2007 – EFFETTI DELLE MAREE ATMOSFERICHE IN UNA CAVITÀ DEL CARSO TRIESTINO: RISULTATI PRELIMINARI
pubblicato su : Atti e Memorie della Commissione Grotte “E. Boegan” Vol. 41 pp. 17-24 Trieste, 28 febbraio 2007
ATMOSPHERIC TIDE EFFECTS IN A TRIESTE KARST CAVE: PRELIMINARY RESULTS
SUMMARY
The preliminary results of the analysis of atmospheric pressure and air temperature data, recor-ded in the Trieste Karst cave “Abisso di Trebiciano” (n. 17 VG) from January 8th to July 9th 2006, are reported. The measurements were performed by means of a programmable data logger, placed at about 100 m below the cave entrance and with a sampling period of 5 minutes. The analysis of the recorded data evidenced the presence of a semidiurnal cycle whose amplitude is about 0.85 m°C and 44.4 Pa for temperature and pressure, respectively. Moreover, a phase-lag between temperature and pressure was observed, with temperature peaking about 1 h 45 min earlier than pressure. Finally, in the periodogram analysis of atmospheric pressure, a 34-h peak was observed; however more data are needed to validate this result.
RIASSUNTO
EFFETTI DELLE MAREE ATMOSFERICHE IN UNA CAVITÀ DEL CARSO TRIESTINO: RISULTATI PRELIMINARI
Vengono presentati i risultati preliminari dell’analisi delle misure di pressione atmosferica e temperatura dell’aria registrate dall’8 gennaio al 9 luglio 2006 nella grotta del Carso triestino denominata “Abisso di Trebiciano” (n. 17 VG). Le misure sono state effettuate ad una profondità di circa 100 m dalla superfcie, mediante uno strumento digitale programmabile con un intervallo di campionamento di 5 minuti. Il numero totale di misure disponibili per ciascun canale, in seguito al controllo di qualità, è 51955. La statistica descrittiva relativa alla temperatura ha evidenziato la notevole stabilità termica della cavità, caratterizzata da un campo di variazione di soli 0,09 °C, dedotto da un valore minimo di 11,126 °C e da un valore massimo di 11,216 °C, mentre il valore medio è pari a 11,169 °C. Successivamente i dati di ciascun parametro sono stati fltrati mediante una media mobile calcolata su 288 punti, corrispondenti ad un intervallo di tempo pari a 24 ore, e sono stati confrontati. Da tale analisi è emerso che, in particolar modo nei mesi più freddi (febbraio e marzo), ad una diminuzione della pressione atmosferica è corrisposto un aumento di alcuni centesimi di grado della temperatura dell’aria, dovuto allo spostamento verso la superfcie della massa d’aria più profonda, caratterizzata da una temperatura maggiore rispetto a quella degli strati superiori. Dai valori sperimentali di pressione e temperatura non fltrati sono state successivamente calcolate per ciascun parametro le medie orarie, la cui analisi ha evidenziato la presenza di un ciclo semidiurno di ampiezza pari a circa 0,85 m°C per la temperatura e a 44,4 Pa per la pressione. Il ciclo medio giornaliero delle pressione e della temperatura, dedotto dalle medie orarie, è stato quindi interpolato con una fun-(*) Federazione Italiana Maricoltori ONLUS/ONG (NGO), Via Guerrazzi, 2, I – 34131 Trieste – e-mail: andrea.bussani@tin.it
zione trigonometrica costituita dalla somma delle armoniche diurna (periodo: 24 ore) e semidiurna (periodo: 12 ore): questa procedura ha permesso di rilevare una differenza di fase tra i due segnali semidiurni, con un anticipo della temperatura pari a 1 h 45 min. Infne, attraverso l’analisi dei perio-dogrammi, calcolati per entrambi i parametri, è stata osservata una componente armonica di periodo pari a circa 34 h nel segnale di pressione, la cui validazione necessita di ulteriori misure.
