UNIVERSITY OF PANNONIA GEORGIKON FACULTY

Doctoral School of Animal and Agro-Environmental Sciences Headmaster: Dr habil. Angéla Anda Doctor of the Hungarian Academy of Sciences

WATER BALANCE ANALYSIS OF LAKE BALATON, WITH SPECIAL REGARD TO THE METODOLOGICAL AND REGIONAL QUESTIONS OF EVAPORATION.

THESES OF DOCTORAL (PHD) ESSAY

Written by: BALÁZS VARGA Supervisor: Dr habil. Angéla Anda Doctor of the Hungarian Academy of Sciences

KESZTHELY 2009

CONTENTS

1. BACKGROUND

3

2. AIMES OF THE STUDY

3

3. MATHERIALS AND METHODES

4

3.1. Calculation of areal rainfall on the lake surface and on the catchment

4

3.2. Methodes used to determine evaporation, measurement locations 6 3.3. Calculation of water balance and natural water reserves change 7 4. RESULTS 4.1. Analysis of the water balance components of Lake Balaton

8 8

4.2. Studies on methodes used to measure and calculate evaporation 13 4.2.1. Effect of the location on potantial evapotranspiration

13

4.2.2. Effect of the measurement method on evaporation values

16

4.2.3. Effect of the method used for water level determination on pan evaporation 18

5. THESISES

19

6. LIST OF PUBLICATION IN THE TOPIC OF THE DISSERTATION 21

2

1. BACKGROUND Lake Balaton is the largest lake in Central Europe, but its mean depth is only 3.3 metres and its water mass is only around 2 km³. Compared with this relatively small water quantity, the lake surface, or potential evaporation surface, is approximately 600 km². Following the extremes in the water level during the period 2000–2004, detailed studies on the water balance of Lake Balaton are again in the focus of interest. The main question is whether the water balance of the lake has been modified by the possible effects of climate change since the last major water balance studies, carried out in the 1970s, and if so, what factors were responsible for the substantial decline in the water level during the period in question and what factors are likely to have a fundamental influence on the sustainability of the water balance in the future. Even experts versed in this subject were surprised by the sudden drop in the water level in the early years of this century and by the equally sudden resolution of the crisis in 2005. These rapid changes definitely demand attention in order to prepare us for the consequences of similar phenomena in the future. Following the regulation work carried out in 1863, Lake Balaton became a lake regulated from above, and the only means of keeping the water level between the desired levels was by regulating the amount of water allowed to flow down the Sió Canal. Apart from rainfall on the surface, the water reserves of Lake Balaton are determined primarily by the rainfall conditions in the catchment area. Despite several attempts at revising the methodology, the greatest uncertainty in establishing the water balance of the lake is still the precise estimation of surface evaporation. No solution has been found for the continuous monitoring of evaporation conditions above the open water, so at present data from shore-based stations are used for the calculations. The present analysis of evaporation in Keszthely Bay was designed to determine territorial differences in evaporation, together with a revision of the methodology used to measure evaporation.

2. AIMS OF THE STUDY One major aim of the work was to carry out a long-term climatic and statistical analysis of the main components of the water balance of Lake Balaton. Analyses of the quantity of precipitation falling on the lake surface 3

and the annual quantity of evaporation, including its distribution over the year, were supplemented by the study of influx. Seeing the modification in the quantity of water entering the lake from the catchment basin, the aim was to determine changes in rainfall conditions in the catchment basin between 1921 and 2007 in order to identify the sub-basins most sensitive to the effects of climate change and those of special importance in the long-term sustainability of the water balance. After an analysis of the main components of the water balance, a further aim was to analyse changes in the water balance and natural water reserves of the lake. Due to uncertainties experienced when calculating evaporation, it was aimed to investigate evaporation in the Keszthely Bay region in order to demonstrate differences in the evaporation values calculated from data recorded at three measurement sites: at the meteorological station, on the lake shore and above the open water. It was hoped, on the basis of the results, to determine whether the evaporation values recorded at the meteorological station and currently used for calculations, were suitable for calculating the evaporation of the lake surface. Methodological studies on evaporation measurement were designed to demonstrate whether the evaporation values determined using various methods were comparable, and to elaborate correction factors, based on a 19-year database, that could be used to make the results obtained with various types of evaporation pans comparable. By elaborating and testing a method for continuous evaporation measurement it was hoped to make the determination of evaporation more accurate.

