Cadernos Lab. Xeolóxico de Laxe ISSN: 0213-4497 Coruña. 2005. Vol. 30, pp. 145-166

Hydrogeological study of A High Mountain Area (Serra da Estrela, Central Portugal): a multidisciplinary approach Estudo hidrogeológico de uma área montanhosa (Serra da Estrela, Portugal central): uma abordagem multidisciplinar

ESPINHA MARQUES, J.1;MARQUES,J.M.2; CHAMINÉ, H.I.3,9;AFONSO,M.J.3; CARREIRA, P.M.4; FONSECA, P.E.5; CABRAL, J5; MONTEIRO SANTOS, F.A.6; VIEIRA, G.T.7; MORA, C.7, GOMES, A.8;TEIXEIRA,J.9; SAMPER, J.10; PISANI, B.J.10; AGUIAR, C.11; GONÇALVES, J.A.12;ALMEIDA,P.G.13; CAVALEIRO, V.13; CARVALHO, J.M.3;SODRÉ BORGES,F.1; AIRES-BARROS, L.2 & ROCHA, F.T.9

Abstract

The results of a preliminary hydrogeological study of the river Zêzere catchment upstream of Manteigas (Serra da Estrela Natural Park, Central Portugal) are presented. In this mountain region, different types of groundwater and surface water (used in several economic activities) occur. The methodology adopted in this study emphasizes the way how Geology, Geomorphology, Geophysics, Geochemistry, Soil Science and Hydrogeology contribute to the description of the hydrological phenomena taking place in the catchment, such as infiltration and aquifer recharge and groundwater flow and geochemistry — allowing to develop bet- ter hydrogeologic conceptual models. The hydrological modelling in course includes the use of the VISU- AL BALAN code, which is being coupled to a GIS. The hydrogeochemical techniques are highlighted as well as its preliminary results concerning major and minor elements. The thermomineral water study includes the identification of the reservoir’s geologic material, the characterization of water-rock interaction and geothermometry.

Key words: Mountain areas, geotectonics, geomorphology, hydrogeology, management of hydric resources, hydrogeochemical techniques, Portugal Centre.

(1) Dep. de Geologia (CGUP), Faculdade de Ciências da Universidade do Porto, Praça de Gomes Teixeira, 4099-002 Porto, Portugal (E- mail: [email protected]). (2) Dep. de Engenharia de Minas e Georrecursos, Instituto Superior Técnico (IST), Lisboa, Portugal. (3) Dep. de Engenharia Geotécnica, Instituto Superior de Engenharia do Porto (ISEP), Portugal. Cadernos Lab. Xeolóxico de Laxe Coruña. 2005. Vol. 30, pp.145-166

(4) Dep. de Química, Instituto Tecnológico e Nuclear (ITN), Sacavém, Portugal. (5) Dep. de Geologia, Faculdade de Ciências da Universidade de Lisboa (LATTEX), Portugal. (6) Dep. de Física, Faculdade de Ciências da Universidade de Lisboa (CGUL), Lisboa, Portugal. (7) Centro de Estudos Geográficos, Universidade de Lisboa, Portugal. (8) Dep. de Geografia, Faculdade de Letras da Universidade do Porto (GEDES), Portugal. (9) Centro de Minerais Industriais e Argilas, Dep. de Geociências, Universidade de Aveiro, Portugal. (10) Escuela Tecnica Superior de Ingenieros de Caminos, Canales y Puertos, Universidad de La Coruña, España. (11) Centro de Investigação da Montanha (CIMO), Escola Superior Agrária de Bragança, Instituto Politécnico de Bragança, Portugal. (12) Dep. de Matemática Aplicada, Faculdade de Ciências da Universidade do Porto, Portugal. (13) Dep. de Engenharia Civil, Universidade da Beira Interior (CECUBI), Covilhã, Portugal. CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 147

