Most studies of the temporal variability of piezo-metric levels that appear in the literature refer to short-term cyclic variations, with periods ranging from hours to days or weeks, such as the variations deriving from tidal waves that may be observed in coastal aquifers. Some series presenting long-term fluctuations have been studied qualitatively, and then compared with rainfall data, with which there is not necessarily a parallel (Hanson & Dettinger 2005).

The most relevant aspects of the present study are the long piezometric series involved and the spectral methodological focus, as there are few bibliographic precedents. Detecting longer cycles requires a long series of records, preferably on a network of monitoring points: there may be a spatial variability in the hydrodynamic parameters of the aquifer affecting the preservation of cycles in some places, while the cycles are unobservable at other places. There were four different cycles detected in this study (not all being detected at all piezometers): decadal, 3.2-year, annual and semi-annual. The decadal cycle is perhaps the most interesting one. It is the cycle with the longest range that is related to sunspot activity, yet it is not expected in waterhead evolution. It is seen more clearly at the piezometers in close proximity to drainage channels at the border of the aquifer or along the main river entering the aquifer. This is the case because climatic variations of rainfall in relation to sunspot activity are amplified by the effect of the drainage basin, which will produce a runoff series well correlated with the climatology. Thus, it has the same effect on piezometers close to where those drainage channels enter the aquifer. The presence of the three-year cycle would have a similar general explanation, yet because the rainfall regime and its spatial distribution are different, variations in the piezometric level would be seen only in the more permeable sectors of the aquifer borders. In this way, the relationship between the surface drainage network and the aquifer recharge would be established as a reflection of the climate. The existence of cycles in the series can therefore be traced to climatic variations and the hydrodynamic characteristics of the aquifer itself, as the signal received by the piezometric levels is modulated by the functioning of the aquifer. Local or distorting phenomena that may be caused by stormy episodes are not manifest in the piezometric levels, which would clearly reflect only the predominant cycles of the regional climate.

Fig. 6. Mapping of statistical significance of the cycles, represented over the transmissivity (m2/day) kriged map: (a) decadal, (b) NAO, (c) annual and (d) semi-annual cycles. The axes are expressed in UTM coordinates.

In conclusion, we underline the importance of long-term monitoring of piezometric levels. Processing this information with the methodology put forth here constitutes, in our opinion, a sound reference for the study of climatic changes.

We are grateful for the financial support given by the Spanish MEC (Project BTE2002-00152) and Junta de AndalucĂ­a (Group RNM122). The third author is a Ramon and Cajal Grant Holder from the Ministry of Education and Science (MEC).

0 0

Post a comment