National Speleological Society Bulletin 51(2): 125-128. December 1989.

WIND CAVE, SOUTH DAKOTA:
TEMPERATURE AND HUMIDITY VARIATIONS 
James Nepstad and James Pisarowicz
Wind Cave National Park

It has long been thought that caves possess constant temperatures and humidities; never changing, even during periods of extreme temperature fluctuations on the surface. Although this may be true in very remote parts of large cave systems, most caves exhibit enough variations in this respect to warrant serious investigation (Wigley & Brown, 1976; Bramberg, 1973; Stark, 1969; Cropley, 1965; Davis, 1960; Little, 1952).
    In his investigations into temperature fluctuations at Greenbrier Caverns and Ludington's Cave, Cropley (1965) noted that cave temperatures would vary widely under influences of surface temperature fluctuations and the varying flow patterns of air and water currents within the studied caves. Such changes in cave temperatures were attributed to a combination of cold wintertime air and water entering these caves, evaporative cooling by less-than-saturated cave winds, and, to a lesser extent, by the conduction of heat from the surface through the rock above the caves.
    Wind Cave, located in the southern Black Hills of South Dakota, aptly named. Winds in excess of 120 kph have been recorded at the entrance to the cave. Conn (1966) provided evidence that such winds are caused by barometric pressure changes with the magnitude of the wind related to the cave's vast size (currently 84 km surveyed) and the small size and number of entranced to the cave (2 natural entrances each approximately 300 cm²; two man-made entrances, one 3 m² and the other sealed by an elevator door).
    Few caves experience the volume of airflow which Wind Cave exhibits. On average, almost 3000 m³ of air enter or leave the cave per hour when the Walk-In Entrance is open. With such huge exchanges of air, Wind Cave provides an excellent study site to evaluate not only the effects of cave winds upon temperature and humidity variations but also the effects of man-made intrusions such as digging entrance tunnels and shafts into the delicate cave environment.

Method

Between 11 November 1984 and 25 March 1988 temperature and humidity readings were recorded on the surface and at 14 standardized sites located along the tour route in Wind Cave (Figure 1). These readings were taken on 109 different days during this time period although all sample sites were not sampled each day (Table 1). A sling psychrometer (Princo Instruments, model D-430) was used to record the temperatures and humidities using the standard wet bulb/dry bulb method. Wind speed and direction at the Walk-In Entrance to the cave were recorded as well as whether or not the door to the cave was open or closed. Miscellaneous observations such as sky cover, precipitation, unusual wet or dry conditions in the cave at the recording sites, etc., were also noted. It took approximately 1.5-2 hours to traverse the cave to make the required measurements with the sites sampled in the same sequence on each day. As all the data collection locations were on the tour trails, travel time form location to location was not a significant factor. Thermometers were read to within 0.5° F. The resulting uncertainty is approximately 3% (Wefer, 1989). In addition to the standardized sample locations several "deep" caving trips collected temperature and humidity readings during the study.
    Ten different thermometers were used in the study because of breakage in the cave. Each thermometer was standardized against a laboratory thermometer in a constant temperature environment in the temperature ranges normally found in the cave. The thermometers chosen for use in the cave were matched pairs in that each pair gave identical readings in the temperature ranges expected to be found in Wind Cave. Thermometers were not calibrated beyond the temperature ranges found inside the cave.

