Vízügyi Közlemények, 2001 (83. évfolyam)
4. füzet - Rövidebb tanulmányok, közlemények, beszámolók
612 Ga l sa A.—Salát P.-Cserepes L. Qualified evaluation of the hydraulic conductivity of a geological profile of the Hungarian low land Alföld, utilising measured heads of wells "V by Attila GALSA, Dr. Péter SALÁT and Dr. László CSEREPES geophysicists The paper describes a quality-controlled inverted procedure, 1 which determines the horizontal and vertical hydraulic conductivity of geological layers using the measured piezometric levels of deep wells. The measurements were made in a geological profile of the Alföld. The hydrogeological model made for the profile investigated is shown in Figure 2. Similarly to other systems of the Alföld, the flow takes place in the Pleistocene sediments, where the hydraulic conductivity is known to be the highest. Preliminary measurements indicated that the permeability of the underlying red-clayey Levantine layer is probably lower. The hydraulic conductivity of the Upper Pannonian layers, consisting mostly of sand and sand-stone, is in between that of the Pleistocene and Levantine layers. A clayey-marl layer represents the base-formation, from the point of view of subsurface flows. The hydrogeological model was made on the basis of evaluating 609 data of wells. The data were grouped according to distance and elevation above sea level. The group of well data were characterised by the undisturbed water level and the standard deviation. The latter consists of two terms, the one which contains the measurement and model errors, and the other one, which is temperature dependent. Figure 3. shows the distribution made on the basis of the of static well water levels. The computer algorithm altered the value of horizontal conductivity coefficient Kx and of the coefficient of anisotropy e=Kx/Kz (where Kz is the vertical coefficient of hydraulic conductivity) in such a way as to minimise the difference between measured and calculated water levels. The difference was expressed as the weighed square- or weighed absolute difference. The operation of the programme is illustrated by Figure 4. The variance of the layer-parameters is calculated by the information matrix taken at the minimum of the criteria-function. Only the ratio of the horizontal conductivities can be assessed on the basis of theoretical relationships. Therefore the parameter estimation was made by the Logan-Schieder relationship, using the data of 25 well tests (Kx= 104 m/s for the Pleistocene layer). The calculations were made for three different media models of different complexity. In the single-layer model the alluvial complex of the Pleistocene, Levantine and Upper-Pannonian layers was considered a single layer. Results of the calculation are shown in Table 1, in which the profile was evaluated in parts (See Figure 3.) and also jointly. The layer-parameters of the two-layer model (Pleistocene-Levantine + Upper Pannonian) and the three-layer model (Pleistocene +Levantine + Upper Pannonian), obtained by optimisation, are shown in tables II. and III. The distribution of static water levels, obtained by the layer-parameters of Table III., are shown in Figure 5. Compared to Figure 3. the largest difference is the lack of large-depth positive anomaly at 100 km and 200 km distances. Since the model does not consider the distorting effect of inhomogeneous temperature distribution (the thermal anomaly of Tiszakécske and Hajdúszoboszló) this difference cannot be eliminated. The hydrogeological heterogeneity of the layers shows in the high variance of the estimated parameters. While the difference between measured and calculated changed only by a few per cent for the media models of different complexity, the variance of the estimated data was multiplied. Since the bulk of the flow occurs in the Pleistocene layer, the variance of the coefficient of anisotropy was, eventually, the lowest for this layer.
