The study considered the years with hydrologic extremes takink place during the last century. The hydrologic extremes were correlated with the state of the unconfined aquifer. The hydrologic extremes were detected through a statistical analysis of the series of precipitations and the alternative water resource in the unconfined aquifer. The election of the research areas was based on the avalability and quality of freatic records, longitude of the series and the coherence and consistence. In this case, the network efficiency was also analyzed.
The regional averages of the water table depth were considered, supposing affected that they were affected only by climatic variations. Bearing in mind that the several water resources states in the region respond to various relationships between the variables precipitation, evaporation and evapotranspiration, we carried out a serial water balance and correlations between the precipitation variability and the water table depth.
The variability of the water levels with time was visualized in hydrograms. The periodic oscillations were analyzed with spectral analysis techniques and the tendencies were detected by means of conventional statistical analysis.
It in the studied zone there was an acceptable correlation between the precipitation and the behaviour of the unconfined aquifer. Here the unconfined aquifer suffers quasi-periódic variations that might be considered in projects for the use of groundwater resources.
Therefore, it can be said that precipitation produces decreasing and rising of water level. he relationship between the ENSO and the precipitation in the Southeastern South America (SSA) has already been denonstrated by several authors. The present work aims at establishing relationships between the water table and the ENSO. This correlation also provides useful information about the predictability of the future water table levels for sustainability of groundwater resources.
Then, the main conclusion of this work is that the water table level is predictable in terms of a predictor like El Niño, and we believe the coefficients of correlation between the water table levels and the predictors could probably be improved by means of the application of an multiregressive analysis.