The Role of Groundwater in the Acidification of the Hydrosphere - Examples from Small Catchments in the Bohemian Massif

Z . HRKAL, J. BUCHTELE, E. TlKKANEN, A. KA pA H O s J. SANTRUCEK N GU -BULL 439, 2002 - PA GE 99

The role of groundwater in the acidification of the hydrosphere - examples from small catchments in the Bohemian Massif

ZBYN EKHRKAL,JOSEF BUCHTELE, EEVATIKK ANEN,ASKO KApAHO &JAROMIRSANTRUCEK

Hrkal, Z., Buchtele,J.,Tikkanen, E., Kapyaho, A.& Santrucek,J. 2002:The role of groundwa te r in the acidifica tion of the hydrosphere - exam ple s from small catchment s in the Bohem ian Massif.Norges qeoloqiske undersokelse Bulletin 439, 99-105 .

The aim of the pre sented study was to assess the im pact of the groundwa ter on the degree of acidificati on of sur face w aters in small catchm ent s. Quantit ati ve hydr ogeological analy sis based on th e result s of the SACRAMENTO hydrolog ical mo de l was aime d at th e assessment of th e contribu tio n of baseflow repr esentin g th e groundwater dis charg e in the tot al runoff. Co mparison of single pH measureme nts in surface wate r wi th the corr espondin g actual and mo delled w ater discharge was used as th e info rmatio n in th e qu antitative part of the data processing. The obtai ned result s showed differen t prot ective pow ers of th e aq uifers against acidifica tio n,a conspicuous decrease in the soil buffer capac ity of th e Krusne hory Mts. and a significant influence of th rou gh fall precipitati on s on the gr oun dwater qu alit y.

Zbynek Hrkal,JosefBuchtele, Jaromir Santrucek, Cha rles University, Dept. ofHydrog eology, Czech Republic. Eeva Tikkanen, Asko Kapyah o, University ofHelsinki, Fin land.

Introduction ity of surface water and groundwater could have been The Czech Republic is one of several countries where th e obta ined only from th e monitoring of small experimental effects of anthrogenic atmospheric acidification first began catchments. However, in small catchme nts in th e Czech to reveal th emselves in th e early 1960s. However, the mass Republ ic, as in other countr ies, th e contribution of ground extinction of forest vegetation, first in the highe r parts of the water discharge and quality to surface water was most ly Krusne hory Mts.and later in th e Jizerske hory and Krkono se neglected. Comparisons were made between th e amount Mts., demonstrated the tru e extent of acidification.This pro and quality of precipitation on the one hand, and the same vided conclusive evidence of th e widespread nature of the for runoff on th e oth er hand, without considering the vari acidification, but some doubts still prevailed about th e ability of the ground wat er component in th e surface water. degree of acidification of groundwat ers. Some regional In recent years, an investigation has been sponsored by studies (Hrkal 1992,Hrkal & Fottov a 1999) reported evidence Grant 11310005 of th e Min istry of Education of the Czech of progre ssive acidification of shallow, near-surface aquifers Republic 'Material and energy flows in the uppe r parts of th e in the crystalline bedrock compl exes which revealed a con Earth', and this has been carried out within the 5th siderable decrease in alkalinity leading to th e complete dis Framework Programm e sponsored by the Europ ean Union . appearance of bicarbonates,decrease in pH,and an increase The LOWRGREP project 'Landscape-use optimisation with in AI and Be concentrations. Whilst the quality of surface regard to ground water resources protection in mo untain waters is affected by atmospheric deposit ion directly and hardrock areas' was intended to assess the contribution of im mediately, the impact of rainfall on th e chemical composi groundwater im pact on the deg ree of acidificat ion of sur tion of groundwate rs is delayed. Owing to th e contact of face waters. ground water with th e bedrock enviro nment, and also pri marily with soil strata, th e ground water enjoys a degree of Method of data processing potential buffering against acidification, and consequently The selection of test sites was th e first step of the wor k. For may play an important role in th e resulting chemistry of sur this reason, data from th e experimental small catchment face water discharge. However, th e proportion of ground network GEOMON (Fottova 2000) have been used. The water in the total discharge varies considerably in the course degree of acidification in single catchm ents is controlled by of the year, and consequently a similar variation can be seen groundwater to a varying extent. The comparison of pH in the degree of its impact on th e final chemical composi input, represented by atmospheric precipitation and tion of surface waters. throughfall, and the output, represented by pH and alkalin More precise data on the relationship between th e quality of surface runoff, was used to assess the deg ree of resis- NGU -BULL 43 9, 2002 - PAGE 10 0 Z. HRKAL, J . BUCHTELE, E. Tl KKANEN, A. KApA HO e J . SANTRUCEK

