Field Trip Guidebook Understanding Upland Recharge for Geologic Assessments
Leaders: Nico Hauwert, Ph.D., PG and Scott Hiers, PG City of Austin Watershed Protection Department Heather Beatty, PG Texas Commission on Environmental Quality Tabor Water-Quality Protection Land February, 18, 2010 Guest speakers: Brian Cowan, UT Geological Sciences research assistant Raymond M. Slade, Jr., hydrologist Mustafa Saribudak, Ph.D., P.G., Environmental Geophysics Assoc., Principal Geophysicist-Geologist Jack Holt, Ph.D., Univ. of Texas Institute of Geophysics Geophysicist Jean Krejca, Ph.D. Zara Environmental Principal cave biologist (note: cave specialist Peter Sprouse and cave biologist Krista McDermid substituted for Dr. Krejca during the trip)
Acknowledgements: Kevin Feliksa, contract technical writer for the City of Austin, assisted in editing and formatting this guidebook. David Johns, MS, PG and Ed Peacock PE of the City of Austin Watershed Protection Department provided technical review for the guidebook.
ii
Table of Contents Orientation .......................................................................................................................................... 1 Safety Discussion .............................................................................................................................. 1 Nearest Hospitals .............................................................................................................................. 1 Acknowledgement ............................................................................................................................. 1 Introduction ......................................................................................................................................... 3 Tabor Tract ........................................................................................................................................ 3 Surface Karst Process and Karst Recharge Features ...................................................................... 3 Karst Survey and Feature Evaluation................................................................................................ 6 Determining a Protective Setback for Karst Features ....................................................................... 9 Stop 1 – Quad Border Sink............................................................................................................... 11 Solution Sinkholes ........................................................................................................................... 14 Collapse Sinkholes .......................................................................................................................... 15 Soil-piping sinkholes........................................................................................................................ 15 Stop 2 – Sandbur Cave .................................................................................................................... 16 Small Sinkhole/Karst Depressions .................................................................................................. 17 Stop 3 – Barker Ranch #1 ................................................................................................................ 23 Stop 4 – Flint Ridge Cave and Sinkhole ........................................................................................... 27 Identifying Caves ............................................................................................................................. 29 Cave Maps....................................................................................................................................... 29 Cave Radio Location ....................................................................................................................... 30 Geophysical Surveys....................................................................................................................... 31 Photo Tour of the Upper Flint Ridge Cave ...................................................................................... 32 Submitted Papers ............................................................................................................................. 35 Does Rainfall Infiltrate Through Upland Soils To Recharge The Edwards Aquifer? ........................ 36 Ineffectiveness of Structural BMP’s on the Recharge Zone............................................................ 40 Resistivity and Natural Potential Anomalies over Flint Ridge Cave ................................................ 44 Urban Geophysics: A Mapping Of Mount Bonnell Fault And Its Karstic Features In Austin, Tx..... 48 Groundpenetrating radar and other geophysical investigations at Flint Ridge Cave, Austin, Texas .................................................................................................................................................... 64 Appendix A. A Recharge Classification for Karst Features .............................................................. 73 Appendix B. Some Assessments Regarding the Tabor Tract .......................................................... 85 Appendix C. TCEQ Flowchart – Assessing the Probability for Rapid Infiltration.............................. 97 Appendix D. City of Austin Setback Criteria – (COA Environmental Criteria Manual) ..................... 98 Appendix E. TCEQ Guidance RG-348: TCEQ Setback Criteria .................................................... 105 Appendix F. Zara Environmental: Key References for Endangered Karst Invertebrates............... 117 Appendix G. Maps of Karst Faunal Regions in Central Texas ....................................................... 119 Appendix H. Some Karst Species................................................................................................... 122 Appendix I. Student Geophysical Survey Reports References ...................................................... 126 References......................................................................................................................................252
iii
Orientation Safety Discussion
Biting – Rattlesnakes Tripping – Loose rocks, holes, barbed wire Poking – Cacti, barbed wire Insects – Chiggers, ticks, spiders, mosquitoes Plants – Green briars (thorns) and poison ivy Dehydration – not enough water, other fluids
Nearest Hospitals 1) Seton Southwest is at 7900 FM 1826 (7.3 miles) Go east on Green Emerald Terrace. Turn left at Brodie Lane. Turn left at West Slaughter Lane. Drive through Circle C and turn right at RR 1826. The hospital is on the left. 2) St. David’s South Austin Hospital at 901 West Ben White Blvd. (9.2 Miles) Go east on Green Emerald Terrace. Turn left at Brodie Lane. Turn right at US 290 and enter the highway. Take South 1st Street exit; hospital is on the right before you get to South 1st Street. If you miss the turn in, turn right on South 1st Street and follow signs to the hospital. Acknowledgement We gratefully acknowledge and thank the Wildlands Conservation Division of the Austin Water Utility for granting access to the Tabor Tract for the research discussed in this guidebook and for this field trip.