POVZETEK
VLIP ATMOSFERSKE BIBAVICE V JAMI NA TRŽAŠKEM KRASU
Predstavljajo se zacˇetni rizultati analiz meritev atmosferskega pritiska in temperature zraka, ki so bile opravljene od 8. januarja do 9. julija 2006 v jami “Labodnici (Abisso di Trebiciano)” št. 17 VG na Tržaškem Krasu. Meritve so bile izvršene približno 100 m globoko od površja s pomocˇjo pro-gramirane digitalne naprave s cˇasovnim presledkom petih minut. Analiza podatkov je pokazala pri-sotnost poldnevnega ciklusa z amplitudo 0.85 m°C za temperaturo in 44.4 Pa za pritisk. Poleg tega je bilo razbrati razliko v fazi med dvema poldnevnima signaloma s predcˇasnostjo temperature za 1 h in 45 min. Z analizo periodograma atmosferskega pritiska je bilo opaziti v signalu približno 34-urno harmonicˇno komponento. Vsekakor so potrebni dodatni podatki za ugotovitev veljavnosti pre-dstavljenih rizultatov.
Introduction
In recent years a considerable number of research campaigns, aiming at the acquisition of atmospheric pressure and air temperature data in Trieste Karst caves, have been carried out by means of programmable data loggers placed in selected hollows. The use of such instruments has permitted the collection of time series of temperature and pressure at short sampling period (5-20 minutes) and, for the best instruments, at high resolution (pressure: about 15 Pa; temperature: about 0.003 °C). The frst analyses of the pressure data recor-ds collected during the research campaigns evidenced clearly the presence of atmospheric tides in the pressure signal, a well known phenomenon thoroughly described, for example, by Chapman and Lindzen (1970). Surprisingly, evidence of atmospheric tide was detected also in the air temperature measurements recorded in a Trieste Karst cave, where a semi-diurnal signal with an amplitude of about 0.0022 °C was observed (Bussani, 2004). More unexpectedly, the comparison between such signal and the corresponding semidiurnal con-stituent of atmospheric pressure extracted from the 1961-1990 time series recorded at Trieste University (stravisi, 1994), revealed the presence of a phase-lag, with the cave temperature reaching its maximum about 2.61 hours earlier than the external atmospheric pressure. These results are essentially comparable to those obtained by sondag et al. (2003) who analysed air temperature and atmospheric pressure data collected in Karstic caves of two areas of Brazil, with temperature peaking about one hour earlier than pressure: the semi-diurnal variation of temperature was interpreted as a consequence of the adiabatic expan-sions and compressions of the cave atmosphere, but no hints were given about the phase-lag between temperature and pressure. Very similar results were also obtained by Chen et al. (2003) from the analysis of air temperature and atmospheric pressure data recorded in an underground laboratory, but the presence of the phase-lag led to the conclusion that adiabatic expansions and compressions could not account for the whole phenomenon. The differen-ce between the phases of the two signals was then explained by Wu et al. (2003) by means of a thermodynamic model that included the effect of heat conduction between air and rock.
It has to be kept in mind that the observed semidiurnal signals are caused by the so-called thermal atmospheric tides, i.e. tides induced by solar heating and resonance effects
that take place in the atmosphere, while the gravitational atmospheric tides, originating from the lunisolar attraction on the atmospheric air mass, are comparatively negligible.
This paper deals with the effects of the thermal atmospheric tides (atmospheric tides, in the following) upon atmospheric pressure and air temperature data collected in the Trieste Karst cave “Abisso di Trebiciano” (n. 17 VG) during the frst half of 2006. Descrip-tive statistics, mean daily cycles and harmonic analysis of both parameters are performed and compared to what is reported in the literature.
Data collection
A Driesen und Kern Plog 520 thermobarometer, originally designed for the marine en-vironment, was placed in the Trieste Karst cave “Abisso di Trebiciano” n. 17 VG (fg. 1) at a depth of about 100 m below the surface. The cave entrance, sealed by a locked trap door, is at 341 m above sea level (a.s.l) and has a total depth of about 330 m. The instru-ment, whose technical characteristics are given in tab. 1, was placed at about 1 m above the cave foor and at about 2 m from the nearest rock walls.
