal distribution
      3.2  Floods
      3.3  Low water conditions
  4. Evaporation estimation for Lake Balaton

1. Introduction

The hydrological impact task of the project is aimed at the production of future hydrological scenarios based on the output of regional climate models.

VITUKI-NHFS and VIDRA (of INHGA) conceptual hydrological models were used to produce long term hydrological series. Mostly the Tisza River Basin (the largest - by drainage basin size - tributary of the Danube) and its subcatchments have been studied with special emphasis on the Upper Tisza and the Mures/Maros rivers. The entire Danube was also covered resulting as the superposition of separate simulations for the Danube down to the Carpathian Basin, the Drau/Drava, and the Tisza, while the River Sava Basin was captured by a simplistic correlation scheme. Out of the lower Danube tributaries only the Arges Basin was covered. (Fig. 1a.)

Figure 1a: Hydrologic study regionsFigure 1a: Hydrologic study regions

top

2. Meteorological data analyses

The raw REMO5.7 ERA40 (1961 – 2000) and REMO5.7 A1B (1951 – 2050) data produced by the Max Planck Institute for Meteorology in Hamburg was further processed to fit the needs of the hydrological models used in CLAVIER. The area of CLAVIER dataset is presented in Figure 1b.

Figure 1b: The area of the CLAVIER datasetFigure 1b: The area of the CLAVIER dataset

For the statistically corrected data a downscaling procedure was performed by VITUKI based on the elevation distribution functions so that the dataset in the needed fine (10 km) spatial resolution could be obtained. An example for the temperature downscale is shown in Figures 2 and 3.

Figure 2: Example of original (0.25 deg. resolution) temperature fieldFigure 2: Example of original (0.25 deg. resolution) temperature field

Figure 3: Example of downscaled (0.1 deg. resolution) temperature fieldFigure 3: Example of downscaled (0.1 deg. resolution) temperature field

The transient model simulations were carried out in the period of 1951 - 2050. The validation was related to the period of 1984 - 2000. The sub-basin temperature and precipitation projections were analysed together with the investigation of the impact on the flow conditions.

Analysing the mean areal precipitation and temperature values, computed based on the A1B scenario (periods 1961-1990 and 2021-2050), for each of the sub-basins within the Tisza River Basin the following projections can be made:

  • A general increase of the mean annual 2m air temperature for all of the subbasins, with 1.4 ÷ 1.6 °C, with smaller changes for the spring period (0.8 ÷ 1.2 °C) and greater changes for the other seasons, especially winter (1.5 ÷ 2 °C);
  • Great spatial variability in tendencies is foreseen for the annual precipitation with a slight increase up to 3.5% in Upper Tisza mountainous catchments and overall decrease in other sub-basins with 3 ÷ 10 %, while definite increase for winter periods (14 ÷17% in Upper Tisza), and general decrease in all other seasons with some exceptions in the highly elevated parts of the Upper Tisza.

Annual and seasonal changes in the temperature and precipitation values for the Tisza River Basin are presented in Fig. 4-8, and Fig. 9-13, respectively.

Figure 4: Difference of the ANNUAL mean temperature values for Tisza catchment
Figure 5: Difference of the SPRING mean temperature values for Tisza catchment
Figure 6: Difference of the SUMMER mean temperature values for Tisza catchment
Figure 7: Difference of the AUTUMN mean temperature values for Tisza catchment
Figure 8: Difference of the WINTER mean temperature values for Tisza catchment

Figure 9: Difference of the ANNUAL precipitation amount for Tisza catchment
Figure 10: Difference of the SPRING precipitation amount for Tisza catchment
Figure 11: Difference of the SUMMER precipitation amount for Tisza catchment
Figure 12: Difference of the AUTUMN precipitation amount for Tisza catchment
Figure 13: Difference of the WINTER precipitation amount for Tisza catchment

From the analysis of the mean areal precipitation and temperature values, computed based on the A1B scenario, for each of the sub-basins within the Mures and Arges River Basins, we can see the following potential climate change:

  • A general increase of the mean annual 2m air temperature for all the basins, with 1.4 – 1.6 °C, with smaller changes for the spring period (1 – 1.2 °C) and greater changes (1.5 – 2 °C) for the other seasons (see Fig. 14-18);
  • The temperature changes are greater with ~ 0.2 °C for the Arges River Basin comparing with the Mures River Basin, for all the seasons (see Fig. 19-23);
  • A general decrease of the annual precipitation with -3 ÷ -5.5 %, but with an increase of 5.5 ÷ 8 % in Mures Basin and 8 ÷ 10 % for Arges Basin in winter, and general decrease in all the sub-basins for the other seasons, going up to – 15 ÷ -25 for some sub-basins in the spring and summer months mainly (Fig. 24-33).