INTRODUCTION special emphasis on surface water/groundwater interactions. Maximum subsurface temperatures This paper is strongly connected to one of the experienced by Caldas de Manteigas thermal waters most crucial water-related research issues of this will be recorded by two types of chemical geother- millennium: “High Mountain Areas Hydrology” mometers. Such information will be extremely help- (AURELI, 2002). Special emphasis will be dedicat- ful in geothermal resource evaluation, since it also ed on thermal (e.g., Caldas de Manteigas thermal reflects the depth of groundwater circulation, based waters; CARVALHO, 1996) and non-thermal on an understanding of regional tectonics and geot- groundwaters issuing in Serra da Estrela High hermal gradients. Mountain area (Central Portugal). Surface water sys- In order to address what is happening on the tems will also be considered. The selected study area interrelation between local surface waters (recharge is located in the vicinity of a regional morphostruc- waters) and groundwaters, an integrated multidisci- ture — Bragança–Vilariça–Manteigas fault zone plinary approach is being launched, under the scope [BVMFZ] — on the sector that intersects the cen- of the HIMOCATCH R&D Project “Role of High tral massif of Serra da Estrela. Mountain Areas in Catchment Water Resources, The study area corresponds to the river Zêzere Northern/Central Portugal: Serra da Estrela and Serra do drainage basin upstream of Manteigas village and Marão case studies”. presents specific geomorphologic, climatic and geo- tectonic characteristics which certainly contribute to GEOMORPHOLOGICAL AND CLIMA- control local thermal groundwaters recharge and TOLOGICAL BACKGROUND circulation. Besides the presence of important ther- mal water resources (allowing the installation of the The Serra da Estrela (Fig. 1) is part of the Caldas de Manteigas Spa), the research area is char- Cordilheira Central, an ENE-WSW mountain range acterised by the existence of other strategic ground- that crosses the Iberian Peninsula, and is the highest water resources (e.g., high quality drinking water for mountain in the Portuguese mainland. Associated bottling and domestic use at Manteigas village) to a maximum altitude of 1993m a.m.s.l., this which seem to be also strongly dependent on geo- mountain shows particular climatic and geomor- morphology (recharge areas) and geotectonics phological characteristics that play an important (active faults responsible for groundwater circula- role and impact on the local water cycle, and, partic- tion). Surface water resources have been taken into ularly, on the hydrogeological sub-cycle. consideration because of the basin’s contribution to The Zêzere river catchment upstream of the the storage at the Castelo do Bode large dam, the village of Manteigas and its surroundings corre- main source for Lisbon’s water supply. sponds to an area of ca. 28 km2 with an altitude A broad characterization of the hydrogeologic ranging from 875m a.m.s.l., at the streamflow gauge regime existing in the region is presented, in terms measurement weir of Manteigas, to 1993m a.m.s.l., of its geotectonic, climatic and geomorphologic at the Torre summit. The relief of this sector of features. Relevant aspects of the geophysical behav- Serra da Estrela is dominated by two major plateaus, iour of the BVMFZ, based on seismicity and mag- separated by the NNE-SSW valley of the Zêzere netotelluric studies, are reviewed. The hydrologic river (VIEIRA, 2004; VIEIRA et al. 2005): the importance of soils and land cover is also examined. Torre-Penhas Douradas plateau (1450-1993m Preliminary hydrogeochemical results on the a.m.s.l.), located in the western side, and the Alto da local surface waters and shallow and deep ground- Pedrice–Curral do Vento plateau (1450-1760m waters will be presented and discussed. One of the a.m.s.l.). These plateaus are composite, show flat earliest tasks consisted of the hydrogeologic inven- surfaces at distinct altitudes and present a few wide torying of surface and groundwater ocurrences. valleys. Late Pleistocene glacial landforms and These waters were monitored for one hydrologic deposits are a distinctive feature of the Zêzere year. During this period, two fieldwork campaigns catchment, since the majority of the plateau area were carried out. The hydrochemical results (major was glaciated during the Last Glacial Maximum (e.g., and minor elements) were used to derive informa- DAVEAU et al. 1997, VIEIRA, 2004). tion on groundwater’s geochemical history, with a 148 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005)

The Serra da Estrela climate (DAVEAU et al., tion and aquifer recharge processes, the type of 1997; VIEIRA & MORA, 1998; VIEIRA, 2004) is flow medium (porous vs. fractured), the type of Mediterranean with dry and warm summers; the groundwater flowpaths, or the hydrogeochemistry. wet season extends from October to May, with a The main lithotypes occurring in the region are mean annual precipitation of ca. 2500mm in the (fig. 3): i) Variscan granitic rocks; ii) Precambrian- Torre summit, while the plateaus show more than Cambrian metasedimentary rocks; iii) alluvium and 2000mm. The main precipitation control factors Quaternary glacial deposits. The most important seem to be the slope orientation and the altitude. In regional tectonic structure is the NNE-SSW fact, the western side of the mountain presents a Bragança-Vilariça-Manteigas fault zone, which con- larger number of days with rainfall, but a slightly trols the thermomineral occurrences. lower total amount than the eastern part, which in The Bragança-Vilariça-Manteigas left-lateral turn shows a smaller number of days with rain. A strike-slip fault zone (BVMFZ) is one of the major general raise in the precipitation with the altitude is structures of the late-Variscan fault system network noticeable. However, on a local scale, the distribu- in NW Iberia (Fig. 1). Its reactivation during tion of the precipitation is hard to interpret due to Cenozoic times by the alpine compressive tectonics, its relation to the behaviour of the air mass fluxes together with the reactivation of major ENE-WSW and to complex air divergence and convergence trending reverse faults (such as the Seia-Lousã mechanisms controlled by the mountain morpholo- fault), originated the uplift of the Serra da Estrela gy. Mountain as a horst in a pop-up structure Monthly temperature averages (VIEIRA & (RIBEIRO et al., 1990). A left-lateral movement MORA, 1998) from Penhas Douradas, Lagoa with upthrusting of the eastern block towards Comprida and Penhas da Saúde meteorological sta- WNW can be put in evidence, representing the pre- tions reveal that Serra da Estrela is characterized by dominant tectonic style of the reactivated BVMFZ a simple thermal regime. The warmest month is July in Plio-Quaternary times; fault slip-rates ranging and the coldest is January. Mean annual air temper- from 0.2 to 0.5 mm/year (see CABRAL, 1989, atures are below 7ºC in most of the plateaus area 1995) for the Upper Pliocene to Quaternary tecton- and, in the Torre vicinity, they may be as low as 4ºC. ic activity. The available data concerning snow precipita- Instrumental seismicity associated to the tion are scarce and of poor quality. Nevertheless, BVMFZ demonstrates its present day activity (e.g., the hydrologic importance of snow provides good RIBEIRO, 1984; MOREIRA, 1985; VELUDO, reasons for intensifying the research concerning the 2004). It is presented here the location (Fig. 4) of snowfall and snow cover patterns. So far, the spatial the epicentres of several seismic events that and temporal irregularity of snow related phenom- occurred in the BVMFZ or in associated structures, ena has been referred in earlier studies (e.g., between 1964 and 2004. The epicentres distribution ANDRADE et al., 1992; MORA & VIEIRA, 2004). indicates that whole segments of the BVMFZ are Since the snowfall above 1700m a.m.s.l. may repre- active. The magnitude of the seismic events ranging sent a significant fraction of the annual precipita- from 1 to 6, demonstrate the variable seismic activ- tion, the aquifer recharge from snowmelt will be ity in the area. estimated through the use of isotopic methods and The BVMFZ has been the subject of geophys- geomathematical modelling. ical studies in the last decades, which allowed a bet- ter characterisation of different tectonic structures GEOTECTONICAL, GEOPHYSICAL connected to the BVMFZ. During 1996 and 1998, AND HYDROGEOLOGICAL SETTING thirty magnetotelluric (MT) soundings were carried out in the northern tip of the Vilariça basin (MON- The Serra da Estrela mountain is located in the TEIRO SANTOS et al., 2000, 2002). The interpre- Central-Iberian Zone of the Iberian Massif tation of these data produced an image of the inter- (RIBEIRO et al., 1990). The geological conditions nal electrical resistivity distribution of the basin (fig. represent an essential part of the hydrologic setting 5). The main characteristics of the MT models and since they impose some of the main features of the the interpretation of its features are as follows hydrogeologic systems (Fig. 2), such as the infiltra- (MONTEIRO SANTOS et al., 2000): i) the upper- CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 149

