THE INFLUENCE OF BAROMETRIC PRESSURE FLUCTUATIONS ON CAVE DRIP RATES

Cave drip-water studies have been widely used to investigate recharge processes and characteristics in karst systems. Commonly overlooked, or characterized as ‘noise’ in these studies, are small unexplained fluctuations in the drip rates. To investigate fluctuations in more detail, we analyze data collected between 2009 and 2011 at a drip site in Cave Without A Name in central Texas, ~35 km NW of San Antonio, Texas. Other data collected include a full suite of surface environmental parameters logged at 10-minute intervals, as well as hydrologic, geochemical, and atmospheric data measured at the same interval at the drip site.

From these data, four periods with varying drip rate and little or no rainfall were analyzed to investigate the potential effects of barometric pressure changes. Results show a strong anti-correlation between barometric pressure and drip rate, with the relationship varying depending on the drip rate. Barometric pressure has little influence on drip rate during periods of high flow (>400mL/hr) and periods of low flow (<150mL/hr), while periods of moderate flow (150 to 400mL/hr) are strongly influenced by barometric pressure. During moderate flow, fluctuations in barometric pressure can account for changes in the drip rate of as much as 65-70mL/hr, or a change in drip rate of ~25%. Our preliminary results are consistent with the findings of Genty and Deflandre (1998) and support their proposed hypothesis that, as pressure increases, air bubbles entrained in microfissures in the epikarst compress and result in a volume loss and a decrease in hydraulic head. This results in a decrease in discharge. The opposite effect is also true.

At our site, and at higher flow rates, these effects are damped by flow in larger conduits or fractures. Conversely, when the drip rates are very low, the effect diminishes and disappears; perhaps due to drainage of fractures with entrained air and/or a dominance of diffuse matrix flow over fracture flow. To expand further on the Genty and Deflandre conceptual model, we also propose that subtle features observed in the drip-rate data are the effects of gasses moving into and out of liquid solution and further enhancing the barometric-related volume changes. A new model is being developed and tested which includes both barometric effects and gas-solubility effects on bubble volume.