I.' 18' assessment is based on the median from these arithmetic averages.Castany et al.(1970) are of the opinion that reliable ,....""~.....",,,;;l 00 km N results can be obtained from a series of measurements over = * Poland a minimum of 10 years,whereas currently the longest period of monitori ng extends only over 5 years. Consequently, the presented results should be viewed with some caution.

: ~Prague To incorporate a mo re precise baseflow assess ment , the CZECH R£J>UB LIC SACRAM ENTa soil moi sture accounti ng model has been used (Burnash 1995).This model, in combination with the Anderson snow model (Anderson 1973),enables continuous Slovakia simulations to be carried out over a period of several years. For simulations of the annual rainfall -runoff process, the daily time series are usually used.This is a conceptual water o Krusne hory Mts. • Jezeri catchment balance model with lumped inputs, with the possibility for ~ Sum ava .MIs. • Lysin. and Pluhuv Bor catchment in the Slavkovsky forest its implementation in a semidistributed mode.

"* Na Lizu catchmen t There are altogether six runoff components generated by the mod el: Fig. 1. Location of the test sites. DIR - direct runoff, from those part sof the basin whi ch become impervious after saturation, trvrty of ind ividua l catchments against acid atmospheric IMP - the runoff from the permanently impervious part deposition. It was found that the greater the difference of basin, bet ween th e pH of atmospheric water and in runoff and a SU R - surface runoff, higher alkalinity in surface water,the greater was the buffer IN - interflow, ing capacity of the hydrogeological environment and conse SUP _. supplementary baseflow, i.e., essentially the sea quently its resistance against acid deposition. sonal component of baseflows, However,the basic information concerning the qual ity of PR M - primary baseflow, i.e., the long term part of base gro undwaters was still missing when assessi ng the degree flow. of grou ndwa ter impact on th e final chem istry of surface The daily precipitation and temp erature data typical for water. Mo nito ring of th e groun dwater quality within the th e catchment area have been used as the input data. The LOWRGREP project comm enced in two small catchments model runoff simulation was then calibrated to the mea (Jezeri and Na Lizu, see Fig. 1) as late as in the beginning of sured surface runoff data and the model assessment of the the year 2001.The first results from these localities revealed baseflow compa red with the minimum runoff data. that th e ground water has a more alkaline reaction that the The quantitative part of the data processing was fol surface water and, consequent ly, the gro und water should lowed by analysis of the groundwater impact on the final restrain the impact of 'acid flash floods ' on the pH of surface chemical compos ition of the surface flow. The following waters.On the other hand, it is obv ious that the chemist ry of meth od was used to express the impact of groundwate r on groundwater dep ends closely on the type of host rock envi the chemistry of the surface runoff. The hydrogram of the ronment. ln other words, the pH of groundwater confined to total runoff has been separated into th ree parts. One part granite will differ from th at, for example, in serpentinite. includes th e period when th e ground water discharge con Therefore, th e chemical composi tion of surface water does stitutes at least 50% of th e total runoff and when the not depend only on th e quantitative parameter, i.e.the pro groundwater chemi stry has most effect on the quality of portion of groundwater in th e total runoff, but also on the surface waters.The op posite case is represented by a period quality of th e groundwater. when the contribution of groundwater discharge in tota l The work was carried out in two main stages; firstly, a runoff decreased below 25%. In such a case, the quality of qualitative assessment, which in turn provided data for a atmospheric water plays a decisive role in the quality of the quantitative analysis. The quantitative hydrogeologica l tota l runoff. The period during whic h the contribution of analysis was aimed at an assessment of the contribution of groundwa ter discharge in total runoff varied between 25 baseflow, representing th e gro und water discharge in the and 50% was considered a mixed period in wh ich none of totaI runoff. th e explore d parameters plays a decisive role. Comparison The first estimate of quantity of the baseflow is derived of single pH measurements in surface water with th e corre from Castany's method (Castany et al. 1970).This calculates sponding actual water discharge was used as th e initia l the long-term,gro undwater runoff assessment in two steps. information in th e data processing. Seasonal variati ons in First, the arithmetic average from a 30-day period of mini the pH of atmospheric waters expressed as monthly aver mum total runoff in each hydrolog ical year expresses th e ages were treated separately. lowest year's runoff. Then, the long-term gro undwate r Z. HRKAL, J . BUCHTELE, E. TIKKANEN, A. KApAHO & J . SANTRUCEK NGU-BULL 4 3 9 , 2 002 - PAG E 1 01