Fig 1. Morphotectonic features from Central Portugal, Serra da Estrela mountain region. Major faults: PCTSZ – Porto- Coimbra-Tomar strike-slip shear zone; VCRFZ – Vigo-Vila Nova de Cerveira-Régua fault zone; VRPFZ – Verin-Régua- Penacova fault zone; BVMFZ – Bragança-Vilariça-Manteigas fault zone; SLFZ – Seia-Lousã fault zone. most lithologies of the Parautochthonous show an tary units. The models suggest that this conductor is average resistivity of 100-200 ohm m; ii) the transi- deeper within the northeastern part of the basin, tion zone between Lower Allochthonous- whose floor tilts towards the north and southwards; Parautochthonous and the Parautochthonous is v) at depths of 2km, in the western part of the stud- represented by a 350m thickness layer with a resis- ied area, the resistivity increases up to values greater tivity of 200-500 ohm m; iii) the Autochthonous than 4000 ohm m. We interpret the deepermost part metasediments (dominated by quartzites, pelites and of this layer as the Iberian gneiss basement; vi) the graywackes), display distinct resistivity values, rang- resistivity gradients revealed in the upper crust were ing from 500 to 4000 ohm m, defining different lay- associated with the main fault that controls the for- ers; iv) the sedimentary filling of the tectonic basin mation and evolution of the tectonic basin in is a good conductor, presenting average resistivities Quaternary times. of 20-100 ohm m. The low resistivity of this zone The studies in course (e.g., ESPINHA MAR- is mainly due to the water content of the sedimen- QUES, in prep.; ALMEIDA, in prep.; AFONSO et 150 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005) al., 2005) are taking into consideration the hydroge- words, that infiltrates — as consequence of a pre- ologic importance of the fracture network, with cipitation event, directly influences the aquifer particular emphasis on understanding the fault sys- recharge as well as the short-term stream response. tem geometry, the processes of opening and sealing The major factors usually pointed out as affecting of fractures, the stress fields and the seismic pump- infiltration — and runoff — are the amount and ing occurrence. For this purpose, scale studies from characteristics of precipitation (or irrigation), the outcrop to satellite imagery will be accomplished. In soil physical and chemical features (e.g., saturated particular, remote sensing and tectonic analysis will hydraulic conductivity at the surface, clay mineralo- be used to characterise the main tectonic linea- gy, presence of water-repellent substances), previ- ments. Data obtained from computational analysis ous soil water saturation, surface slope and rough- of digital terrain models and from geological/geo- ness, land cover and amount of evapotranspiration morphological structures are complemented by (e.g., DINGMAN, 1994). field data acquired using traditional mapping meth- The soil features in this sector of Serra da ods. Results will be integrated in a Geographic Estrela result from the way how the formation fac- Information System (GIS). tors act. The soil system is described (JENNY, In order to understand groundwater’s recharge 1994), by the following factors: parent material, cli- and circulation related phenomena (particularly mate, topography, organisms and time. Human thermomineral waters), infiltration and recharge action is often referred as an additional item to be areas are being identified and delimited by means of considered (e.g., GAUCHER 1981). The soil study hydrogeochemical, isotopic (MARQUES et al., included several fieldwork campaigns carried out 2005) and hydrogeomorphologic criteria. In addi- through 2004. During these campaigns, soil samples tion, groundwater circulation paths must be charac- were collected (Fig. 6) in order to obtain a physical, terised and groundwater contribution to streamflow chemical, geochemical and mineralogical characteri- estimated. Thus, lithology, water chemistry, mor- sation. phostructure, climate, soil type and land cover The main soil physical properties considered in should altogether be considered. the study are texture, structure, bulk density, particle An important issue connected to the infiltration density, porosity, colour, water retention and and aquifer recharge processes consists in the iden- hydraulic conductivity. Other properties considered tification of areas of prevailing fractured or porous are pH, organic matter content, cation exchange circulation mediums. In particular, the porous medi- capacity and exchangeable cations. Soil mineralogy ums are dominant in the alluvium and Quaternary focuses on clay mineralogy; soil geochemical analy- glacial deposits as well as in the most weathered sis considered 36 chemical elements. The soil granites and metasedimentary rocks. Porous medi- hydraulic conductivity was studied using the Guelph ums usually occur at shallower depths (typically less permeameter field method. During 2004, around 50 than 50m). On the other hand, fractured mediums tests were conducted in most of the soil sampling occur in poorly weathered granitic or metasedimen- sites. tary areas. Such mediums may be present very close The soil clay mineralogy and geochemistry is to the surface (especially on granitic outcrop domi- closely related to a detrital origin (absolute predom- nated areas, with thin or absent sedimentary cover) inance of quartz, mica/illite and feldspars; more or below the referred porous geologic materials. significant values of Al, Fe and K). Nevertheless, some distinctive features are evident, such as: sam- INFILTRATION, SOILS AND LAND ples related to granites show higher amounts of COVER phyllosilicates (but show a decrease in illite whereas kaolinite increases); samples showing more signifi- Soils are especially relevant in hydrologic stud- cant values of Al, Fe and K are those related to ies, as they contribute to control both the volume granites and to glacial deposits located on slope and the water chemistry in hydrologic systems. The and/or base of slope sites, whereas Ca shows high- amount of water that moves from the topographic er values in samples related to glacial deposits locat- surface into the soil or the rock masses — in other ed on slope, base of slope and plateau sites (ROCHA et al., 2005). CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 151