3. MATERIALS AND METHODS Studies were made on the whole surface of Lake Balaton to analyse trends in the main components of the water balance. The database was put at our disposal by the Balaton Branch of the Central Transdanubia Environment and Water Directorate and by the Environment and Water Research Institute Non-Profit Co. The original form of the data series, which included data on rainfall to the lake surface, influx determined by rainfall to the water catchment basin, drain-off and the transfer of mine water, lake evaporation and the 4

monthly water quantity drained off via the Sió Canal during the 1921–2007 period, was processed to give derived data. In the course of statistical analysis of the data series, observations on past trends were based on the evaluation of regression lines fitted to the data series and on 30-year averages calculated every ten years throughout the period, on sliding 30-year averages with an increment of ten years throughout the period recommended by WMO. 3.1. Calculation of rainfall on the lake surface and on the catchment The precipitation on the whole surface of Lake Balaton was calculated using the Thiessen polygon method. Mean areal rainfall was obtained in the form of the mean quantity falling on each polygon, weighted with the size of the polygon. The rainfall was calculated for each sub-basin, and these values were used to calculate the value for the whole of the lake surface. The mean rainfall on the lake surface was determined after weighting the rainfall means for the sub-basins. When characterising rainfall conditions over the whole of the catchment area, Balaton catchment basin was divided into 5 sub-basins. In delineating these sub-basins, consideration was given not only to their geographical location, but also to their role in the water balance of the lake and to special climatic features. The Zala catchment basin to the west of the lake was given special attention due to its outstanding role in the water balance and to its special climatic features, and was thus divided into two parts along a line stretching from Keszthely via Zalacsány to Tilaj. The delineation of the other three subbasins followed the standard classification, and they could be denoted as the north, south and east catchment basins, which contained 5, 6 and 3 rainfallmeasuring stations, respectively. The 25 meteorological stations covering the whole of the catchment basin are mostly equipped with conventional Hellmann precipitation meters. Automatic sensors were installed in the earlier years of the century, mainly at the larger stations. The areal rainfall was calculated from the arithmetical means of the data from stations in the given sub-basin.

5

3.2. Methods used to determine evaporation; measurement locations Climatic data from Siófok and Keszthely were used to calculate the monthly evaporation from Lake Balaton. Surface water loss was calculated using the original Meyer formula between 1921 and 1974, and using a modified version of this formula between 1975 and 1985. The original data series obtained from VITUKI was evaluated when analysing the time series of surface evaporation and in water balance analysis. Measurements were carried out at three locations for the precise determination of evaporation in Keszthely Bay. The database was based on observations from the meteorological station maintained by the National Meteorological Service at the Department of Meteorology and Water Management in Keszthely (Georgikon Faculty, Pannon University). The second location was an evaporation measurement station set up directly on the lake shore, on the premises of the Keszthely Yacht Club, in 2006. The third measurement site, in operation since autumn 2006, is on a measurement column set up by the Balaton Integration Non-Profit Co. for measurements above the lake surface. The relative humidity and air temperature data required for calculating evaporation from the open water areas of Keszthely Bay were taken every 10 minutes with the same method at all three locations, using a data collector of the DO-9406 type, placed in a standard thermometer box placed at a height of 2 metres. The Antal formula widely used in Hungary was applied for evaporation studies in the Keszthely Bay region (ANTAL 1968). This empirical method for determining the daily value of potential evapotranspiration is also known and applied abroad.

where

PET = 0.74 (E – e)0.7 (1 + αT)4.8 α: thermal expansion coefficient of the air (273K–1) E: saturation vapour pressure associated with daily mean temperature (mbar) e: daily mean vapour pressure (mbar) T: daily mean temperature (°C).

Pan evaporation measurements were mostly conducted at the Tanyakereszti Meteorological Station, where five types of evaporation pans 6

were operated (A, U, GGI-3000, INEP and 20 m² pans). The results obtained with these pans during the period 1982–2000 were statistically compared using the t-test. The water level in the evaporation pans was recorded once a day at 7 am using Piche tubes, while the water temperature in the pans was measured between 7 am and 1 pm. The central component in the evaporation measurement station operating on the shore of Lake Balaton since June 2006 is an A-type evaporation pan, with the requisite water level and water temperature sensors. When the station was set up a capacitive sensor was used to record the data, but this was later replaced by a hydrostatic sensor. During the testing stage, the automatic registration of the water level every 10 minutes was complemented by conventional measurements on the water level using a Piche tube at 7 am. 3.3. Calculation of water balance and natural water reserves change The water balance of Lake Balaton was calculated in the usual manner with the following formula:

where

∆K = Cs + H – (P + VK + L) Cs: precipitation falling on the lake surface (mm) H: surface inflow (mm) P: evaporation from the water surface (mm) VK: water utilisation directly from the lake (since 1971) (mm) ∆K: changes in the water reserves of the lake (mm) L: water quantity drained trough the Sió Canal (mm)

The effect of possible modifications in the hydrological and meteorological components of the water balance on the water balance of the lake was studied by analysing changes in the natural water reserves (∆KT): ∆KT = (Cs + H) – P