Selection of test sites from GEOMON small catch Two pairs of experimental sites have been selected on men t network th e basis of the above-mentioned results.The first is repre The process of acidification in the Czech Re public has been sented by the Jezerf and Na Lizu catchments, characterised monitored in 44 small catchments of th e GEOMON system by an identical geological background but different intensi (Fig. 1). Each of th ese catchments represents a different geo ties of impact of acid deposition.The catchments Lysina and logical, morphological or vegetatio n type. They also differ Pluhuv Bor (Kram & Hruska 1994), an example of extreme from each other in the intensity of impacts of atmospheric differences of lithological com position (acid granite at deposit ion. The method of data collection and assessment Lysina and ultrabasic serpentinite at Pluhov Bor), represent of the hydrochemical budget and element flow corresponds the second pair of test sites. with international standards, and the result s achieved are therefore comparable with localities abroad. Results from selected catchments As follows from Table 1, all studied catchments are The Jezeri and Na Lizu catchments affected by acid atmospheric deposition.The average pH of These two localiti es were selected in orde r to prov ide evi atmospheric precipitation was always below 5,independent dence for the degree of resistance of th e geologica l environ of th e position of th e locality. The long -term arithmetic ment against th e prolong ed impact of acid atmospheric means of pH in throughfall are always lower th an those for deposition . It is not eworthy that both catchments are very precipitation at th e same locality. The differences between similar to each other asfar asth eir geology and morphology the pH of precipitation and throughfall are at th eir highest in are concerned. Bot h lie in mo untain areas at altitudes rang th e vicinity of th e sources of th e pollution . ing between 475 and 924 m (Jezerf) and 828 to 1024 (Na As follows from this five-year mon itorin g and collection Lizu) m a.s.l., and both are underlain by Proterozoic meta of data, th e nature of the geology causes various resistance morph ic rocks, paragneisses in particu lar.However, they dif to acidification. The difference between pH of atmospheric fer considerably in th e intensity of anthropogenic imp act. precipitation or th roughfall and pH of surface runoff was The Jezeri catchment lies in the Krusne hory Mts and used as a parameter expressing th e degree of buffering belongs to th e so-called 'Black triangle', i.e., the territory of against acidification. Using this parameter, the stud ied th e eastern part of Germany (former GDR),the southwestern catchm ent scan be divided into different group scharacteris part of Poland and the northern part of the Czech Republic, tic of