Figure 14: Difference of the ANNUAL mean temperature values for Mures catchment
Figure 15: Difference of the SPRING mean temperature values for Mures catchment
Figure 16: Difference of the SUMMER mean temperature values for Mures catchment
Figure 17: Difference of the AUTUMN mean temperature values for Mures catchment
Figure 18: Difference of the WINTER mean temperature values for Mures catchment

Figure 19: Difference of the ANNUAL mean temperature values for Arges catchment
Figure 20: Difference of the SPRING mean temperature values for Arges catchment
Figure 21: Difference of the SUMMER mean temperature values for Arges catchment
Figure 22: Difference of the AUTUMN mean temperature values for Arges catchment
Figure 23: Difference of the WINTER mean temperature values for Arges catchment

Figure 24: Difference of the ANNUAL precipitation amount for Mures catchment
Figure 25: Difference of the SPRING precipitation amount for Mures catchment
Figure 26: Difference of the SUMMER precipitation amount for Mures catchment
Figure 27: Difference of the AUTUMN precipitation amount for Mures catchment
Figure 28: Difference of the WINTER precipitation amount for Mures catchment

Figure 29: Difference of the ANNUAL precipitation amount for Arges catchment
Figure 30: Difference of the SPRING precipitation amount for Arges catchment
Figure 31: Difference of the SUMMER precipitation amount for Arges catchment
Figure 32: Difference of the AUTUMN precipitation amount for Arges catchment
Figure 33: Difference of the WINTER precipitation amount for Arges catchment

top

3. Hydrological results

REMO5.7-ERA40 (1961 – 2000) and REMO5.7-A1B (1951 – 2050) data produced by the Max Planck Institute for Meteorology, Hamburg were used as climate change scenario. These climate change simulations were first pre-processed to fit the needs of the hydrological models (error corrected/adjusted dataset – by WegCenter, Graz; temporal refinement of corrected data to 6-hour time step – by INHGA, Bucharest, downscaled down to 10 km spatial resolution – by VITUKI, Budapest).

The hydrological simulations were produced in “natural flow” conditions, without taking into account the influence of the reservoirs.

The hydrological model parameters reflect the present day land used – land cover influence on the basins hydrological response.

Model calibration was performed using observed historical data and regional parameters to some adjustments using the ERA40 data.

top

3.1 Changes in mean flow and seasonal distribution

Based on the hydrological simulations, the analysis of the mean seasonal and annual discharges simulated time series show the following impact of the A1B climate change scenario:

In order to estimate the A1B climate change scenario impact on the hydrological regime the variations of the mean monthly simulated discharges for some selected 30 years periods (1961 – 1990 as reference period and 2001 – 2030, 2011 – 2040, 2021 – 2050 as representative periods for the future) were also analysed.

The preliminary results in most cases indicate slight decrease of annual mean flow throughout the region (Fig. 46), with significant spatial variability and even some increase for the high elevation zones in the Upper Tisza sub-catchments (Fig. 47). The decrease of spring runoff is compensated by the flow resulting from thaw during late winter.

Figure 46: Comparison of the mean monthly simulated discharges for the selected 30 years future periods, Tisza - Szolnok
Figure 47: Comparison of the mean monthly simulated discharges for the selected 30 years future periods, Tisza - Tiszabecs

The similar analyses were carried out for some selected representative hydrometrical stations from the Mures and Arges River basins.