7

4. RESULTS 4.1. Analysis of the components of the water balance of Lake Balaton The input side of the water balance is determined primarily by the rainfall, directly by that falling on the lake surface and indirectly in the form of influx from the catchment basin. The mean annual rainfall sum for 1921–2007 was 618 ± 108 mm, while this figure was only 601 mm for 1971–2000. The least rainfall since measurements were begun was recorded in 1932, with only 433 mm over the lake surface, while the wettest year was five years later in 1937, with a total of 905 mm precipitation. There was thus a difference of 472 mm between the minimum and maximum values, with a mean absolute deviation of 92.3 mm. Linear trend fitting revealed a decline of 62.5 mm/100 years in the rainfall sum, but the R² value (0.021) was not significant at the 5% level. The 86-year data series was divided into 30-year periods, sliding by 10 years each time. The decline could also be detected using the means suggested by WMO, particularly during the second half of the test period. When analysing trends in rainfall over the year, no month was found when the change in the rainfall quantity was significant at the 5% level over the 1921–2007 period. When the annual rainfall data for the lake surface were interpolated using a neural network and the results were plotted, within-year rearrangements were observed, in contrast to the trends for the whole period (Fig. 1). The May and September rainfall maximums characteristic of the first half of the period gradually gave way to a pronounced grand maximum by the 1950s, the spring peak being shifted to June, while the period of higher rainfall in autumn gradually appeared earlier (first in August, later in July). By the 1950s a June–July rainfall maximum had developed.

8

Figure 1. Within-year rearrangement of rainfall between 1921 and 2007 by interpolation using a neural network Later this grand maximum became less pronounced, which is unfavourable for the summer water regime of the lake, as the increase in water reserves from influx during the summer and autumn months makes up a relatively small proportion of the total annual sum, and the water quantity lost through evaporation can many be replaced by rainfall on the lake surface. The mean areal value of the annual rainfall sum calculated for the whole of the catchment basin using the data from all 25 stations was 684.4 ± 111.4 mm over the period 1921–2007, while the figure for 1971–2000 was considerably lower (657.3 mm). Reductions in rainfall significant at the 5% level were only observed at certain stations in the northern part of the Zala subbasin, while for the other four sub-basins, although the steepness of the trend line had a negative sign in all cases, only a few individual stations (Siófok, Fonyód, Nagyvázsony, Keszthely) recorded a decline in rainfall sum that was significant at the 5% level. In three of the four sub-basins where most of the stations did not record significant changes in rainfall, the stations where changes were found to be significant were located on the lake shore (Keszthely, Fonyód, Siófok). In line with WMO recommendations, climate norms were developed for each measuring site using moving time periods of 30 years with an increment of 10 years. At most of the stations found in the northern part of the Zala subbasin rainfall trends were described by a similar curve, the only exception being Zalaegerszeg, where a gradual decline in rainfall could be observed over 9

the whole period. In the southern part of the Zala sub-basin, a curve similar to that recorded in the northern part was detected at stations in Felsırajk, Vése, Zalakomár and Zalavár, but the highest 30-year mean was observed later at most of these stations, during the 1951–1980 period. The northern sub-basin of the Lake Balaton catchment area presented a varied picture as regards rainfall trends. In the Tapolca Basin there were constant rainfall means from the 1921– 1950 period to the 1941–1970 period, after which a sudden decline led to a further period with constant rainfall means at a lower level. The Nagyvázsony region was characterised by a dynamically decreasing trend throughout the whole period. Balatonalmádi was also characterised by a drop in rainfall, but only from the 1962–1990 period onwards. Compared to these stations, the rainfall conditions in Tihany and Szentantalfa remained constant, which could be due to the geographical location of the measurement equipment. In terms of rainfall trends, the southern sub-basin was one of the most variable parts of the catchment area. While the rainfall was constant over time in Balatonkeresztúr and Balatonlelle, the rainfall tended to increase to varying extents during the first part of the period at the other stations, up to the 2nd–4th climate norm, depending on the station. In the second half of the period, however, there was a reduction in rainfall, which continued to the end of the period in the case of Fonyód and Nagybajom, while in Somogyvámos, and especially in Marcali, this trend changed towards the end of the period, giving way to a moderate increase. The eastern sub-basin of the Lake Balaton water catchment area is of less importance for the lake water balance, but changes in the rainfall conditions were nevertheless worthy of attention. Declining trends were observed at all three stations. This was most pronounced in Siófok and more moderate in Lepsény, while at the Karád station, situated further to the south, the decline was followed by an increasing trend at the end of the period. From the point of view of the lake water balance, the distribution of rainfall over the year is also of outstanding importance. Several studies on rainfall conditions in Hungary treat the disappearance of the secondary maximum as a recent phenomenon, but this is not supported by the data of the 25 stations representative of rainfall conditions in the Balaton catchment area. At ten of these stations (Sümeg, Zalavár, Zalakomár, Vése, Tapolca, Balatonkeresztúr, Marcali, Fonyód, Balatonlelle, Siófok) the ratio of rainfall in August over the last thirty years has increased significantly, by 5–10%, compared to the whole period, and at several stations it has exceeded the rainfall sum in June. 10