1
2
Introduction This field trip provides an opportunity to directly observe a wide variety of features that are typically encountered in environmental assessments and to share updates on recent studies that apply to site assessments of the Edwards Aquifer. We will use examples from actual site assessments to examine common misunderstandings and range in perceptions of site recharge among professional geoscientists. We will also show how different technologies are used to assist in site evaluation where karst is present, revisit recharge to the Edwards Aquifer based on recent studies, and examine the surfacesubsurface interaction of karst ecosystems. This guidebook is provided to supplement information presented on the fieldtrip. Tabor Tract The Tabor Tract is one of 64 Water Quality Protection Lands (WQPL) purchased by City of Austin (COA) with bond money after three different bond elections including the passage of Proposition 2 in 1998. From 1998 through 2008, the COA purchased 9,127 acres of land directly and acquired another 14,527 acres in conservation easements to fulfill a mandate to help protect Austin's water quality and quantity. The Tabor Tract spans 301 acres of undeveloped land located within the Bear Creek Watershed, about 12 miles southwest of downtown Austin. Two other WQPL tracts; Reavley and Edwards Crossing (218 total acres), are adjacent to and east of the Tabor Tract. The Tabor Tract is completely within the Barton Springs Segment of the Edwards Aquifer Recharge Zone, and is bisected by Bear Creek. Surface geology consists of the Cretaceous-age limestone of Kainer Formation of the Edwards Group, which includes the Dolomitic, Kirschberg, and Grainstone Members (as well as the Basal Nodular Member that does not outcrop on this tract but is present in the subsurface). Numerous springs and karst recharge features such as small soil-filled sinkholes (or karst depressions), small soil-piping sinkholes, solution cavities, solution-enlarged fractures (fissures), sinkholes and caves are visible on the surface of this tract. During our fieldtrip we will examine several features that are examples of those listed in the geologic assessment table that follows from the TCEQ assessment guidance document (TCEQ, 2004). Surface Karst Process and Karst Recharge Features The creation and evolution of karst landscape is regulated by the dissolution of carbonate rocks due to physical and chemical weathering (erosion and dissolution). This process is seen in the formation of karren, which is the often minor sculpturing of the bedrock surface by rainfall, sheet flow, channelized flow, and percolating flow (White, 1988). Karren formed in bedrock is influenced by the amount of rainfall, the lithology of bedrock, fracture and jointing within rock, and soil cover and vegetation surrounding the rock. Surface dissolution of calcium carbonate is dependent on water-bearing carbon dioxide produced primarily by decaying plants and animals. The production of carbon dioxide is the greatest in areas where organic matter is abundant. Carbon dioxide gas tends to be high in soils. As a result, surface karren and karst features are variable and their pathways into subsurface bedrock can be complex. These surface micro-features (karren), together with pre-existing subsurface dissolution features can create a network “surface capillaries” for rainwater infiltration. 3
4
Table 1 PROJECT NAME: Tabor Karst Workshop FEATURE CHARACTERISTICS EVALUATION
GEOLOGIC ASSESSMENT TABLE LOCATION 1A
LATITUDE
1C*
2A
LONGITUDE
FEATURE TYPE
2B
3
POINTS
FORMATION
4
5
DIMENSIONS (FEET)
TREND (DEGREES)
X
S1-Quadboarder Sink S2 -Sandbur Sink S3 -Flint Ridge S4 -Barker Bat Cave
10029875.00 10031277.00 10029658.00 10031003.00
SH 3073083.00 C 3072680.00 C/SH 3070394.00 C 3073456.00
20 30 30 30
Y
5A
Z
Kainer 350 500 4 Kainer 2 5 5.4 Kainer 270 300 10 Kainer 8 10 9
DOM
FEATURE ID
1B *
6
7
DENSITY (NO/FT)
APERTURE (FEET)
8A
8B
INFILL
RELATIVE INFILTRATION RATE
TOTAL
10
0 0 0 0
0 0 0 0
11
12
SENSITIVITY
CATCHMENT AREA (ACRES)
TOPOGRAPHY
40
X X X X
1.6
X X X
* DATUM:___________________North American Datum 1983 State Plan Texas Central 4203 - Lambert Conformal Conic 2A TYPE C
TYPE
2B POINTS
Cave
30
8A INFILLING N
None, exposed bedrock
SC
Solution cavity
20
C
Coarse - cobbles, breakdown, sand, gravel
SF
Solution-enlarged fracture(s)
20
O
Loose or soft mud or soil, organics, leaves, sticks, dark colors
F
Fault
20
F
Fines, compacted clay-rich sediment, soil profile, gray or red colors
O
Other natural bedrock features
5
V
Vegetation. Give details in narrative description
MB
Manmade feature in bedrock
30
FS
Flowstone, cements, cave deposits
SW
Swallow hole
30
X
Other materials
SH
Sinkhole
20
CD
Non-karst closed depression
Z
Zone, clustered or aligned features
5 30
12 TOPOGRAPHY
Cliff, Hilltop, Hillside, Drainage, Floodplain, Streambed
I have read, I understood, and I have followed the Texas Commission on Environmental Quality's Instructions to Geologists. The information presented here complies with that document and is a true representation of the conditions observed in the field. My signature certifies that I am qualified as a geologist as defined by 30 TAC Chapter 213.