Fig 2. Some aspects of the study area and the research work: a) Zêzere valley; b) Nave de Santo António, Cântaro Magro and Cântaro Gordo area; c) hydrochemical field analysis; d) snow cover at Nave de Santo António; e) soil profile in a granitic area; f) glacial deposit at Manteigas. 152 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005)

Fig 3. Geological map of the Serra da Estrela region, Central Portugal (simplified after OLIVEIRA et al., 1992)

Land cover has an important impact on a num- the unsaturated zone and the aquifer, will provide ber of hydrological processes. The amount of infil- estimates of aquifer recharge in the catchment. For tration varies considerably depending on whether this purpose, the computer code VISUAL BALAN vegetation is or is not present and on vegetation V2.0, developed at the University of A Coruña type. In the study region, several categories of land (ETSI), will be used (SAMPER et al. 1999, 2000, cover such as grassland, heathland, or forests are 2005). more favourable for infiltration than granitic or VISUAL BALAN V2.0 is a lumped hydrologic metasedimentary outcrop areas or exposed soils, code which solves the water balance equation in the mostly because of a runoff slowing effect. soil, the unsaturated zone and the aquifer. The code It is included (Table 1) a preliminary description requires only a few parameters and incorporates of some hydrologically relevant soil system charac- user-friendly interfaces for data input and post-pro- teristics at each sampling site. Soil classification cessing of results. It evaluates hydrologic compo- according to the FAO-UNESCO criteria is being nents in a sequential manner. In addition to the produced but is not yet available. water balance equation in the upper soil, the code also solves the water balance equations in the unsat- HYDROLOGICAL MODELLING urated zone and in the underlying aquifer. This allows the computation of daily groundwater levels Hydrological models are being used to evaluate as well as basin water discharge rates. Computed water resources in the basin. These models, which heads and streamflows can be compared to meas- solve the water balance equations in the upper soil, ured values for the purpose of model testing and CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 153

Fig 4. Seismicity connected to the Manteigas-Vilariça-Braçança fault system and surrounding structures, recorded between 1964 and 2004 (VELUDO, 2004). Triangles – seismic stations (FOZC – Vila Nova de Foz Côa; MTE – Manteigas; PBRG – Bragança; PVRL; Vila Real); circles – earthquake epicentres. 154 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005)

Fig 5. 2-D resistivity model obtained from joint TE-TM mode data (apparent resistivities and phases) collected across the Vilariça graben. The line marked with a b shows the approximate location of the basin (MONTEIRO SANTOS et al., 2000).

Table 1 — Main soil system characteristics of the studied area. CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 155

Fig 6. Water and soil sampling points; major ion hydrogeochemical results. 156 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005) calibration. VISUAL BALAN V2.0 accounts for Serra da Estrela Natural Park will certainly help to irrigation sources and return flows. It also considers develop and improve conceptual models compris- snow precipitation, melting and runoff. Two ing the main components of hydrogeologic sys- advanced options have been incorporated recently: tems, namely: i) recharge areas, ii) groundwater cir- i) Automatic parameter estimation using groundwa- culation zones and iii) discharge areas. Such an ter level and streamflow data and ii) Sensitivity approach is particularly suitable for handling a great analyses of hydrologic components to model parameters. variety of data that must be spatially integrated in a Since VISUAL BALAN V2.0 is a lumped coherent manner (Fig. 8) which, as referred earlier, hydrologic model, it is more suited for homoge- is the case of the study area. neous and small basins. Therefore, as the studied The methodological framework followed in this basin is complex and exhibits a large spatial variabil- research offers important benefits for regional ity affecting most of its features, the basin is subdi- water resources management, especially in what vided into several hydrologically homogeneous sub- concerns to groundwater exploration, to the defini- basins where independent water balances can be tion of well-head protection areas and to the assess- performed. Results of each sub-basin are subse- ment of aquifer pollution vulnerability. quently lumped to obtain the results for the whole basin. VISUAL BALAN V2.0 has recently been cou- The hydrological model is based on tempera- pled to a GIS (SAMPER et al., 2005). This proce- ture and precipitation data from Penhas Douradas dure was carried out by extending the capabilities of meteorological station (1383m a.m.s.l.). The possi- the original code. For that purpose, a pre-processor bility of using auxiliary data from other stations has been developed as an input interface to VISU- such as that at Manteigas (815m a.m.s.l.) is open to AL BALAN V2.0. Beginning from a digital eleva- consideration. A preliminary definition of sub- tion model and using the geomorphologic data in basins, closely related to the definition of hydroge- the GIS the pre-processor interprets model input omorphological units, is illustrated (Fig. 7). Several data: sub-basin delineation, drainage network, mor- criteria are being considered in order to accurately phologic parameters (average slope, characteristics, represent the complexity of the hydrological sys- soil type and land cover/land use). Additional tem: geology, geomorphology, hydrogeology, cli- improvements are expected to be able to perform mate, soil type and land cover. During this delimita- modelling of complex basins considering the spatial tion process, special attention has been paid to the variation of model parameters (distributed parame- available geological and geomorphological maps ter model) and surface runoff propagation. GIS will (e.g., FERREIRA & VIEIRA 1999; VIEIRA, 2004). provide average parameter values for each sub-basin Due to the strong temperature and precipitation delineated by the pre-processor in the first step. vertical gradients, climate data has to be analyzed Available meteorological data from different sta- carefully. Preliminary model testing and calibration tions will be processed in the GIS to create maps has been performed by comparing computed that describe the spatial variability of weather vari- streamflows to measured values from Manteigas ables. This information will be then processed to streamflow gauge station. obtain series of average values for each sub-basin, in the same way as morphologic parameters. GEOGRAPHICAL INFORMATION Connectivity between sub-basins will be established SYSTEM APPLICATION and so flow accumulation will be calculated for each sub-basin. A further step is planned in the future to Geographical Information System technology be able to apply the balance equations to smaller is particularly appropriate for handling hydrogeo- areas, taking full advantage of the capabilities of the logical data (SINGHAL & GUPTA, 1999). The GIS. development of a GIS applied to this sector of CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 157