Some of the rain falling in the catchment area appears in the water balance of Lake Balaton in the form of influx, the most important component on the input side. The annual influx calculated for the whole of the lake surface is dependent on the rainfall quantity, but in terms of the long-year mean (878 mm) the water entering the lake from influx is one and a half times the quantity received from rainfall directly onto the lake surface. The mean value of influx was 874 ± 300 mm between 1921 and 2007. The data series had a range of 1681 mm, while the mean absolute deviation was 230.4 mm. Based on the steepness of the linear trend line, the influx declined by 332 mm/100 years, shown by the R² value (0.078) to represent a significant change at the 5% level. The moving mean calculated from the data of 10 years also gives a clear illustration of the declining trend. The reduction in influx can also be detected from the 30-year means, especially during the second half of the period. The means were above the mean for the entire period right up to 1971–2000, after which there was a sharp decline in the values. This steep drop in influx could be attributed in part to the substantial rainfall deficit in the Lake Balaton area between 2000 and 2004, which was most pronounced in the western part of the Zala sub-basin, and also to the fact that 10 of the 15 years with the lowest influx fell between 1990 and 2007. With the exception of December and January, the linear trend was negative in every month, though it was only significant in February (48.1 mm/100 years), June (41.1 mm/100 years) and July (40 mm/100 years). During the 1921–2007 period, the mean evaporation of Lake Balaton was 900.7 ± 71.4 mm. The greatest quantity of water (1073 mm) passed from the lake surface to the atmosphere in 1946 and the least (723 mm) in 1970. The data thus had a range of 350 mm, with a mean absolute deviation of 57.1 mm. The values of the bottom and top quartiles were 857 and 949 mm, respectively, so half the years are likely to have lake evaporation between these two figures. The linear regression trendline revealed a decrease of 29.7 mm/100 years, but this was not statistically significant, so it can be said that over the entire period the annual evaporation of the lake did not change. The moving mean calculated from 10 year means, however, pinpointed periods when the change was more pronounced. Similar results were obtained when sliding 30-year means with a 10-year increment were analysed. The 1921–1950 and 1931–1960 means showed that these periods had somewhat more intensive evaporation than the 11

period as a whole, while the 1941–1970 mean was similar to that of the whole period. After this a substantial reduction in evaporation was observed. The 1950–1980 and 1961–1990 mean values were indicative of a period with only moderate evaporation, after which the intensity of evaporation gradually rose again. With the exception of March and May, evaporation from the lake exhibited significant changes in every month, though the sign of this tendency differed. The lake evaporation decreased from April to September, when evaporation was more intense, while the data revealed an increase in the autumn and winter months, from October to February. The change was greatest in October (17.2 mm/100 years) and November (32.7 mm/100 years), but the increase in evaporation was not negligible in the winter months either. The greatest reduction was recorded in June and July, with values of 27.7 and 26.9 mm/100 years. The rearrangement of evaporation within the year could be clearly seen when the database was interpolated and plotted with the help of a neural network. Although no change was observed in March on the basis of the raw data, it is clear from Figure 2 that the period of more intensive evaporation has shifted to an earlier date, while the similar period in November can be seen to have shifted to a later date.

Figure 2. Rearrangement of evaporation within the year between 1921 and 2007, using interpolation and a neural network The quantity of water allowed to drain out of the lake via the Sió Canal is artificially regulated, so no far-reaching conclusions can be drawn from this component of the water balance. Apart from the mass of water entering the 12