___________________________________________
Date Sheet __1__ of _1___
TCEQ-0585-Table (Rev. 10-01-04)
5
Hillside Hillside Hillside Hillside
The epikarst (or subcutaneous zone, White 1999) is the “uppermost zone of exposed karstified rocks, in which permeability due to fissuring and diffuse karstification is substantially greater and more uniformly distributed in area, as compared to the bulk rock mass below” (Klimchouk, 2000). The Texas Commission on Environmental Quality (TCEQ) defines epikarst as the upper part of the bedrock at the surface or beneath the soil that is characterized by increased fracturing and enhanced limestone dissolution. The epikarst is a zone of significant water storage and transport. The intricate network of dissolution features in the epikarst zone function to channel surface water and allow it to infiltrate down-gradient to larger vadose flow paths, such cave streams, which function as conveyances for water to reach the water table. Epikarst is extremely important in routing water, nutrients, organic matter, and soil from the land surface and from the rooting zone into the subsurface where they can move laterally to seeps and springs, or to vertical collector structures that channel them downward into cave networks (Aley, 1997). Epikarst is recognized as an important storage zone of groundwater (Klimchouck, 2004). In some areas (e.g. Perrin et al., 2003), storage in epikarst can be more significant than storage in the phreatic zone. Because the water storage and drainage contributions of the epikarst and soil have not been distinguished, we will generally combine these zones for the purposes of this guidebook. Karst Survey and Feature Evaluation Karst surveys evaluate the significance of features for recharge and hydrologic connection to the underlying aquifer (TCEQ, 2004; Appendix A). Some surveys evaluate the biodiversity of caves on a site (USFW, 2006; Appendix F and G). Karst recharge features consist of sinkholes (dolines), solution-enlarge fractures, solution cavities, and caves that convey surface materials into the subsurface. For definitions of these features see Instructions to Geologists for Geologic Assessments (TCEQ, 2004). The identification of these features and preservation of significant recharge features is the primary goal of completing the environmental assessment for the COA land development process (COA Land Development Code 25-8-121). Caves have naturally strong hydraulic connection to the aquifer, frequently represent sites of major recharge, and are important habitat for karst species. For these reasons, they rank relatively high in karst surveys. See page 29 for further information on identifying caves. Karst features are widespread across the Edwards Aquifer outcrop (Lindley, 2004) although caves and sinkholes are not necessarily obvious to people only familiar to them through large gaping commercial caves (Hauwert, 2009, Appendix A). Ranchers commonly disposed of trash in sinkholes and cave entrances or fill them to protect livestock from injury (Hauwert, 2009). Large internal drainage sinkholes are frequently modified to serve as ranch stock ponds. Historically, some creek swallets were intentionally filled to allow greater surface flow for purposes including mill production. Recent anthropogenic erosion has increased an order of magnitude over naturally high post glacial erosion rates (Pimentel, et al., 1995; Cooke, et al., 2003). Consequently many low-lying sinkholes have become filled. Also, with natural erosion and infilling of surface depressions with sediment, many karst recharge features have little expression visible at the surface. Dense vegetation tends to localize around recharge features and can obscure the surface expression of a feature. Consequently, many if not most caves known today required some degree of work to access them. Some caves known today 6
(such as Djeridoo Cave, located on private property just south of the Tabor tract) appeared simply as a pile of rocks. As drainage features, each sinkhole contains a drain that is potentially well-connected hydraulically to the aquifer, although the drain may be obscured by soil fill. Rapid infiltration can occur in soil-filled sinkholes, either through macropores or eventual failure of ephemeral sediment plugs (White, 1988). Red, clayrich terra rossa clays, and siliceous rock fragments are assumed to be residual from slow denudation (erosion and dissolution) of the overlying landscape (Quinlan, 1978; Cooke et al, 2006; Hauwert, 2009). The presence of terra rossa clays may indicate very slow rate of soil movement through a feature. A recommended practice to systematically locate all recharge features on a site is to walk the survey area in closely spaced transects of 50 ft apart or smaller, marking features with flagging and locating them using a Global Positioning System unit (GPS). To classify and evaluate a potential feature, a geologist should hand-excavate the feature if filled with loose rocks, soil, and debris (“when in doubt, dig it out”). The duration and depth of excavation depends on the likelihood that the feature is naturally significant for recharge, has subsurface connection, and serves as important habitat for karst species. Besides the rapid infiltration rate discussed above, both COA and TCEQ methodologies use catchment areas to aid in evaluating the amount of potential recharge to a feature and its sensitivity. Through erosion and dissolution, large flows to features tend to create strong hydraulic connection with the subsurface. Catchment areas for large internal drainage sinkholes can be estimated in several ways including:
Mapping defined drainages entering sink and catchment divides in the field using a sufficiently accurate GPS unit and visually estimating the limit of surface flow paths to a feature. The presence of channels and flow debris may assist in the catchment delineation. Locating drainage divides on available surface-elevation contour maps. Although surface contour maps are generally insufficient to map entire catchment areas for internal drainage basin sinkholes without field mapping, they may complement catchment delineation in areas where subtle catchment divides ascend slopes. Surface contour maps may be a suitable tool to delineate catchments for creek swallets and karst features without well-defined drainage to which runoff is diffuse. Surveying the surface topography around the sink using professional survey techniques.
The mapped catchment areas can be checked by:
7
Estimation of sinkhole catchment area from sinkhole bowl volume. Upland sinkhole bowl volume formed by recent surface dissolution directly relates to the size of the catchment area within the Barton Springs Segment (Hauwert, 2009). From regression of measured sinkholes in the following figure, the bowl volume multiplied by 300 approximately equals the catchment area that can be expected for upland solution sinkholes. Cave volume exposed in collapsed sinkholes largely is not formed by recent surface dissolution but by subsurface dissolution.
Therefore, this relationship should not be extended to estimate sinkhole catchments for collapsed sinkholes. Relation of Upland Solution Sinkhole Bowl Volume to Catchment Area 250,000
y = 282x R² = 0.90
Catchment Area (m2)
200,000
150,000
100,000
50,000
0 0
100
200
300
400
500
600
700
3
Bowl Volume (m )
Measurement or estimate of the rainfall and flow to the sink during a wet period when soils are relatively saturated. If the flow entering the sink exceeds the rainfall volume over the estimated catchment area, then the estimated catchment is too small. Well-defined drainages, multiple drainages entering a sink, and signs of significant flow to a sink indicate a relatively large catchment area.
In the TCEQ’s Instructions to Geologists, a significant recharge feature is defined as a karst feature with a well-defined surface opening or a sinkhole without a surface opening that has a catchment area greater than 1.6 acres (equivalent to the area of a circle with a radius of 150 ft). Therefore, accurate delineation of the catchment area is required to determine the significance of features. According to the TCEQ, the catchment area for a feature receiving only diffuse runoff can be determined by simply multiplying the width by the length of any upslope area that could plausibly drain to the feature (TCEQ, 2004). Delineation of the catchment area for a feature located on a hillside or a flat upland area can be difficult. When determining the catchment area for a recharge feature, geologists often overlook the importance of the soil and epikarst zones effects of both storing water and routing surface and subsurface water toward a recharge feature. Most guidance used today to determine setback area for recharge features focuses on only the surface flows to a feature. A water balance of a sinkhole microbasin on nearby J17 WQPL tract measured
8
800
10% of the longer term recharge drained into the sinkhole drain, while the remaining 90% of recharge infiltrated in upland soil and minor recharge features (Hauwert, 2009). For short intervals around rain events half or more of the infiltration occurs in the sinkhole drain. Lindley (2006) generally measured infiltration rates of one inch over 6 to 16 hours through background soil and soil-filled small sinkholes (see page 18). Speleological investigations completed by A.B. Klimchouk in arid mountains of Central Asia demonstrated a substantial shaft flow in the vadose zone after long periods without precipitation, and revealed the existence of numerous "hidden" shafts beneath karren fields (Klimchouk et al., 1979, 1981; Klimchouk, 1987, 1989). Cave shaft entrances exposed at the surface may be truncated exposures of a stair-step profile cave system that at one time conveyed unsaturated vadose flows to the water table (White, 1988). Within the Barton Springs Segment, Midnight Cave in the Circle C Metro Park on Slaughter Creek demonstrates prolonged drainage of vadose flows down a shaft to the water table. Determining a Protective Setback for Karst Features Two methodologies for determining protective setback are provided in the back of this guidebook: Section 1.10.4 from the COA Environmental Criteria Manual and TCEQ RG348. The