low, medium or high resistance (Table 1). with the highest air pollution in Europe.The Na Lizu catch ment is situated in the Sumava Mts.This terr itory is consid Table 1. Re sistance of individual catchmentswithin the GEOMON net ered to be one of the c1eanest parts of the Czech Republic. work to acidification expressed as a relationship between the pH of Even though th e pH values of precipitation on both catch rainfa ll and pH of runoff. ment s are very similar, the large difference in th roughfall pH c values is quite significant. J:= I .2 a.:! a. ~ c~ c ~ Comparable geology and mo rpho logy are responsible GO Cl GO :> ~ :§. 3: 0 3: u '"c ~ ~ ~ GO .;;; for similar runoff conditions in both catchments.The model GO~ GO ~ Cl ..o~ ..0 a. c J: GO J: GO J: '"GO .-0 determination of baseflow in th e Na Lizu catchment (5.2 li s) a. u a. u a. u ~ GO c , c o'"Cl ~:t: ~:t: ~~ correspond s well with th e baseflow estimated using ~15 GO 0 GO 0 '0 GO c :t: c :t: c 'iij :E GO > :> .- :> .- :> GO u Castany's method (4.5 li s).However,some discrepancy exists Cl '" ~ 'C ~ 'C ~ ~ '" Lysina gra nite 4.66 4.26 4.14 -0.12 -0.52 between mod el baseflow estimation (8.0 li s) and Castany's Loukov granite 4.62 4.85 5.14 0.29 0.52 method assessment (4.5 li s) in the Jezerf catchment. The Udvou paragneiss 4.16 4.06 4.8 0.74 0.64 specific runoff of groundwater in the Jezerf and Na Lizu loucek catchments is 3.1 I/s/km2 and 5.2 IIs/km2, respectiv ely. No Uhlirska gra nite 4.4 3.8 5.18 1.38 0.78 differences exist in the groundwater runoff coefficien t Lesni gran ulite 4.49 4.08 5.53 1.45 potok 1.04 (share of groundwater discharge in the tot al runoff); in the Jezeri parag neiss 4.64 3.65 5.76 2.11 1.12 E Jezerfcatchment this is equal to 42%,whereas in the Na Lizu Spalenec granulite 4.81 4.21 6.03 1.82 1.22 :c::l catchment it amounts to 41%. Modry .. marble 4.7 4.6 6.28 E There are considerable differences, however,exists in the potok 1.68 1.58 Cervik sand stone 4.54 4.15 6.12 1.97 1.58 chemical compos ition of surface water (Fig. 2).Although th e Polomka paragneiss 4.82 4.36 6.66 2.3 1.84 average pH values of atmospheric water from the period amphibolite 1995-2000 in both catchments are more or lessidentical (pH Salacova paragnei ss 4.7 4.68 7.05 4.5-4.6),the quality of surface water is very differen t.The pH Lhota 2.37 2.35 ofthe surface runoff in th e Na Lizu catchment,c.7.0,is extra Na Lizu paragneiss 4.45 4.34 6.72 2.38 2.27 Anensky ordinarily stable during the course of th e year. In the caseof gnei ss 4.62 4.16 7.05 2.89 po tok 2.43 the Jezerf catchment, th e average pH is one unit lower.A Pluhuv se rpentinite 4.66 4.49 7.33 2.84 2.67 close relationship between the surface water quali ty and bor water discharge can be observed in both of th ese catch- N GU -BUL L 439, 2002 - PAG E 102 Z . HRKAL, J . BUCHTELE,E. TlKKAN EN , A. K,ii,PAHO & J. SANTRUCEK