lake, the quantity of water drained was also greatly dependent on the current water management concept, which changed on several occasions during the period tested. The mean quantity of drained-off water between 1921 and 2007 was 573±384 mm. An analysis of the data series revealed that right up to 1993, it was necessary every year to reduce the water level artificially to a greater or lesser extent, while in 1993 and later between April 2000 and September 2005, the sluice-gates remained closed, except when the Sió riverbed was rinsed out. The data series had a range of 1791 mm, with a mean deviation of 305 mm. The values of the bottom and top quartiles were 275 and 850 mm, respectively. Trend analysis indicated a significant reduction in the quantity of water drained off, with a value of over 400 mm/100 years. The significance of this change is indicated by the fact that the mean value for 1921–2007 was 574 mm, while the mean for the last 30 years was only 379 mm, a reduction of 35%. 4.2. Studies on methods used to measure and calculate evaporation 4.2.1. Effect of the location on potential evaporation Evaporation in the Keszthely Bay area was recorded in 2006 and 2007, but neither measurement period covered the whole of the intensive evaporation period. Measurements above the open water surface were begun on 6 September 2006. The comparative analysis of the evaporation values calculated from the daily data did not reveal any significant differences between the potential evapotranspiration at the lake-shore station and at the meteorological station. The evaporation calculated from data recorded above the open water of the bay during the period from 6 September to 15 November was significantly different at the 5% level from that recorded both on the lake shore and at the meteorological station. During the period tested, the potential evapotranspiration calculated above the bay was 14.8% higher than that calculated at the station. It was thus concluded that in the autumn months the potential evapotranspiration above the open water was considerably greater than the data from the meteorological station, which are currently used to calculate the evaporation figures included when compiling the water balance. The differences were confirmed by calculating cumulative evaporation values for the given period, which showed that the evaporation above the open 13

water was more intense even at the beginning of the test perio period than on the lake shore or at the meteorological station. This difference gradually increased and could be attributed not to a few days of unusually intense evaporation, but to a tendency observed throughout the period.

Figure 3.. Comparison of the evap evaporation on the shore of Lake Balaton and above Keszthely Bay between 6 September and 15 November 2006 The total evaporation above the lake surface during this period was 186 mm, which was significantly higher than the 158 mm recorded on the lake shore or the 162 mm calculated at the meteorological station. In the present studies, the relationship between tthe evaporation data calculated at the two locations was found to be linear. The evaporation values recorded at the meteorological station (x) could be used to calculate the evaporation above the open water (y) at a satisfactory level of reliability using tthe formula y = 0.92x + 0.52. In 2006 the pan evaporation at the meteorological station was compared with that at the lake shore station. The evaporation sums recorded between 12 August and 29 November 2006 appeared to be relatively similar, but the difference nce was found to be significant at the 5% level. For the whole of the observation period, the mean absolute deviation calculated from daily sums was 0.67 mm, while the deviation between the two stations expressed as a percentage of the mean value, again ca calculated from daily sums, was 4.3%. In the case of heavy rain or strong wind, extremely great differences were found, and it appeared probable that these were not the result of errors arising from the measurement techniques. It is definitely not recommende recommended to compare the two approaches on such days. 14

The comparison of evaporation figures calculated for 2007 was complicated by the failure of the DO loggers, resulting in a lack of data during certain periods. Between 1 April and 9 June no differences could be detected between the evaporation recorded at the meteorological station and on the shore of the lake, as in the previous year. Over a total of 70 days, 314 mm evaporation was measured at the station, compared with 312 mm on the shore. Between 18 June and 17 August 2007 a difference significant at the 5% level was detected when comparing evaporation data recorded at the meteorological station and above the lake surface, with a value 31.3 mm higher above the open water (354.6 mm) than at the meteorological station (323.4 mm). Based on the statistics of 58 measurement days, the difference between the evaporation at the two locations was 8.9%. From 18 August 2007 data were available for comparison from all three locations. On the basis of 78 days of measurement, no significant difference was found between the evaporation recorded on the shore of Lake Balaton (193.4 mm) and above the open water (195.6 mm), a result that differed from observations in previous periods. At the same time, the evaporation calculated from data recorded at the meteorological station (176.6 mm) was significantly (5%) lower that that measured on the shore and above the open water, with a difference in evapotranspiration of 9.7% between the meteorological station and the open water. Cumulative evaporation was calculated in 2007 for a longer period than in 2006, from 18 June to 31 October, using the data from the meteorological station and the open water in Keszthely Bay. A difference in evaporation was again perceptible right from the beginning of the measurement period, amounting to a 101.5 mm (16%) higher evaporation sum over the open water after 142 days, which was significant at the 5% level. This suggests that much more exact values could be obtained when calculating evaporation if it were possible to use values recorded above the lake surface. Due to the longer data series, the evaporation above the lake surface could be calculated with a considerably higher coefficient of determination than in 2006 using the equation y = 0.7x + 0.61. This means that in the future it will be possible to make an approximation of evaporation above the open water even if data recorded above the lake surface are not available. Pan evaporation was measured continuously at both locations in 2007 from 1 March to 31 October. In March and April the results obtained at the two locations were very similar, with slight differences only on a few individual 15