Fig 7. Sub-basin limits: 1 – Eastern plateau; 2 – Zêzere valley eastern slopes; 3 – Lower Zêzere valley floor; 4 – Nave de Santo António col; 5 – Upper Zêzere valley floor; 6 –Zêzere valley western slopes; 7 – Cântaros slopes; 8 – Lower western plateau; 9 – Upper western plateau.

HYDROGEOCHEMICAL APPROACH resources of a potential area for development. In the present chapter, some of the geochemi- Surface manifestations of thermomineral cal techniques employed in thermomineral water waters circulation are a subject of great scientific investigations in the Caldas de Manteigas area, in and economic interest. Thermomineral waters and order to update local and/or regional conceptual shallow cold groundwaters spurting out in the same circulation models, are outlined. Preliminary results area should be observed and studied in detail, as on the major and minor element composition of they provide a significant amount of information at local surface waters and of shallow and deep relatively low costs. This information may be used groundwaters will be presented and discussed. in the appraisal of the thermomineral water The hydrogeochemical study in course illus- 158 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005)

Fig 8. Data integration in a GIS (modified after SINGHAL & GUPTA, 1999). trates the considerable number of ground and sur- collection. The following methods were applied for face water types coexisting in a relatively restricted chemical analyses performed at the Laboratório de area, thus reflecting the hydrologic complexity of Mineralogia e Petrologia of Instituto Superior the basin. Técnico (LAMPIST, Lisbon): atomic absorption In order to embrace the various hydrologic sub- spectrometry for Ca and Mg; emission spectrometry systems occurring in the region, three main water for Na, K, Li, Rb and Cs; colorimetric methods for categories were considered: i) shallow cold ground- SiO2,Fetotal, F and Al; ion chromatography for waters; ii) deep thermomineral groundwaters and SO4,NO3 and Cl; potentiometry for alkalinity, here iii) surface waters from the river Zêzere and its trib- referred to as HCO . Representative data of the utaries. 3 A hydrogeologic inventory and the definition of waters sampled during the 1st (September 2003) a network of surface and ground water monitoring and 2nd (April 2004) fieldwork campaigns are pre- points (springs, boreholes and streams) was carried sented (tables 2 and 3). out (Fig. 6). Subsequent fieldwork campaigns were At Caldas de Manteigas area, the thermominer- conducted in September 2003 and April 2004 in al waters (with output temperatures around 45ºC) order to collect water samples for chemical and iso- are characterised by the following main features: topic (oxygen-18, deuterium and tritium) analysis. i) relatively high pH values ( ∼ 9). The resulting hydrogeochemical information (major ii) TDS values usually in the range of 160 to and minor element composition) is being applied to 170 mg/L. estimate geochemical evolution of groundwater, iii) HCO is the dominant anion. including its origin and interaction between water 3 and the aquifer rock minerals. iv) Na is the dominant cation. Temperature (ºC), pH and electrical conductivi- v) the presence of reduced species of sulphur ∼ ty (ìS/cm) of the waters were determined in the (HS- ∼? 1.7 mg/L). field. Total alkalinity was measured a few hours after vi) high silica values (usually around 50 mg/L) CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 159