J czcr i ca tc hme nt ;'\a Lizu catchmcnt

140 r------,r------, 7.5 60 r------,------~ 7.5 o total runoff ' 20 f------50 . baseflow 7.0 7.0 '00 1------40 6.5 60 f----.,------ff------+-i I--~H ~ l 30 60 6.0 20 40 5.5 10 20

5.0 1995 1996 1997 1998 1999 2000 1995 1996 1997 1995 1999 2000

7.5 7.5 I - pHrunoff - . - pH precip itation I

6.5 6.5

. \ I · pH i \ pH 5.5 I 5.5 t, ; I I I ... \! '; I r: . I " I'. /". j 4.5 . . \ . -v . \ • \ . \ .'- r. ."...... " \ I < ' / ....1 _ I '/ II''v' V 1. .1 '. i ./ ',...... ; It \' 3.5 1995 1996 1997 1995 1999 2000 1995 1996 1997 1998 1999 2000

Fig. 2. Relationship bet ween pH runoff and total runoff and baseflow,Jezeri and Na Lizu catchment s.

ments. A sharp drop in pH occurs when the proportion of of th e same acidity of precip itation at both catchments, the the baseflow decreases.This pH drop is more abrupt at the fast grou ndwater circulation and slow dissolution of soil Jezerfcatchment,down to 5.1-5.5,in comparison with values minerals cannot fully neutral ise the huge amount of acids in of onl y 6.0-6.3 at th e Na Lizu catchment. th e th rough fall. There are two possible solutions to explain these differ ences: The Pluhuv Bar a nd Lysina catchments 1.Soils in both catchments are poor in carbonate miner Both catchments are affected by atmospheric depos ition to als.lf any were orig inally present, they have long since been the same degree. On the other hand, they are markedly dif dissolved due to the infiltration of acid rain. Alum inosilicates ferent as regards geology and hydrogeology. Both catch in soil play the main role in the buffe ring of H+ ions. As their ments lie in the Slavkovsky les upland in western Bohemia dissolution is controlled kinetically, th e best buffering prop (Fig. 1) and are close to each other. Consequently, the cli erties of soil are attained during dry periods , i.e., at the time matic and mor phological conditions and vegetation are of the highest proportion of groundwater in surface runoff. identi cal in both catchmen ts.The area of each catchment is Afte r torrent ial rains or snow melting the surface water is about 0.2 km' and the altitude varies between 800 and 950 more acid, not only for the higher proportion of atmospheric m a.s.l.Both catchments are largely covered by spruce forest. water in surface runoff but also for th e shorter ti me of The basic difference lies in the bedrock geology.The Lysina hydrolysis of soil minerals. catchm ent occurs exclusively in granite of the Kynzvart 2. Throughfalls have a greater impact on groundwater Massif, and the local soil has the character of peat gley.The acidification than precipitations. Acid gases (sulphur and Pluhuv Bor catchment is underla in by serpent inite that is nitrogen oxides, and perhaps also hydrogen fluoride) are to characterised by clayey weathering and very low permeabil a large exte nt adsorbed on th e needles of conifers and ity. together with dry deposit ion are washed out by rain and The results of th e precipitation/ runoff mode l at the infiltrate into soil. Movement of adsorbed gases and dry Lysina and Pluhuv Bor localities are less accurate relative to deposition is strictl y limited and they remain in th e vicinity th ose of the preceding pair of catchments. This fact is of the ir sources. Concentrations of H+ ions are ten times demo nstrated by the lower degrees of correlation bet ween higher in throughfall than in precipitat ions at Jezeri.ln spite the measured and the simulated values, wh ich for the Jezeri Z. HRKAL , J. BUC HTELE, E. T1KKANEN, A. K.4PAHO & J. SA NT RUCEK NG U-B ULL 4 3 9, 2002 - PAGE 10 3

Pluhuv Bur catchme nt Lysina ca tc hme nt

20 ,------:------, 20 ,------, 5.0 o tota l runoff . baseflow

15 7.5 15 - pH runoff 4 5

:I: ~ 10 ~ 10 C.