The comparison of the mean monthly simulated discharges, for the selected 30 years future periods and the reference period (Figures 48 and 49), indicate the following impacts of the A1B climate change scenario:

  • Mures – Arad (basin area = 27280 km2; mean basin altitude = 618 m)
    • A general decrease tendency of the mean monthly discharges for the March –November period, in the future periods comparing with the reference period;
    • There is a significant decrease indication of the mean monthly discharges for all the selected future periods, for August, September and Octoberp;
    • There is a clear continuous decreasing tendency, as we go in the future, for June, November and December;
    • Comparing with the reference period, the simulated mean monthly discharges variations indicate an increasing only in the winter season months, especially for February and December;
    • The most stable regime is for the mean monthly discharges in January.
  • Arges – Budesti (basin area = 9328 km2; mean basin altitude = 442 m)
    • A general decrease tendency of the mean monthly discharges for the April –October period, in the future periods comparing with the reference period;
    • There is a significant decrease indication of the mean monthly discharges for all the selected future periods, for April, May, July and September;
    • There is a clear continuous decreasing tendency, as we go in the future, for April, May, June, November and December, and a small increasing tendency for August;
    • Comparing with the reference period, the simulated mean monthly discharges variations indicate an increasing in the winter season months (especially for February and December), and for November in the first two future periods;
    • We have small variations in the future, around the values from the reference period, for January, March and August.
  • Figure 48: Comparison of the mean monthly simulated discharges for the selected 30 years future periods, Mures - Arad
    Figure 49: Comparison of the mean monthly simulated discharges for the selected 30 years future periods, Arges - Budesti

top

3.2 Floods

No clear picture can be drawn about possible changes in the flood conditions. While more frequent winter floods are expected, the decrease of mean flow in some seasons is not followed by the decrease of flood peaks. Torrential type of flood events may occur even more frequently, while the frequency of floods with long duration and large volume may become lower.

Figures 50-67 show the results of statistical analyses of the annual maximum discharges for the following rivers: Upper Tisza (Fig. 50-51), Lower Tisza (Fig. 52-53), Szamos (Fig. 54-55), Sajó (Fig. 56-57), Hármas-Körös (Fig. 58-59), Aries (Fig. 60-61), Tarnava (Fig. 62-63), Mures (Fig. 64-65), Dambovita (Fig. 66-67).

Figure 50: Estimated 30-year return level, Tisza - Vásárosnamény
Figure 51: Time-dependent histogram, Tisza - Vásárosnamény

Figure 52: Estimated 30-year return level, Tisza - Szeged
Figure 53: Time-dependent histogram, Tisza - Szeged

Figure 54: Estimated 30-year return level, Szamos - Csenger
Figure 55: Time-dependent histogram, Szamos - Csenger

Figure 56: Estimated 30-year return level, Sajó - Felsőzsolca
Figure 57: Time-dependent histogram, Sajó - Felsőzsolca

Figure 58: Estimated 30-year return level, Hármas-Körös - Gyoma
Figure 59: Time-dependent histogram, Hármas-Körös - Gyoma

Figure 60: Estimated 30-year return level, Aries - Turda
Figure 61: Time-dependent histogram, Aries - Turda

Figure 62: Estimated 30-year return level, Tarnava - Mihalt
Figure 63: Time-dependent histogram, Tarnava - Mihalt

Figure 64: Estimated 30-year return level, Mures - Ludus
Figure 65: Time-dependent histogram, Mures - Ludus

Figure 66: Estimated 30-year return level, Dambovita - Lunguletu
Figure 67: Time-dependent histogram, Dambovita - Lunguletu

top

3.3 Low water conditions

The study of the Lower Danube low water conditions indicates the possibility of more expressed and longer low flow periods. Some 25% increase of low water levels/discharges is possible to cause problems for the users who require considerable amount of water. The other problem is limitation of water intake. The navigability of the river section may also deteriorate to a certain extent. Figures 68, 69 and 70 present the number of low water occurrences for Mures, Lower Tisza and Lower Danube, respectively.

Figure 68: Low water occurances, Mures - Arad
Figure 69: Low water occurances, Tisza - Szeged
Figure 70: Low water occurances, Danube - Bazias

top

4. Evaporation estimation for Lake Balaton

Lake Balaton is one of the most valuable natural resources of Hungary. It plays a significant economic role in the country via its tourism industry: it is the second most popular tourist destination in the country, after Budapest. The main attractions are water sports (including water skiing, sailing, swimming and fishing, just to name a few) thus maintaining a proper lake water level is crucial for these activities.

Since the beginning of this decade, however, significant weather extremes have been observed in Hungary. Initially, lake water level had dropped significantly due to a drought (which made many consider supplementing the missing water from the outside sources) followed by a year of abundant precipitation resulting in a rapid rise in water level reaching the allowable upper maximum that eventually led to opening the Sió Canal to drain the surplus water, something that has happened rather seldom in the recent past.