Results and Discussion

Figure 2 indicates variation in temperature throughout the sampled areas as a function of air direction at the Walk-In Entrance. Several items are significant in this figure. First note the large temperature differentials in the Post Office [PO] (t = 9.619, df = 107, p < .001), Rookery [RKY] (t = 12.353, df = 107, p < .001), Flowstone [FLO] (t = 6.618, df = 107, p < .001) and Church [CHR] (t = 6.182, df = 107, p < .001). Significant differences also observed at the Upper [UE] and lower [LE] elevator landings (t = 2.222, df = 107, p < .05 and t = 2.290, df = 107, p < .05, respectively). The temperature differentials are not significant at the other sample locations because of air flow direction.
    In general the air outside the cave is colder and drier than the air underground. During the period of this investigation the outside temperature averaged 6.9° C and the relative humidity 56.7%. When rising barometric pressure forces this colder, drier air into the cave, the temperatures in the cave near the entrance drop as the cave gives up heat to warm the colder surface air. These data indicate that air movement into the cave at the Walk-In Entrance results in significant differences in cave temperature as far into the cave as the Church (approximately 300 m direct distance from the Walk-In Entrance). Although it is tempting to use distance measures in this analysis, because of the extremely complicated, three-dimensional nature of Wind Cave, what those distance would be are unknown. Nevertheless, it appears that significant temperature fluctuations result from air movement shifts as far along the tour route as the Church.
    Also noteworthy are the significant temperature differentials at both elevator landings. Despite the elevator doors and a building over the elevator shaft, air flow down the shaft significantly effects cave temperatures at these locations.
    The general shape of the data in Figure 2 is also revealing. Moving along the graph from the left, temperatures can be seen to rise until they appear to stabilize in the area of the cave called the Top of the 87 [T87] and remain fairly constant until the Second Crossroads [SXR] where the temperatures drop again. The next four readings are relatively constant with more variability in the last two sets of data, the Lower Elevator [LE] and Upper Elevator [UE] landings.
    Using correlated t-tests on overall temperature readings reveal that the data separate into five separate clusters. The first cluster involves the Post Office and Rookery temperatures. The temperatures at these locations are not statistically different from each other but are statistically different from the other cave temperatures. In a similar manner, the Flowstone and Church make another cluster. A third cluster involves the Temple, Assembly Room, Eastern Star and Garden of Eden locations. The Top of the 87, Coliseum, Fairgrounds and Second Crossroads comprise the fourth cluster. The final cluster includes the two elevator landing data sets.
    The entrance cluster (PO and RKY) is separated from the second cluster (FLO and CHR) by a series of trails which constricts before opening into large passage again in the vicinity of the second pair of locations. perhaps this constriction along the main corridor of the tour trail accounts for the separation of these four locations into two separate clusters.
    The third cluster location, (TEM, ASS, ES, GDN) have in common that they are near the elevator entrances, but separated by some distance form the direct effect of the air movement up and down the elevator shafts. In a similar vein, the two elevator data collection locations are colder than adjacent sampling points probably because of air leaking into the cave down the elevator shaft. Such leakage of surface air into the cave at these locations tend to depress these temperatures despite the elevator doors and the building which sits atop the elevator shaft.
    Finally, the eastern cluster (T87, COL, FG, SXR) consists of locations either in the upper levels of the cave (T87, COL, FG) or, as in the case of the Second Crossroads [SXR], at the farthest distance (straight line) from the Walk-In Entrance. The temperatures in this cluster appear to be the most stable and correspond most closely to temperature readings that have been taken on "deep" trips into Wind Cave (12.5-12.8° C).
    Figure 3 indicates the relationship between location, the direction of air flow and relative humidity. Interestingly, all of the locations with statistically significant differences--the Temple [TEM] (t = 2.619, df = 99, p < .05); Assembly Room [ASS] (t = 2.621, df = 107, p < .05); Second Crossroads [SXR] (t = 2.439, df = 69, p < .05); Eastern Star [ES] (t = 2.814, df = 107, p < .05) and Garden of Eden [GND] (t = 3.210, df = 107, p < .01)--are relatively far from the Walk-In Entrance, fairly close to the elevator entrances, or seem to connect at about the same cave elevation through a passage named "Summer Avenue."
    Summer Avenue is an interesting place along the tour route in Wind Cave. No matter which direction the air is moving at the Walk-In Entrance, the air movement is always from east to west through this passage. The reason behind this unusual air flow is not currently understood and has not been addressed in any literature about Wind Cave. Whatever the reason for this air-flow pattern, each of the locations with significant differences in humidity as the air flow changes are along passages on approximately the same level near the elevator entrances.
    The humidity data collected show less of a tendency to "cluster" than the temperature data. In this regard, only the elevator locations (LE and UE) may be referred to as a cluster. These locations relative humidities are not statistically different from each other, Yet the Lower Elevator/Assembly Room and Upper Elevator/Easter Star differences are statistically significant using a correlated t-test (t = 3.166, df = 107, p < .01 and t = 2.511, df = 107, p < .05, respectively). Leakage of colder, drier air into the cave down the elevator shaft is responsible for the dry conditions in the cave near the elevator shaft.
    Although vast quantities of air move into the cave through the Walk-In Entrance, statistically significant differences were not observed at the entrance sampling locations. This finding can be attributed to two factors. First, there was great variability in the humidity readings in this part of the cave. For example, in the Post Office, relative humidities as low as 60% were recorded during winter months when the cave is inhaling large quantities of air and as high as 100% when the cave was exhaling. Second, a large measurement uncertainty factored into all of the humidity data. This was because of the relatively gross precision of the temperature data. With temperatures read to within 0.5° F the uncertainty in relative humidity was approximately 3% (Wefer, 1989). Such an uncertainty in the measurement of this variable would tend to mask any real differences that might actually exist.
    Statistically significant differences were also noted between the Fairgrounds and Second Crossroads (t = 2.913, df = 69, p < .05) and the Second Crossroads and the Assembly Room (t = 3.021, df = 69, p < .01). Perhaps these differences are related in some way to the Summer Avenue phenomenon.
    A reexamination of results from Figures 2 and 3 for the elevator cluster indicates the double effect of dry air entering the cave. Not only does this surface air lower the temperature of the cave air in its own right, but the drier air also evaporates water from the cave environment. This drying action further cools the cave since evaporation required large quantities of heat. This same mechanism of cave cooling has been previous noted at Lehman Cave (Stark, 1969) and at Greenbrier Caverns (Cropley, 1965). The entrance cluster data indicate a similar trend insofar as the humidity differential at the Rookery, resulting from air flow direction, was significant (t = 1.89, df = 107, p < .05, one-tail).
    At one time (prior to 1890) the only significant opening to the surface at Wind Cave was the blowhole near the Walk-In Entrance. The construction of man-made entrances allowed for a huge increase in airflow into and out of the cave, bring with it changes in the cave climate.
    The greatest harm to the cave itself may come form the evaporation of moisture in the cave. Many of the cave's speleothems are directly dependent upon the amount of water available. Stalactite growth may be slowed or even stopped when less dripping water is available. Such was the case at Carlsbad Caverns until air locks at the elevators slowed this process.
    There is also considerable evidence showing that aragonite tends to form in preference to calcite in areas with high evaporation rates. (Hill & Forti, 1986). Thus, a change in cave climate could possibly change the very chemical structure of the speleothems in the cave.
    Cave fauna will also be disturbed by a change in the cave climate. Today Wind Cave's fauna is relatively sparse and one can wonder whether it was more abundant sometime in the recent past. Animals which have evolved in the cave's environment over thousands of years probably have little tolerance for major temperature changes. Many of these animals live on moist surfaces. When evaporation takes place, these surfaces can become remarkably cool. Different species of bats prefer different environments in the cave for roosting and they also could be disturbed by a change in the cave climate.
    Unfortunately, there is no base line to assess the changes that opening up large entrances into Wind Cave has wrought. The climate of Wind Cave is complex and probably affects the cave in ways that we have yet to understand.

