Land subsidence in Venice and Porto Corsini

— This pape r is a con t r ibu t ion to t h e s t u d y of t he problem of subsidence in t he P o del ta and in t h e Vene t i an lagoon. W e h a v e compared t h e d a t a supplied by t i de gauges in Tries te , Venice a n d P o r t o Corsini (Fig. 1) dur ing t h e period 1896-1966, and in Genoa in t he period 1894-1966. Al though d a t a were recorded in s l ight ly d i f ferent periods, these periods a re sufficiently over lapping to allow for a significant s ta t i s t ica l compar i son . W e no t e t h a t t h e m o v e m e n t of t h e a n n u a l mean sea level in Genoa a n d Tr ies te is g rea t ly dif ferent f r o m t h a t of P o r t o Corsini and Venice and , in order to s epa ra t e t h e componen t due to eus ta t i c effect f r o m a n y possible m o v e m e n t of t h e coast , we calcu la te t h e polynomial coefficients, t h a t best fit t h e real movemen t s . Approxi m a t i n g pol inomials are of first degree for Genoa and Tr ies te and of second degree for P o r t o Corsini a n d Venice. W e have eva lua te t h e coefficients of crosscorrelat ion firstly on original d a t a and then sub t r ac t i ng f r o m original d a t a t he best fit pol inomials . T h e correlat ion is very signif icant and t h e t r end is similar in Tries te and Genoa. T h e power spec t rum analysis wi th t h e direct Four ie r ' s me thod show long periods wich are shared by t ide gauges. W e conclude es t ima t ing 1,35 m m / y e a r t h e increase of t h e mean sea level for t h e eus ta t i c effect in t he Medi te r ranean 3,9 m m / y e a r t h e subs idence in P o r t o Corsini and 1,4 m m / y e a r t h e subsidence in Venice.

1. -STATEMENT OF THE PROBLEM An analysis of long time records of tide gauges located in differents stable areas should give the secular sea level variation.Unfortunatly it is not easy to prove the stability of stations where mareographs are located.
We have collected data on four italian stations located in geologically different areas covering about 70 years and attempted to estimate the secular variation of the sea level and also the subsidence of some of the stations.To have more accurate results the analyses of a larger number of stations should be made; however we feel that the data that we analyse give consistent and accurate results.
2. -AVAILABLE DATA Data are supplied by tide gauges in Trieste, Venice, Porto Corsini and Genoa, they are yearly averages of observations carried out in the periods listed in Table 1.Iii the period 1922-1034 in Porto Corsini occasional recordings were made which allowed for an interpolation that covers the whole lapse of time 1896-1966.It was impossible to perform the same operation in Genoa, because the local tide gauge had been damaged in the course of various war events and remained inactive between 1911 and 1927.
The tide gauges of Genoa and Trieste are located on solid rocks those of Venice and Porto Corsini are on thick sediments which are known to subside.
Table 1 summarizes the periods for which we have information.Pigs.2, 3, 4, 5, concerning the recorded tide gauge data, clearly show that the movement of the mean annual sea levels in Porto Corsini and Venice greatly differs from that observed in Genoa and Trieste.A survey of the data suggest that in each station there is a secular trend on which periodic oscillations are placed on.
In order to separate the periodic movement of the sea level and the component due to glacial eustatic effect from the secular trend, we computed best fit polynomials of 1st degree for Genoa and Trieste and 2nd degree for Porto Corsini and Venice.
The correlation coefficients between the raw data and the data after subtracting the secular trend are given in Tables 2 and 3.It is  obvious that the correlation is very significant after removal of the secular effect which proves the physical significant of the data for the long periodic variations of the sea level.The power spectrum was also obtained, by Fourier's direct method.The plot of the power spectrum clearly show the following periods (in years) which are shared by all tide gauges: The statistical test of Walker on the power spectrum show that the peaks relate to period of 20.0 and 12.5 years are highly significance (97,5% confidence level); the others have a confidence level lower than 60%.
It should be noted that some periods related to the Genoa recordings are slightly different from those recorded by the other tide gauges, from this it is point out also a period at 25 years which does not appear in the other records.This difference is most probably due to the different situation in the sea basins.Also the power spectrum concerning Genoa data is based on a less complete time series.As we mentioned the tide gauges in Venice and Porto Corsini are located on thick sediments in areas wich are known to subside.The linear secular trend in Trieste and Genoa which are located in stable rock of geological different areas are: 1.40 mm/year and 1.36 mm/year in perfect agreement.This linear trend is very probably due to eustatic sea rise.
Genova e Trieste e delle curve di II grado per Porto Corsini e Venezia.I coefficienti (*) This work has been carry out by the contribution of C.N.R. (Lavoro eseguito con il contributo del C.N.R.).(**) Institute of Geodesy and Topography University of Bologna (Italy).

Table 2 -
CORRELATION COEFFICIENTS FOR THE RAW DATA

Table 3 -
COEFFICIENTS OF CORRELATION AFTER SUBTRACTION OF LINEAR