6 .5

1995 1996 1997 199R 1999 2000 1995 1996 1997 199R 1999 2000

8.0 pi I precipi tation - pll runo ff

7.5 7.5

7 0 7.0

6.5 6.5

60 6.0 :I: :a c. \ 5.5 ./., 5.5 \ . / . 5.0 t , , . . \ 5.0 ~ ~ '\ . .'. ~ ! \ ! \.. .\ 4.5 " j', .' '. _ ~ / ; .J ' J.. 1' <>. ./r.,_, - ...., I \' \ ...... , \. - ...., I 4.5 J ...., .' " 4.0 ' J 4.0

3.5 3.5 1995 1996 1997 199R 1999 2000 1995 1996 1997 1998 1999 2000

Fig. 3. Relation ship between pH runoff and tot al runo ff and baseflow, Pluhuv Bor and Lysina catchments.

and Na Lizu catchments were 0.89 and 0.86, respectiv ely, to th e total runoff, th e geochemical character of local rocks whereasat th e Lysina locality it was only 0.57 and at Pluhuv at Pluhuv Bor is responsible for a high resistivity against acid Bor even less (0.5).The model results suffer from th e poor deposition .Eventhough th e pH of atmospheric water isonly quality of th e abundant data on precipitation. These data 4.7,the surface water maintains a neut ral or slig htly alkaline come from th e area of Marianske Lazne, and are not suffi reaction thr ough th e whole year. The value of pH of surface ciently representative to characterise th e actual precipita water varies between 6.3 and 7.95.The main reason for this tion in th e model region . However, another analysis and is a hydrolysis of basic silicates, e.g.serpenti ne.Their dissolu interpretation of th e results showed th at th e accuracy of ti on is often congruent, much faster than th e weathering of simulation is sufficient in th is case to meet th e project objec acidic rocks,and is able to buffer free H+ ions. Fig. 3 demon tives. strates a fair relationship between the proportion of The low permeability of serpenti nite is reflected in th e groun dwater discharge and th e total runoff, including pH, in runoff conditions of th e Pluhuv Bor catchme nt. The mod surface wate r. elled groundwater discharge constitutes onl y 19% of the The influence of groundwat er on the chemistry of dis total runoff. The specific discharge of th e ground water is charged surface wat er is obvious in the Lysina catchment. A also low, amo unting to 1.8 lis/km' du ring th e monitored similar trend exists here too, i.e., th e close relationship period. between the pH of surface wat er and the proportion of The modelled, specific groundwater runoff in the Lysina groundwater runoff in the total runoff.Although th e loading catchment (3.3 I/ s/ km') includi ng its runoff coefficient (22%) of acid atmospheric deposition in the Lysina catchm ent is is,due to th e completely different geology, greater th an th at similar to th at of th e Pluhuv Bor catchment, th e quality of in th e Pluhuv Bor catchm ent. The mod el determination of surface water in th e catchments is comp letely different.The baseflow in th e Lysina catchment (004 lis) corresponds fairly Lysina runoff is th e most acidic amo ng all th e investig ated well with th e baseflow estimated using Castany's method catchm ent s (Table 1),which is obviously due to th e lithology (0.7 l/s). of the soil profile.The groundwater itself in th e Lysina catch Hydrogeological conditions considerably affect th e ment is very acid because of humi c acids released from th e quality of water drain ed from both catchments. Although gley horizon rich in peat, and consequentl y it is not able to th e contr ibution of gro undwa ter runoff is small with respect neutr alise acid rainfall. This fact is parti cularly evident in NGU -BULL 4 3 9 , 2002 - PAGE 104 Z. HRKAL, J. BUCHTELE, E. TIKKANEN, A. KApAHO & J. SANTRUCEK