representing a considerable percentage of The role of CO2 is strongly related with the pH + - total mineralization. of the solution since 2H + 2HCO3 is, in fact, vii) high fluoride concentrations (up to 7 2H2CO3,or 2CO2 + H2O. mg/L). - Group II) This second group of waters encloses surface (Zêzere river – sampling points As indicated by the chemical composition of close to Covão da Ametade and Jonja stream) the Caldas de Manteigas thermomineral waters, the and shallow groundwaters (Nave de Stº reservoir rock should be mainly granite, being the António and Espinhaço de Cão spring waters). thermomineral waters mineralization strongly dom- This group of waters is also characterized by inated by the hydrolysis of plagioclases. The strong relatively low mineralization, but presents clear HCO3-Na signatures of Caldas de Manteigas ther- Na-Cl geochemical signatures. The Na-Cl facies momineral waters can be clearly seen in the Stiff found within some of the waters of this group diagram (Fig. 6). could be ascribed to the local use of NaCl to Concerning the non-thermal waters sampled in promote snowmelt in the roads during the the studied region (surface waters and shallow Winter season. These geochemical signatures groundwaters), two groups can be defined through can be clearly seen in the Piper diagram (see Fig. the geochemical signatures derived from the Stiff 9). In the case of Espinhaço de Cão spring the and Piper diagrams (Figs 6 and 9), namely: rather different Stiff diagrams indicates that - Group I) This first group encloses the so- water chemistry could be strongly controlled by called “normal” surface (Zêzere river – sam- other sources of anthropogenic contamination pling point close to the Caldas de Manteigas rather than the local use of NaCl for snow spas) and shallow groundwaters (Covão do Boi, melting, since there are important differences in Jonja, Paulo Luís Martins, Bisa and N. Srª. de the Na/Cl ratios from the 1st to the 2nd field Fátima spring waters). All of these waters work campaign. The different geochemical sig- belong mainly to the HCO3-Na facies (a relative- natures found in Espinhaço de Cão spring can ly higher Ca concentration was found only in be clearly detected in the field through the high- Bisa spring during the field work campaign of er electric conductivity values (see Tables 2 and 3). September 2003), displaying different total dis- solved solids (TDS) values. The lower mineral- Chemical geothermometry is one of the most ization found in Zêzere river and Covão do Boi, important geohydrologic tools in the exploration of Jonja and Paulo Luís Martins spring waters thermomineral water resources. The main objective reflects low water-rock interaction associated of geothermometric interpretation is to use the with short surface/underground circulation chemistry of hot springs in order to estimate chem- paths. These spring waters could be considered ical and physical properties of the reservoir fluid. as good signatures of local recharge. The rela- This methodology depends upon the temperature tively higher mineralization detected in Bisa and dependence of the concentrations of certain N. Srª. de Fátima spring waters could be species, the chemical equilibrium between minerals ascribed to long shallow underground flow and water and various chemical reactions. During path, allowing higher water-rock interaction. It the last two decades many chemical geothermome- should be stated that Bisa spring is located in a ters have been proposed, both qualitative and quan- forestated area where soils are enriched in titative. Those used in this paper include the chal- organic matter. This important source of addi- cedony and quartz geothermometers (FOURNIER tional CO2 could be responsible for a higher & TRUESDELL, 1974 in: FOURNIER, 1977; water-rock interaction, expressed in the follow- TRUESDELL, 1975 – cooling by conduction), the ing equation: feldspar (Na/K) geothermometer (WHITE & ELLIS, 1970 in: TRUESDELL, 1975) and the + - 2 2NaAlSi3O8 + 9H2O + 2H + 2HCO3 = K /Mg geothermometer (GIGGENBACH, 1988). + - The results obtained are presented (see Table 4). Si2O5Al2(OH)4 + 2Na + 2HCO3 + 4 Bearing in mind that the reservoir fluid may H4SiO4 160 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005) CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 161

Fig 9. Major ion hydrogeochemistry: a) September 2003 campaign; b) April 2004 campaign. 162 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005) become mixed with cold groundwater at shallow In future studies involving the interpretation of levels and could change its chemistry by leaching chemical geothermometry special emphasis will be and reaction with wall rocks on the way to the sur- put on the evaluation of the overall chemical char- face, the results obtained from the application of acteristics of the waters, as they correlate to equilib- several chemical geothermometers should be inter- rium with alteration minerals. preted with great caution. In the near future, the results of chemical geothermometry will be corre- CONCLUDING REMARKS lated with the results from the geophysical and geo- tectonical surveys that will be performed in the Crystalline rocks, particularly granites and region, under the scope of the HIMOCATCH metasedimentary units, dominate the Serra da R&D Project. Such information will be extremely Estrela Mountain region. The hydrogeological char- helpful in thermomineral water resources evalua- tion, since it also reflects the depth of groundwater acterisation of this kind of hard-rocks is complex, circulation, based on an understanding of regional due to the high heterogeneity and anisotropy of the tectonics and geothermal gradients. fracture network that stores and conducts the water. Using reservoir temperatures given by the It is usual to consider that the flowpaths are mainly quartz geothermometer (applied to Caldas de governed by the fissured medium hydraulic conduc- Manteigas AC2 and AC3 borehole waters), and con- tivity, faulting and weathering, resulting on non- sidering a mean geothermal gradient of 30ºC/km, continuous productive zones. Nevertheless, it is we can estimate a maximum depth of about 3.2km clear that in the Variscan Iberian Massif, lithology reached by the Caldas de Manteigas thermomineral and structure play a major role on the productivity water system. This value was obtained considering that: of regional hydrogeological units and related water wells. depth = (T - T ) / gg = (100 - 5) / 30 = 3.2 r a Some differences detected in local shallow cold km groundwater’s characteristics could be the result of water circulation paths varying in length and resi- where T is the reservoir temperature (ºC), T r a dence time and/or anthropogenic contamination, the mean annual temperature (ºC) and gg the geot- such as the local use of NaCl for snow melting. The hermal gradient. combined chemical and isotopic data suggests that Geochemical data of Fonte Santa thermal the Caldas de Manteigas thermomineral waters spring waters seems to corroborate a mixing process with local shallow groundwaters. This trend could be derived from regional groundwater requires the shallow groundwaters diluting Fonte sources. The quartz geothermometer indicates Santa thermal spring waters to be derived from local reservoir temperatures of approximately 100ºC. infiltration. The mixing process should be responsi- Combining information ascribed to the geo- ble for the higher Ca and Mg dissolution. Effectively, the Ca and Mg concentrations are lower chemical and isotopic signatures of groundwaters, in AC2 and AC3 thermal borehole waters, whereas hydrogeologists can strongly support their conclu- these metals are higher in Fonte Santa thermal sions on the origin of waters and recharge areas, spring waters. Furthermore, Ca and Mg are very groundwater quality and contaminant processes, often added to cooled waters by a reaction with the water-rock interactions occurring at depth and rock, as stated by several authors, particularly with resource renewability. regard to waters of the French Massif Central (e.g., MICHARD et al., 1978, 1981; CRIAUD & FOUIL- LAC, 1986). Further fieldwork campaigns will clari- ACKNOWLEDGEMENTS fy the above mentioned hypothesis, based on sys- In order to address the specific scientific issues tematic isotopic (18O and 2H) signatures of the on what is happening about the interrelation waters. CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 163 between local surface waters (recharge waters) and edge to Geodyn – Present to Past POCTI-ISFL-5- groundwaters, an integrated multidisciplinary 32. The authors acknowledge Prof. L. C. Gama approach is being launched under the scope of the Pereira (Coimbra) for detailed reviews that helped HIMOCATCH Project “Role of High Mountain Areas to improve the clarity of the manuscript. in Catchment Water Resources, Northern/Central Recibido:9/5/2005 Portugal”, granted by the Portuguese Foundation for Aceptado:24/6/2005 Science and Technology (FCT), contract Nr. POCTI/CTA/44235/2002. PEF and JC acknowl-