References

Bamberg, S.A. Environments in Lehman Caves, Nevada. NSS Bulletin 35(2): 35-47. April 1973.

Conn, H.W. Barometric Wind in Wind and Jewel Caves, South Dakota. NSS Bulletin 28(2): 55-69. April 1966.

Cropley, J.B. Influence of Surface Conditions on Temperatures in Large Cave Systems. NSS Bulletin 27(1): 1-10. January 1965.

Davis, W.E. Meteorological Observations in Martens Cave, West Virginia. NSS Bulletin 22(2): 92-100. July 1960.

Hill, C. & Forti. A. Cave Minerals of the World. National Speleological Society. Huntsville, AL. 1986.

Little, W. Significant Air-Streams in Ogof Ffynnon Ddu. Transactions Cave Research Group 2(2): 135-148. 1952.

Stark, N. Microecosystems in Lehman Cave, Nevada. NSS Bulletin 31(3): 73-81. July 1969.

Wefer, F.L. On the Measurement of Relative Humidity in Cave Meteorology Projects. Nittany Grotto News 36(1): 6-14. Winter 1989.

Wigley, T.M.L. & Brown, M.C. Cave Meteorology. In T.D.Ford & C.H.D. Cullingford (eds). The Science of Speleology: 329-344. Academic Press, New York. 1976.