days. During this period the evaporation from A pans at both locations was also at par on the basis of statistical analysis. In May and June, however, significant differences were found between the two locations, and a similar situation was observed in July and August. The evaporation from the two pans no longer differed significantly in September and October. 4.2.2. Effect of the measurement method on evaporation values Due to differences in their location and size, the evaporation pans responded differently to changes in the meteorological factors that have a decisive effect on evaporation. The analysis aimed to identify parameters that could be used to compare evaporation data from individual pans and if necessary to convert them (Table 1). The evaporation recorded for the INEP pan, which has double, insulated walls and is sunk into the ground, exhibited the greatest similarity to the water loss from the A pan, though relatively great deviations were observed in late summer (August) and in the autumn months, when the absolute value of evaporation is substantially lower. From the spring months until August, the evaporation from the A pan and the INEP pan can be regarded as practically identical. Table 1. Monthly means of the ratios between evaporation from the A pan and from pans regularly used in Hungary (20 m², U, GGI-3000, INEP) in Keszthely between 1982 and 2007 Month/Pan Annual April May June July Aug. Szept. Okt. type mean A-pan 1 1 1 1 1 1 1 1 2 20 m pan 0.75 0.76 0.82 0.81 0.85 0.86 0.97 0.83 2 20 m pan SD +0.11 +0.11 +0.05 +0.05 +0.08 +0.11 +0.19 +0.06 U-kád 0.85 0.87 0.87 0.87 0.87 0.84 0.91 0.87 U-kád SD +0.09 +0.09 +0.07 +0.05 +0.07 +0.09 +0.16 +0.05 GGI-3000 1.05 1.04 1.06 1.10 1.09 1.11 1.19 1.09 GGI-3000 SD +0.12 +0.13 +0.17 +0.18 +0.1 +0.20 +0.20 +0.11 INEP-kád 0-99 0.99 1.01 1.02 1.10 1.07 1.12 1.05 INEP-kád SD +0.12 +0.12 +0.16 +0.12 +0.19 +0.15 +0.22 +0.12 The evaporation from the GGI pan did not differ substantially from that of the A pan, though the difference was slightly greater than that found when 16

comparing the INEP and A pans. Again, greater differences and standard deviations were observed for the GGI pan in the autumn months. From spring to autumn, the difference in the evaporation from the A and GGI pans did not exceed 10%, so if necessary they can substitute each other. Among the two pans sunk into the ground, the water temperature of the GGI pan responded more rapidly to changes in air temperature. Evaporation from the GGI pan exceeded that from the A pan to a moderate extent from April to June and to a considerable extent from July to October. The evaporation from the 20 m² pan was considerably less than from the A pan. Of all the pan types the evaporation from the 20 m² pan was the lowest of all the five types tested, with the exception of three years. The reason for this more modest evaporation could be partly due to the difference in size and partly because the 20 m² pan is sunk into the ground. The evaporation from the 20 m² pan was twice as high as that of the A pan in September and three times as high in October. The evaporation ratio from the 20 m² pan tended to rise gradually over the year. In spring the deviation was considerable (more than 20–25%), while it was slightly lower in autumn, which could be due to the slower warming and cooling of the greater water mass. The evaporation from the U pan was hardly greater than that from the 20 m² pan in terms of the whole year, but substantial differences in evaporation were observed in some months. In spring the evaporation ratio was higher than that observed for the 20 m² pan, while in autumn the evaporation from the U pan was the lower of the two. Again, the greater energy retention of the larger water mass may have been responsible for the enhanced evaporation from the 20 m² pan. The ratios averaged over 19 years were lower in every month than those recorded for the A pan, and only in June of one year and October of another year was the monthly evaporation greater than that of the A pan. The applicability of the calculated correction factors was tested for the results of the test period. The significance level of deviations in the evaporation from different pan types was first determined using a paired t-test, in which case correction is required. In all cases the A pan was taken as the basis of comparison. Significant differences in the evaporation trends were observed in two months for the GGI pan and only in July for the INEP pan. Due to the negligible differences detected and the fact that they are used less frequently, there was no justification for applying a correction for the two smaller pans, which were sunk into the ground. This confirmed conclusions drawn earlier 17