Table 4 - Reservoir temperatures (ºC) of Caldas de Manteigas thermomineral waters, estimated from chemical geothermometry. 164 Espinha Marques et al. CAD. LAB. XEOL. LAXE 30 (2005)

REFERENCES ca da glaciação da Serra da Estrela. Estudos do Quaternário, Lisboa, 1: 41-51. AFONSO, M. J.; ESPINHA MARQUES, J.; MAR- DINGMAN, S. L. (1994). Physical hydrology. Prentice QUES, J. M.; CARREIRA, P.; CARVALHO, J. Hall, New Jersey, 575 pp. ESPINHA MARQUES, J. M. (in prep.). Contribuição M.; MARQUES DA SILVA, M.; SAMPER, J.; para o conhecimento da hidrogeologia da região do BORGES, F. S.; ROCHA, F. T.; FONSECA, P. Parque Natural da Serra da Estrela (sector de E.; GOMES, A.; ARAÚJO, M. A.; TELES Manteigas–Nave de Sto. António–Torre). VIEIRA, G.; MORA, C.; TEIXEIRA, J.; Universidade do Porto (PhD Thesis, in prepara- ALMEIDA, P. G. & CHAMINÉ, H. I. (2005). tion). Hydrogeology of hard-rocks from two key-sec- FERREIRA, N. & VIEIRA, G. (1999). Guia tors in the Portuguese Iberian Massif: examples geológico e geomorfológico do Parque Natural from Porto urban area and Serra da Estrela da Serra da Estrela. Locais de interesse geológi- mountain region. Proceedings Water in Celtic co e geomorfológico. Parque Natural da Serra da Estrela. Edição Instituto da Conservação da Countries: Quantity, Quality and Climate Variability, Natureza/Instituto Geológico e Mineiro, “The Fourth Inter-Celtic Colloquium on Hydrology and Lisboa. 112 pp., 2 mapas. Management of Water Resources”, Univ. Minho, FOURNIER, R. O. (1977). Chemical geothermo- Guimarães. (in press) meters and mixing models for geothermal sys- ALMEIDA, P. G. (in prep.). Estudo tectonoestrutural e tems. Geothermics, 5: 41-50. caracterização geotécnica do maciço granítico Central da GAUCHER, G. (1981). Les facteurs de la pédogénèse. G. Serra da Estrela. Um exemplo da utilização multidis- Delotte, Dison, 730pp. ciplinar numa abordagem holística da Geologia.UBI, GIGGENBACH, W. F. (1988). Geothermal solute Covilhã (PhD Thesis, in preparation). equilibria: derivation of Na-K-Ca-Mg geoindi- ANDRADE, E.; MORA, C.; NEVES, M. & VIE- cators. Geochim. Cosmochim. Acta, 52: 2749-2765. IRA, G. (1992). Desportos de Inverno na Serra JENNY, H. (1994). Factors of soil formation, a system of da Estrela. Contribuição para o estudo da sua quantitative pedology. Dover Publications, Inc.. viabilidade. Finisterra, Lisboa, 27 (53-54): 187- New York, 281pp.. 193. MARQUES, J. M.; CARREIRA, P. M.; ESPINHA AURELI, A. (2002). What’s ahead in UNESCO’s MARQUES, J. M.; CHAMINÉ, H. I.; FON- International Hydrological Programme (IHP VI SECA, P. E.; MARTINS CARVALHO, J. & 2002-2007). Hydrogeology Journal, 10: 349-350. CAVALEIRO, V. (2005) Pattern and spatial CABRAL, J. (1989). An example of intraplate neo- extent of thermal water flow in a high moun- tectonic activity, Vilariça Basin, Northeast tain area (Serra da Estrela, Central Portugal): a Portugal. Tectonics, 8: 285303. preliminary isotopic approach. Abstract, 6th CABRAL, J. (1995). Neotectónica em Portugal International Symposium on Applied Isotope Continental. Memórias do Instituto Geológico Geochemistry, 11-16 September 2005, Prague, e Mineiro, Memória 31, Lisbon, 265 pp. Czech Republic. (submitted) CARVALHO, J. M. (1996). Mineral water explora- MICHARD, G.; EVRARD, M.; FOUILLAC, C. & tion and exploitation at the Portuguese LAMBRET, B. (1978). Acquisition des ions Hercynian massif. Environmental Geology, 27: 252- alcalino terreux par les eaux carbogazeuses. 258. Earth Planet. Sci. Lett., 41: 170-174. CRIAUD, A. & FOUILLAC, C. (1986). Étude des MICHARD, G.; FOUILLAC, C.; GRIMAUD, D. & eaux thermominérales carbogazeuses du Massif DENIS, J. (1981). Une méthode globale d’esti- Central Français. II. Comportement de quel- mation des températures des réservoirs ques métaux en trace, de l’arsenic, de alimentant les sources thermales. Exemple du l’antimoine et du germanium. Geochim. Massif Central Français. Geochim. Cosmochim. Cosmochim. Acta, 50: 1573-1582. Acta, 45: 1199-1207. DAVEAU, S.; FERREIRA, A. B.; FERREIRA, N. MONTEIRO SANTOS, F. A.; ALMEIDA, E.; & VIEIRA, G. (1997). Novas observações acer- CAD. LAB. XEOL. LAXE 30 (2005) Hydrogeological study 165