80 periods when the proportion of baseflow and the total Industrial SO. 0>- ernessons en 70 runo ff exceeds 50%. During this period whe n the surface c 52 60 0 " - water quality is considerabl y influenced by the chemistry of c -V>.. """1500 hJiJ' ~ 50 groundwater, the average pH in the Lysina catchment is ~ '~ I 40 equal to 4.3. c .Q 13 30 :0 "0 20 Discussion eo, C;; 10 The obta ined results have shown that the degree of protec 0 tive power of the aquifer against acidification depends on a 0 0 1850 1875 1900 1925 1950 1975 2000 number of factors. An important factor is the composition of Year soil which is a product of natural weathering of th e bedro ck and the rate of buffering reactions. Neutral or weakly alka Fig.4. Trends in annual coal pro duction in th e nort hern Czech Republic (Most basin).The small inset box show s the nation wide industrial emis line surface water at Pluhuv Bor shows that basic rocks can sions of SO, (from Novak et al.2000). produce soil with a greater protective power against acidifi cation , in spite of a low proportion of groundwater in the surface water. consequently in a reduction of emissions. The prolonged The second important factor is the distan ce between th e impact of acid atmospheric deposition resulted in a monitored catchmen t and the source of th e po lluti on.Some decrease in the buffer capacity of soil of the Krusne hory acidic gases (SO" NO,) are accumulated on coniferous trees crystalline comple x. Consequently, the drained waters suf by adsorption and then oxidised to strong acids. These fered from a decrease in the concentration of HCO; ions and adsorbed gases, as well as dry deposition, result in the heav a very slow decrease in the conte nts of SO." ions (Fig.5). iest acid loading at any locality. A conspicuous decrease in the buffer capacity of the soil The third factor is the duration of exposure of the soil to layer of the Krusne hory Mts.has recently been suppo rted by acid rain. The soil composition, including the soil buffer a study of the Jezerf catchment (Novak et al. 2000). Isotope capacity, may change during the course of tim e. The first analyses of sulphur in water entering and leaving the catch step is a decalcification fol lowed by a hydrolysis of alumi ment show that the sulphur in the runoff comes from the nosilicates. The mechanism of the hydrolytic reaction may soil horizon . Consequentl y, it represents a secondary source change depending on the acidity of the infiltrating water. that was accumulated over long decades and from which The Krusne hory Mts. have been exposed to the sulphur is gradua lly leached and washed out. The study unfavourable cond itions already since the 19th century. by Novak demonstrates that approximately 30% of the total Mining for lignite and its combustion in power plants content of sulphur in drained water comes from organi c located in the Krusne hory piedmont basins began already matter confined to the upper humic soil horizon . in 1860 and culminated in 1988 (Fig. 4). Only in the last The negative role of the SO/, storage in the weat hered decade has the re been a decline in mining for lignite and substrata has also been demonstrated in th e German exper-

80 18

-+-- runoff 16 70 --- ~ -- ~ ----~ . throug hfall (concentrations) ;\ ~ I: -- precipitation runoff 14 60 --- i -':~ :~ ::- n- I\---- : ... : ' \ . . •• • • through fall (linear trend) ~ precipitation ~ : I 12 c "§, 50 .2 ::J 10 !§ e '0. :; 40 'u :t::"' o 8 '" e E: <:: .s 30 Cl 6 E

20 4

o L- ..,.-_..,-_-,- ~ .....------'. 0 _ 1996 1997 1998 1999 2000 2001

Fig.5. Trend s in SO. concentrat ions in the Krusne hory Mt s., 1996-2001 . Z . HR KAL , J. BUCH T ELE, E. TI KKA N EN , A. KApA HO & J. SANTR UCEK N G U - B U LL 439, 2002 - PAGE 105