from the near-unity values of the ratios. Evaporation from the 20 m² pan, except in October, and from the U pan in the summer differed significantly from that of the A pan, so a correction was required when the evaporation figures were converted. When correction factors were applied, the evaporation from both the 20 m² pan and from the U pan was statistically on par with that of the A pan. 4.2.3. Effect of the method used for water level determination on pan evaporation values The testing of an automatic water level sensor was an important part of work on the measurement of the water level in evaporation pans. After the capacitive sensor was put into operation on 1 June 2006, the evaporation calculated using data recorded by the instrument was compared with that measured using conventional methods. The results indicated that the capacitive sensor was not a satisfactory means of recording evaporation. The automatically recorded data exhibited a significant deviation from the water level and evaporation values determined with conventional methods for the same pan. The operational errors of the sensor and the unrealistic water level values were probably caused by condensation on sensor parts not immersed in the water and by the rippling of the water in the pan. Based on 44 sample days, the evaporation recorded using the automatic capacitive meter was 25% lower than that measured under standard conditions, and considerable differences were also observed for the daily means and particularly for the standard deviations. It was concluded from the results that, under the given experimental conditions, the capacitive sensor was unable to give an accurate determination of evaporation. The values recorded using a hydrostatic sensor between 1 March and 31 May 2007 were considerably closer to the results of manual measurements. No significant differences were found in the data for any of the three months examined. The differences in monthly evaporation sums were below 10% in all cases, and the standard deviations were also substantially lower. Based on these test measurements, the new sensor could be suitable for the automatic determination of pan water levels, and can definitely be recommended to complement conventional measurements. If combined with the use of an automatic rainfall meter, changes in water level due to external effects can be efficiently detected, provided measurements are made sufficiently frequently. 18

5. THESISES 1.

2.

3.

4.

The total water balance in Lake Balaton was generally in equilibrium from 1921–2007, and except for the period from 2000–2004 there were no serious problems with the total water balance, which reflects the efficiency of water management and lake level regulation. Since the natural water resources of Lake Balaton exhibited a significant decline from 1921–2007, the maintenance of equilibrium was achieved by allowing progressively smaller quantities of water to drain off through the Sió Canal. The decrease in water reserves was the consequence of the significantly lower influx from the water catchment area, combined with the lower quantity of rainfall on the lake surface in recent decades. The change in the water reserves could not be attributed to differences in evaporation, the annual sum of which did not change during the period investigated. Over the entire catchment basin, a decline in rainfall affecting the whole of a sub-basin was only detected for the northern part of the Zala watershed, while significant rainfall reductions were only demonstrated at a few individual stations in the other sub-basins. The whole of the catchment basin was characterised by a rainfall decrease in May and October, with constant rainfall quantities from July to September and an increase in rainfall in December, especially in the eastern catchment area of Lake Balaton. From 1921 to the 1950s the rainfall peaks in May and September gradually changed, giving a main peak in June. Starting from the 1980s, a rainfall peak developed in August and September and a reduction in rainfall in May and June. Lake surface evaporation from April to September declined during the 1921–2007 period, while the intensity of surface evaporation was enhanced from October to February. In the Keszthely Bay region, depending on the season, the potential evaporation above the open water surface was 8.7–14.8% higher than the evaporation calculated using current methods at the meteorological stations, suggesting that more accurate evaporation values could be provided by using values measured above the open water surface. The present results suggest that the evaporation of the two areas could be converted with a high coefficient of determination. 19

5.

The comparative analysis of evaporation pans showed that the results obtained using INEP and GGI-3000 pans were comparable without correction with the evaporation of A-class pans, while the evaporation of 20 m² pans or U pans could only be compared with that of A pans using a correction factor that varied with the months. In humid, damp environments a hydrostatic sensor can be recommended for the automatic monitoring of the water level in the pans. Values recorded with such sensors could be used to calculate water loss from the pans, and this evaporation value was statistically on par with that calculated using conventional methods.

20

6. LIST OF PUBLICATION IN THE TOPIC OF THE

DISSERTETION Scientific papers published in foreign language in supervised professional periodical: BOLDIZSÁR, A. – VARGA, B. /2006./: Evapotranspiration and development of helophytes. Georgikon for Agriculture Vol. 9 No. 1. p. 91-102. VARGA, B. – BOLDIZSÁR, A. – GIMESI, L. /2007/: Some remarks on water balance components of Lake Balaton. Georgikon for Agriculture Vol. 10 No.1. p.69-82. Scientific papers published in Hungarian in supervised professional periodical: ANDA, A. – VARGA B. /2004/: A Keszthelyi-öböl néhány párolgási sajátossága az elmúlt évtized mérései alapján. Hidrológiai közlöny, 84. évf. 3. p. 65-69. ANDA, A. – VARGA B. /2008/. Párolgásmérı kádak összehasonlító vizsgálata Keszthelyen 1982-2000 között végzett megfigyelések alapján. Hidrológiai Közlöny, 88. évf. 2. sz. BOLDIZSÁR, A. – VARGA B. /2006/: Mocsári növényállományok párolgási jellemzıi. Hidrológiai Közlöny, 86. évf. 4. p.12-15. VARGA, B. /2006/: A Balaton és a Keszthelyi-öböl vízháztartásának hidrometeorológiai vonatkozásai. Légkör, LII. évf. 2. p. 21-26. VARGA, B. – ANDA, A. /2007/: A Balaton párolgásának és energiaháztartásának felülvizsgálatát célzó kutatások és az elsı eredmények bemutatása. Tudományos Közlemények, 7. évf. 1. 2. kötet p. 479-485. VARGA, B. – ANDA, A. /2007/: A Balaton fı vízmérlegtagjainak alakulása 1921-2005 között. Tudományos Közlemények, 7. évf. 1. 3. kötet p. 741-747. VARGA, B. – BOLDIZSÁR, A. – KOCSIS, T. – BEM, J. /2006./: A Balaton párolgás vizsgálatának történeti fejlıdése. Egyetemi Meteorológiai Füzetek No.20. ISBN: 963 463 874 0.