In the next 50 years the climate may become drier, many inhabitants, municipalities, and tourists can be unfavourably affected by the resulting siltation of the coastal region and a shrinking lake water surface area, which makes water sports and transport more problematic. Therefore, despite today’s wetter climate, the idea of a water supplementation system similar to the one built for Lake Velence is the topic of frequent discussion.

The evaporation of Lake Balaton has been changing within a narrow range in the last 80 years, with a difference of 30% between the minimum (723 mm) and maximum (1073 mm) values (Fig. 71). Traditional evaporation estimation methods employed in Hungary revealed significant errors that justified the development of new techniques.

Figure 71: Evaporation from the water balance of Lake Balaton in the period of 1921-2007Figure 71: Evaporation from the water balance of Lake Balaton in the period of 1921-2007

The calculations with the help of long-term water-balance derived monthly evaporation estimates for the shallow, medium-sized Lake Balaton, and regional climate model (REMO) outputs as inputs to the employed potential evaporation equations, have confirmed that shallow-lake evaporation can typically be considered as intermediary between the two welldefined potential evaporation rates: the Penman evaporation rate and the Priestley-Taylor, the so-called equilibrium or wet-environment evaporation rate. By weighting these two evaporation estimates on a monthly basis, the lake evaporation can be fairly well estimated. In addition to this approach, a standard lake evaporation estimation algorithm, the so called WREVAP model has also been applied.

The input data for the evaporation estimation models were provided by the regional climate model REMO of the Max Planck Institute for Meteorology in Hamburg for the time period of 1951-2050. Future boundary conditions for REMO5.7 were taken from the global coupled atmosphere-ocean model ECHAM5/MPI-OM, based on an A1B scenario (which is a medium scenario), while for the past (1960-2000) another REMO5.0 run with reanalysis (ERA40) data as boundary conditions was chosen additionally.

The evaporation estimation for Lake Balaton was achieved by choosing REMO gridpoints around the lake and using ensemble averages of the gridpoint variables. For the 1960-2000 period the weighted-equation based on annual evaporation estimates were compared to the water-balance derived from the annual rates for the calibration and verification years. In the calibration period the combination method underestimated the water-balance, which derived mean annual evaporation rate of 889 mm by a mere 7 mm (~1% relative error) and the 866 mm by 15 mm in the verification period (~2% relative error). Morton’s WREVAP program underestimated it by 23 mm (~2.5% relative error) in the calibration period, and by 22 mm (~2.4% relative error) in the verification period.

After determining the weights with REMO5.7 (driven by ECHAM5/MPI-OM) data for the 1951-2000 period, evaporation for the 2001-2050 period could be obtained (Fig. 72). According to the results no significant changes in mean annual lake evaporation can be expected under the A1B scenario: for the 50-year time period the expected mean annual evaporation rate is 888 mm, which is only a 1-mm increment when compared to the water balance value for 1951-2000 and only a mere 3 mm (~1/3 of a percent relative increase) when compared with the same water balance value for the 1961-2000 period. The Morton model with REMO 5.7 inputs and no tunable parameters yielded 773 mm/year for 1951–2000 and 785 mm/year for 2001-2050, both a significant underestimation, yet the evaporation increase again is under 2% (being close to the limit of error). Considering that Morton’s program performed rather well with REMO5.0 forced by ERA40, there may be some slight underestimation of the range of the seasonal cycle of global radiation in the REMO5.7 data required by the evaporation estimation models, which can be in direct relation with the deficiencies in the global fields. The combination method did not solve the problem because its parameters could be adjusted to be on target with the water balance derived values for 1951-2000.

Figure 72: Evaporation of Lake Balaton in the period 2001- 2050 estimated by different methods for REMO5.7 driven by  ECHAM withFigure 72: Evaporation of Lake Balaton in the period 2001- 2050 estimated by different methods for REMO5.7 driven by ECHAM with

In the light of these results the water supplementation to Lake Balaton from the outside sources, to counteract a possible increase in lake evaporation due to an expected warmer climate, does not seem justified. This may sound as a relief sign for ecologists and environmentalists who were worried about the possible ecological changes in lake flora and fauna as a result of supplying stream water to the lake.

The results of the analysis of the tourism’s sector vulnerability to climate change at Lake Balaton can be found within the results of WP4.

top