A Tisza és a talajvízjárás hosszú távú kapcsolatrendszere Szegeden

Abstract

In this research the fractal properties of the groundwater regime and the Tisza River in Szeged was analyzed using R/S (Rescaled Range) analysis. During the analysis the Hurst exponent (H) was defined in order to reveal whether the
changes of the levels of groundwater and Tisza River were persistent (H<0.5), antipersistent (H>0.5) or random (H=0.5). Regarding the water level time series of the Tisza, the R/S analysis resulted in a value of H=0.65, while the time series of
the 30 groundwater-level monitoring wells varied between 0.62–0.93. The results clearly verify the character of persistence and, thus the processes can be considered as ones amplifying the trend and having long memory elements.
Persistence in water fluctuations means that the within-period changes in the water levels (increasing or decreasing) are expected to continue in the next period as well. By analyzing the spatial distribution of the Hurst exponents for the time
series of certain shallow groundwater-level monitoring wells, it was possible determine which areas are most likely to behave similarly to the Tisza River. The behaviour of the groundwater regime in the respective cases of four wells proved
to be similar to the Tisza. In fact, three of these wells had a direct relationship with the river. The analysis of the geological conditions of the environment of different wells highlighted, the point that the groundwater regime of areas characterized by clay several metres thick has greater persistence than in those areas with a relatively thick sand layer. Consequently, the long memory of a groundwater regime is significantly influenced by the geological structure. Based on the Hurst exponents, it was found that the R/S analysis is suitable for the comparison of time series with different long time (14.5 and 1.5-year-data) and measure (3-day-data, half and 4-hour-data) intervals. This is due to the scale invariant feature of the analysis, since the results of the time series of different wells are similar.