imental catchment Lehstenbach in northern Bavaria References (Lischeid et al. 2000, Manderscheid et al.2000). Ande rson, E. A. 1973: NWSRFS Snow accum ulation and ablatio n model ; NOAA Technical Memo rand um NWS HYDRO- 17, us. Dept. Of Rock massifs such as the Krusne hory gneisses, which are Com merce, Silver Spring, Md . more resistant against acidification,can for longer periods of Burna sh, J.C.R. 1995: The NWS River Forecast System - Catchment time protect groundwaters against th e impac t of acid rains. Modeling.ln.:Computer Mod els of Wate rshed Hydrology (ed. by v.P On th e other hand, after a reduction in the impac t of acid Sing h), Water Resources Publi cat ions Colo., p. 3 11-366, ISBN No. 0 deposition, it may take a long time to revert to th e original 918334-9 1-8 Castany, G., Marg at, J, Albinet, M. & Dellaroziere-Bou illin, O. 1970: conditions and the rock massif may, in contrast, maintain or Evaluati on rapide de ressou rces en eaux d'un e region. In: Atti support the acidification. Groundwater acidification in such Conveg no Inte rnaz.sulle acq ue soutterra nee 1970, Palermo. aquifers will not be easily reversed by a decrease in 50/ Fottova, D. 2000: Assessment of th e mass flux in th e sma ll catchment deposition because of the release of previously stored 50/ -. netwo rk GEOMON in Czech repub lic. - MS Annual Report, Czech Geological Survey. Pragu e This appears to be the major reason for the recorded differ Hrkal, Z.1992: Acidification of grou ndwate r in th e Boh em ian Massif. ences in groundwater qual ity between the Jezeri and Na Norges geologiske undersekelse Bulletin.422.97-102. Lizu catchm ents. Hrkal, Z. & Fot tova, D. 1999: Imp act of at mospheric deposit ion on t he Regardless of th e reasonable harmony between corre ground water quality of the Czech Republic (in French). rtydroqeo loqie.2. Orleans, France, 39-45. lated runoff parameters and chemical composi tion of water, Kram, P. & Hruska, J. 1994: Influence of bedrock geo logy on elemen tal an unexpected discrepancy has been recorded in some peri fl-uxes in two forested catch me nts affec te d by high acidic deposi ods. The period between February and June 1997 in the ti on .App lied Hydrogeology, Vol.2, No.2/94 , 50-58 . Jezeri catchment may serve as a typical example of such dis Lischeid, G., Moritz, K., Bitt ersohl ,J., Alewe ll, Ch. & Matz ner,E. 2000: Sinks of ant hropogenic nitr ogen and sulph ate in the Lehste nbach catc h cord. A flood wave associated with snow melting and ment (Ficht el geb irg e) Lessons learnt concerning reversibi lit y.- Silva increased rainfall during March led to a decrease in th e share Gabreta Vol.4 p.4I-S0.Vimperk. of groundwater in th e total runoff.This condition, however, is Mand erscheid, 8., Schwe isser, T., Lischeid, G., Alewe ll, C. & Matzner, E. surprising ly not reflected in a loweri ng of pH in the surface 2000 :Sulfate pools in th e weathered substr ata of a for ested catc h men t.- Soil Scient ifi c Society American Jou rna l 64: 1078- 1082. water that remained more or less stab le at about 5.4. This Novak, M., Kirshner, J,W., Groscheova, H., Havel, M., Cerny, J., Krejci, R., & fact can be explained by th e chemical composi tio n of the Buzek, F. 2000: Sulphur isotope dy nam ics in tw o Cent ral European rainfall because th e pH of atmospheric water in th e same watersheds affec te d by high at mos phe ric dep osition of Sox. period grad ually increased by two orders of magnitude up Geochim ica et Cosmochim ica Acta. Vol. 64. No.3. pp.367-383. 2000 . Elsevier Science. USA. to 6.0.This means th at there was a period when th e infiltra ti on of rainfall had an unexpected alkaline reaction. This example clearly demonstrates th e complexity of th e whole probl em, which cannot be simply reduced to th e rela tionship between th e rock massif and wate r.Chemical analy sis of a water sample from a stream always represents a mix tu re of surface and groundwater or even atmospheric water, and thei r mutual proportions may vary considerably.On the other hand, it is obvio us th at the volume of ground water in runoff remains one of th e most impo rtant factors influenc ing th e finite quality of th e drained water from the catch ment. Therefore, th e degree of grou ndwater acidification should not be neglected.

Ackno wledgements I wo uld lik e to t han k Drs. David Banks and John Mather for t heir helpful and constructive reviews, which sig nificantly improve d an earlier ver sion of this paper; and also Mrs. D.Fottova fo r th e valuable data fro m t he GEOMON netw ork.