21

Published conference and poster presentations in foreign language: BOLDIZSÁR, A. – VARGA, B. – KOCSIS, T. /2006/: Microclimate and transpiration of reedbeds at changing water level on Lake Balaton. 6th Annual Meeting of the European Meteorological Society, Ljubljana. (CD ISSN 1812-7053) VARGA, B. – BOLDIZSÁR, A. /2006/: Water balance components of Lake Balaton from 1921 until today. 6th Annual Meeting of the European Meteorological Society, Ljubljana. (CD ISSN 1812-7053) VARGA, B. – BOLDIZSÁR, A. – KOCSIS, T. /2006/: Water management’s analysis of Lake Balaton. Ecological Problems of Our Days - From Global to Local Scale International Conference, Keszthely. (CD kiadvány ISBN 963-9639-14-1). Published conference and poster presentations in Hungarian: ANDA, A. – VARGA, B. /2007/: A Keszthelyi-öböl párolgásbecslése. XLIX. Georgikon Napok, Keszthely(CD kiadvány ISBN 978-963 9639-22-5) ANDA, A. – VARGA, B. /2007/: A tó felett végzett elızetes sugárzásra vonatkozó megfigyelések a Balatonon. Környezeti ártalmak és a Légzırendszer XVII. kötet.(CD kiadvány, ISBN: 978-963-87327-1-2) BOLDIZSÁR, A. – VARGA, B. – KOCSIS, T. /2006./: Mocsári növények vízfelhasználása. XLVIII. Georgikon Napok, Keszthely (CD kiadvány ISBN 963- 9639-11-7) VARGA, B. – BOLDIZSÁR, A. – KOCSIS, T. /2006/: A Balaton vízháztartási elemeinek alakulása 1921-tıl napjainkig. Poszter a VAHAVA Projektzáró-konferenciáján VARGA, B. /2004/: A Balaton vízszintváltozásai az éghajlatváltozás tükrében. XLVI. Georgikon Napok, Keszthely (CD kiadvány ISBN 963 9096 962) VARGA, B. /2004/: A Keszthelyi-öböl párolgásának vizsgálata 1991-2002 között. XLVI. Georgikon Napok, Keszthely (CD kiadvány ISBN 963 9096 962) VARGA, B. /2005/: A párolgásmérés módszereinek összehasonlító vizsgálata. XLVII. Georgikon Napok, Keszthely (CD kiadvány ISBN 963 9639 03 6) 22

VARGA, B. /2006/: Szélsıségek a Balaton vízállásának alakulásában 1876-tól napjainkig. XII. Ifjúsági Tudományos Fórum, Keszthely. VARGA, B. /2007/: A Balaton párolgásának jellemzıi. A sport szerepe a turizmus fejlıdésében nemzetközi konferencia, Keszthely. VARGA, B. – ANDA, A. /2007/: Új eredmények a Balaton energiamérleg alkotóinak kutatásában. XLIX. Georgikon Napok, Keszthely (CD kiadvány ISBN 978-963 9639-22-5) VARGA, B. – ANDA, A. /2007/: Kezdeti eredmények a Balaton pontosabb méréssel történı párolgásbecsléséhez. Környezeti ártalmak és a Légzırendszer XVII. kötet.(CD kiadvány, ISBN: 978-963-87327-1-2) VARGA, B. – BOLDIZSÁR, A. – KOCSIS, T. /2006./: A Balaton vízháztartása és az éghajlatváltozás. XLVIII. Georgikon Napok, Keszthely (CD kiadvány ISBN 963 9639 11 7) VARGA, B. – BOLDIZSÁR, A. – KOCSIS, T. /2006/: A Balaton vízmérlegének alakulása az elmúlt évtizedek mérései alapján. V. Természet-, Mőszaki- és Gazdaságtudományok Alkalmazása Nemzetközi Konferencia, Szombathely. (CD kiadvány: ISBN: 9639290-69-6) VARGA, B. – BOLDIZSÁR, A. – KOCSIS, T. /2006/: A Balaton vízgazdálkodásának múltja és prognosztizált jövıje. Környezeti ártalmak és a légzırendszer. XVI. Kötet. p. 213-224.

23