Badania i metodyka prognozowania stanów wód gruntowych (w ramach współpracy WSIEGINGEO Instytutu Geologicznego)

Abstract

W oparciu o metody statystyczne podano sposób określania prawdopodobieństwa, intensywności i znaku tendencji w przebiegu wieloletnich wahań zwierciadła wód gruntowych. Przedstawiono także wyniki analizy tych tendencji, przeprowadzonej na podstawie danych z wieloletnich obserwacji zwierciadła wód gruntowych w wybranych punktach z obszaru wschodniej Polski. Przeanalizowano. Również przy użyciu metod statystycznych wybrane czynniki kształtujące reżim wód gruntowych: opady atmosferyczne, stan zwierciadła wód gruntowych w okresie poprzedzającym, cyrkulację atmosferyczną, niedosyt wilgotności powietrza i temperaturę powietrza. Uzyskane wyniki umożliwiają sformułowanie równań pozwalających prognozować wiosenne stany maksymalne wód gruntowych w wybranych rejonach. THE STUDIES AND METHODOLOGY OF FORECASTING GROUNDWATER LEVEL (WITHIN THE FRAMEWORK OF COOPERATION OF WSIEGINGEO AND GEOLOGICAL INSTITUTE)The paper presents the results of analysis of trends of many-years oscillations of groundwater table and selected factors influencing groundwater regime: precipitation, groundwater table in preceding period, atmospheric circulation, shortage of air humidity and air temperature. The analysis was made within the framework of research on forecasting groundwater regime in Poland and western parts of the Soviet Union, carried out by group of Soviet and Polish hydrogeologists.It was based on data from many-years recordings of groundwater level at selected points and meteorological records from 6 stations.The studies of trends in the course of many-years oscillations of groundwater table are aimed at estimating probability, intensity and sign of the trend. The probability of the trend is calculated using statistical methods. Obtained values of significance of the trend, P, make possible evaluations of probability of its occurrence in many-years changes of groundwater level, using normal distribution function tables. The probability is considered as reliable when it equals at least 95%, that is when its casualness is below 50%.The intensity of reliably established trends may be calculated from regression curve equations. Coefficient of regression curve equations determines the dynamics of many-years variability of groundwater table in meters per year, that is so-called trend intensity.The above discussed methodology was used in analysis of many-years trends in groundwater table oscillations for 23 points from eastern Poland. The majority of wells (about 75%) display trend to rise of groundwater table from the beginning of the fifties to the middle of seventies, whereas negative trend is shown by the remaining quarter of wells.The role of individual factors in shaping changes in groundwater table is most clearly shown by the analysis of correlation matrices of correlation coefficients of two variables or partial correlation coefficients.Of 27 factors proposed (Table 4), 21 entered the equations. Only some variants based on indices of atmospheric circulation and maximum spring water level from the preceding period appear insignificant. The most important factors responsible for the shape of maximum spring groundwater levels include groundwater table levels and precipitation in directly preceding periods. Other important factors include atmospheric circulation, air temperature and shortage of air humidity.The above given sequence of factors appears somewhat different from that obtained for western parts of the Soviet Union.The role of individual factors responsible for shaping groundwater regime markedly depends on conditions of occurrence of these waters in the nature: aquifer lithology, degree of isolation from the surface, depth of occurrence of groundwater table, etc.On the basis of these genetic regularities of the groundwater regime, the authors proposed appropriate methodology of forecasting maximum spring groundwater levels for some parts of Poland. The obtained indices of effectiveness, of this methodology are satisfactory.