Data recording started on January 8th 2006 with a sampling period of 5 minutes. Data retrieval occurred on July 9th 2006, when 52393 measurements for each channel were downloaded. Collected data underwent a quality check procedure consisting essentially in the removal of the recordings taken in the frst two days, markedly affected by the pre-sence of the experimenters who entered the cave the frst day to set up the instrument, and those recorded during the retrieval. Moreover a visual check of all data was performed in order to remove possible spikes, though none was detected. The fnal number of recordin-gs for each channel is 51955.
Statistical analysis
The time-domain analysis consisted mainly in the defnition of the mean daily cycles of each parameter: hourly means of pressure and temperature were computed, averaging all the measurements taken in the same 1-h time intervals, centred about integer hours (e.g. all measurements collected between 00:30 and 01:30 were averaged to give the mean value of 01:00). The mean daily cycles were then interpolated by means of the following trigonometric function, consisting in the sum of the diurnal and semidiurnal harmonics:
A frequency-domain analysis was then carried out: data were smoothed using a run-ning mean flter computed upon 288 points, corresponding to a time window of 24 h, centred on the 145th value. Residuals between experimental data and the fltered signals were then worked out for each parameter. Finally, periodograms were calculated both for pressure and temperature residuals in order to detect the presence of periodic constituents.
Results and discussion
The running-mean fltered time series of atmospheric pressure and air temperature sig-nals, recorded in the cave by the thermobarometer, are reported in fg. 2, while descrip-tive statistics of unfltered quality-checked data (i.e. after the removal of measurements infuenced by the logger setup or retrieval), is given in tab. 2. Broad variations are evident in the pressure signal, presumably induced by the external atmospheric pressure; on the contrary, the temperature fuctuations are markedly damped with respect to the outside: in fact, one of the most striking features of the recorded data is the extreme stability of the cave temperature, whose range during the months under consideration is only 0.09 °C. Moreover, a careful cross-analysis of both the temperature and the pressure signals evi-dences that, especially in the coldest months, temperature frequently peaks in correspond-ence to drops in pressure: this phenomenon could be attributed to the upward displacement of deeper and, in winter months, warmer air, driven by an atmospheric pressure decrease. The mean daily cycles of temperature and pressure and their respective semidiurnal constituents are reported in fg. 3, while the values of the best ft parameters are given in tab. 3. Both cycles show a clear semidiurnal pattern, however the temperature cycle is slightly less smooth than the pressure cycle: this is due to the fact that the amplitude of the temperature signal, unlike pressure, is one order of magnitude lower than the instrument resolution. As a matter of fact, differences between subsequent hourly means of the temperature cycle can be even smaller. Nevertheless, the phase-lag between the two signals is fairly evident and the difference is even clearer for the semidiurnal constituents: apply-ing eq. 1 to the mean daily cycles, it emerges that the frst peaks of the semidiurnal con-stituents of pressure and temperature occur at 10:34 and 8:49, respectively. These results agree with what was reported by Wu et al. (2003), who found a phase-lag of 0.95 rad, cor-responding to a difference of 1 h 48 min. However, the difference observed by Bussani (2004) in a different cave of Trieste Karst was considerably higher (2 h 36 min).
As a point of interest, according to stravisi (1994) the frst peak of the semidiur-nal constituent of Trieste atmospheric pressure occurs at 10:29, which is consistent with the value obtained in the present work (10:34), allowing for the sampling period (5 min-utes), while the phase of the diurnal constituent according to stravisi (1994) is 5.89 hours (5:53), i.e. 2.41 hours greater than the phase observed in the considered cave. However, a good agreement is also present between the best ft values of the diurnal and semidiurnal amplitudes of the atmospheric pressure computed according to eq. 1 (tab. 3) and the cor-responding values obtained by stravisi (1994), who found diurnal and semidiurnal ampli-tudes of 15.5 Pa and 44.9 Pa respectively.