MATEUS, A.; MATIAS, H.; MATOS, L. & logic models and Geographic Information MENDES-VICTOR, L. A. (2000). Systems for water resources evaluation: GIS- Magnetotelluric study of a Plio-Quaternary tec- VISUAL-BALAN and its application to tonic depression: the Vilariça basin (NE Atlantic basins in Spain (Valiñas) and Portugal Portugal). Journal of Applied Geophysics, 44, 1-14. (Serra da Estrela). Proceedings The Fourth Inter- MONTEIRO SANTOS, F. A.; MATOS, L.; Celtic Colloquium on Hydrology and Management of ALMEIDA, E.; MATIAS, H.; MATEUS, A. & Water Resources, Universidade do Minho, MENDES-VICTOR, L. A. (2002). Three- Guimarães. (in press). dimensional magnetotelluric modelling of the SAMPER,J.;HUGET,L.;ARÉS,J.& GARCÍA,M. Vilariça depression (NE Portugal). Journal of A. (2000). Modelos interactivos de balance Applied Geophysics, 49: 59-74. hidrológico: aplicación a cuencas de Galicia. In: MORA, C. & VIEIRA, G. T. (2004). Balance radi- SAMPER, J.; LEITÃO, T.; FERNÁNDEZ, L. activo de los altiplanos de la Sierra de Estrella & RIBEIRO, L. (eds.), Jornadas Hispano-Lusas (Portugal) en una mañana de invierno. sobre ‘Las Aguas Subterráneas en el Noroeste Metodología y primeros resultados. Bol. R. Soc. de la Península Ibérica’. Textos de las Jornadas, Esp. Hist. Nat. (Sec. Geol.), Madrid, 99 (1-4): 37- Mesa Redonda y Comunicaciones. A Coruña. 45. AIH-Grupo Español & APRH. Publicaciones MOREIRA, V. S. (1985). Sismotectonics of ITGE, Madrid. pp. 211-219. Portugal and its adjacent areas in the Atlantic. SINGHAL, B. B. S. & GUPTA, R. P. (1999). Applied Tectonophysics, 117: 85-96. hydrogeology of fractured rocks. Kluwer Academic OLIVEIRA, J. T.; PEREIRA, E.; RAMALHO, M.; Publishers, Dordrecht, 400 pp. ANTUNES, M. T. & MONTEIRO, J. H. TRUESDELL, A. H. (1975). Summary of Section [coords.] (1992). Carta Geológica de Portugal escala III. Geochemical techniques in exploration. 1/500000, 5ª edição. Serviços Geológicos de Proc. Second United Nations Symposium on the Portugal, Lisbon. Development and Use of Geothermal Resources.San RIBEIRO, A. (1984). Neotectonique du Portugal.Livro Francisco, California. Edited by Lawrence de Homenagem a Orlando Ribeiro, 1º Vol., Berkley Laboratory, University of California: Centro de Estudos Geográficos, Univ. Lisboa, 53-79. 173-182. VELUDO, I. (2004). Monitorização da actividade RIBEIRO, A.; KULLBERG, M. C.; KULLBERG, J. sísmica no território do parque arqueológico do vale do C.; MANUPPELLA, G. & PHIPPS, S. (1990). Côa. Relatório de Estágio. Fac.Ciências Univ. A review of Alpine tectonics in Portugal: fore- Lisboa. (Relatório Inédito) land detachment in basement and cover rocks. VIEIRA, G. T. & MORA, C. (1998). General char- Tectonophysics, 184: 357-366. acteristics of the climate of the Serra da Estrela. ROCHA, F.; INÁCIO, M.; ESPINHA MARQUES, In: VIEIRA, G.T. (ed.), Glacial and Periglacial J.; GONÇALVES, P.; AGUIAR, C. & Geomorphology of the Serra da Estrela. Guidebook for CHAMINÉ, H. I. (2005). Clay mineralogy and the field-trip, IGU Commission on Climate Change and geochemistry of soils from Serra da Estrela Periglacial Environments. CEG and Department of Natural Park (Central Portugal): preliminary Geography, University of Lisbon, pp. 26-36. results. Proceedings do VIII Congresso de Geoquímica VIEIRA, G. T. (2004). Geomorfologia dos planaltos e dos Países de Língua Portuguesa, Universidade de altos vales da Serra da Estrela. Ambientes frios do Aveiro. (in press). Plistocénico Superior e dinâmica actual. Universidade SAMPER,J.;HUGET,L.;ARÉS,J.& GARCÍA,M. de Lisboa, 724p. + 1 mapa. (PhD Thesis). A. (1999). Manual del usuario del programa VIEIRA, G. T.; JANSEN, J. & FERREIRA, N. (sub- Visual Balan V.1.0. Publicación Técnica de ENRE- mitted). Environmental setting of the Serra da SA, 5 (99): 1-132. Estrela, Portugal: a short-note. In: BUNCE, R., SAMPER, J.; GARCÍA VERA, M. A.; PISANI, B.; HOWARD, D. & PINTO CORREIA, T. (eds.), VARELA, A.; LOSADA, J. A.; ALVARES, D. & Landscape ecology series, IALE UK. ESPINHA MARQUES, J. (2005). Usin g hydro-