The periodograms of atmospheric pressure and air temperature residuals (fg. 4) confrm the presence of a rather strong semidiurnal constituent and a smaller diurnal constit-uent in both signals; a terdiurnal constituent is also evident in the pressure signal. Some harmonics of smaller amplitude, present alternatively either in the pressure or in the temperature signal, are of dubious relevance and need more data in order to be validated. However, the amplitude of the 34-h peak observed in the pressure periodogram is clear-ly greater than that of the diurnal constituent, thus giving stronger grounds to the hypoth-esis that such peak corresponds to an actual signal and is not merely a spurious result of the data elaboration.
Conclusions
Though the aim of this research is to achieve and study a whole year of atmosphe-ric pressure and air temperature data in a deep Karstic hollow at fne resolution and with a short sampling period, the data acquired in the frst half of 2006 provide suffcient ev-idence to confrm the results recently attained by other authors either in Karstic caves (sondag et al., 2003) or in underground laboratories (Chen et al., 2003; Wu et al., 2003), such as the infuence of thermal atmospheric tides upon temperature signal and the pres-ence of a phase-lag between temperature and pressure. Moreover, the large amount of precise measurements collected in the “Abisso di Trebiciano” cave permitted to obtain very accurate mean daily cycles of pressure and temperature; for the latter parameter, in partic-ular, the differences between subsequent hourly mean values were often far below the detection threshold of the instrument, and only the availability of a considerable number of measurements, combined with the thermal stability of the cave, allowed to resolve the ex-tremely small variations that occurred in the mean daily cycle of temperature.
However the results presented in this work are still to be considered as preliminary, es-pecially as regards the spectral analysis of the atmospheric pressure, where the presence of the 34-h harmonic constituent needs more data in order to be validated: the periodog-rams of the 2nd half of 2006 will be likely either to confrm the detection of this constitu-ent or to provide stronger grounds for its rejection.
In summary, the use of programmable data loggers for the collection of accurate and highly resolved measurements of atmospheric pressure and air temperature in deep caves has demon-strated to be a very effective tool for the detection and characterization of the infuence of ther-mal atmospheric tides on the atmosphere of caves. Moreover, the exceptional thermal stability of the cave chosen for this study can possibly provide some evidence of atmospheric oscilla-tions not yet observed in other studies, but more data are needed to support this hypothesis.
Acknowledgments
The author wishes to thank the Società Adriatica di Speleologia of Trieste for the permis-sion to carry out this research in the “Abisso di Trebiciano”. The author is also particular-ly grateful to Mr. Sergio Dambrosi for his invaluable help during all the stages of this study. This work was supported by the Federazione Italiana Maricoltori ONLUS/ONG (NGO).
ANDREA BUSSANI
BIBLIOGRAPHY
Bussani a., 2004 – Infuenza delle maree atmosferiche sulle misure di temperatura registrate nella grotta ‘C. Doria’ (N. 3875 V.G.). Atti e Memorie della Comm.Grotte “E. Boegan”, 40: 125-131.
Chapman s., Lindzen r., 1970 – Atmospheric tides – thermal and gravitational. D. Reidel, Dordrecht, 200 pp.
Chen F., Wu s., Fan s, Luo J., 2003 – Determining the phases of the semidiurnal temperature and pressure os-cillations by the cross-correlation equilibrium method. Measuring Science Technology, 14: 619-624.
sondag F., van ruymBeke m., souBiès F., santos r., somerhausen a., seideL a., Boggiani p., 2003 – Monitoring present day climatic conditions in tropical caves using an Environmental Data Acquisition System (EDAS). Journal of Hydrology, 273: 103-118.
stravisi i F., 1994 – Trieste 1961-1990 – Pressione atmosferica: dati mensili e statistiche. Università di Trieste, Istituto di Geodesia e Geofsica, Sezione di Oceanografa e Meteorologia. Rapporto interno n. 94/1.
Wu s., Chen F., Fan s., Luo J., 2003 – Phase leading of temperature variations in a cavity caused by heat conduction between air and rock. Chinese Physics Letters, 12: 2192-2194.