TCEQ does not require detailed maps of caves. If there is no known map of a cave footprint, the agency requires an assumed footprint of at least 150 feet around the entrance to a cave opening because 90% of mapped Edwards cave footprints have been found to lie within a 150 ft. circle centered on the opening (TCEQ, 2004). Using the TCEQ methodology, the natural buffer around the assumed footprint should extend a minimum of 50 feet in all directions around the assumed footprint. Where the boundary of the surface drainage area to the assumed footprint lies more than 50 feet from the footprint, the TCEQ buffer should extend to the boundary of the drainage area or 200 feet, whichever is less. Two methods for determining protective setback are provided in the back of this guidebook: Section 1.10.4 from the City of Austin's Environmental Criteria Manual and Texas Commission of Environmental Quality's GR-348. Both use natural setbacks area to preserve the recharge feature and their catchment area or a portion of the catchment area. The City of Austin uses a standard setback distances is 150-ft, but can be extended to maximum of 300-ft measured from a recharge features' orifice, cave footprint or sinkhole rim (or immediate catchment area). As a result, the City typically requires that an applicant provide a cave map so that cave footprint can determine on surface so that the setback distance can be measured.
9
10
Stop 1 – Quad Border Sink Quad Border Sink is a solution sinkhole and a ponded internal-drainage basin with a wetland habitat at the lowest point in the northern end of the sink. This is one of the three sinkhole types we will see on this field trip, the others being collapse sinkholes and soilpiping sinkholes. Quad Border Sink is about 350-ft wide and 500-ft long with a depth of 4 ft. This sinkhole probably started to form when a fracture in the limestone bedrock was enlarged by water dissolving the limestone into a fissure. As a sinkhole matures through dissolution of the surface, it grows in size and it becomes rounded in aerial view and concave in cross section as its conduit drain and catchment areas become larger. When the conduit drain (or aperture) is sufficiently large and the catchment generates sufficient runoff to carry bed load sediment, the sinkhole basin is further shaped by erosion. Sediment covers the bottom of the sinkhole and fills the drain, causing runoff to infiltrate more slowly than an open internal-drainage sinkhole, such as Flint Ridge Sink (the last stop on the field trip). Since many of these ponded internal-drainage basins are found on former ranchlands, erosion associated with livestock may contribute to the sediment accumulation. The catchment area for Quad Border was estimated to be 58 acres by Ford (2000) and 42 acres by Hauwert (current City of Austin coverage). Many geologic assessments misidentified large sinks as non-karst closed depressions. The description of Quad Border Sink in The Hydrogeologic Study for Proposition 2 Watershed Preserves in the Bear Creek Area is an example: “Large Closed Depression, This feature consists of a relatively large, closed depression, approximately 600 feet in diameter, located in the central area of the Edwards Crossing Tract. The depression is about 3 to 4 feet deep with an exposed outcrop of Kainer Formation limestone located along the western edge of the depression. Several small limestone boulders and debris piles are also located within the depression. As seen on Plate 1 a fault is mapped as crossing almost through the center of the depression. A possible origin of the depression may be from dissolution of the limestone bedrock beneath the depression along the fault zone resulting in sagging of the overburden. The feature appears to have fairly deep soil cover and dense grass cover present. Evidence of an organic debris line is present around the rim of the feature, which indicates the possible high water mark of ponded stormwater runoff within the depression. Discussions with residents whose properties are located along eastern edge of the depression indicate that after periods of sustained and significant rainfall, the entire depression will collect with stormwater runoff and will stay wet for several days before slowly seeping in the soil. The recharge potential to the aquifer from this feature is probably medium to high due to the relative large drainage area associated with the depression and the presence of a fault zone located beneath the depression” (Ford, 2000). In the assessment, this feature was mischaracterized as a closed depression which the TCEQ (2004) generally refers to as a “non-karst closed depression”, such as a stockpond or burrow. Sinkholes used by ranchers as stock ponds are frequently (and erroneously) referred to as closed depressions in geologic assessments. In the Barton Springs Segment internal drainage basins are very common, covering at least 10% of the Recharge Zone. Sinkhole bowls are frequently utilized as stock ponds if the aperture drains are sufficiently plugged by clay, debris, or fine-grained sediment from nearby erosion (such 11
as cattle ranching). A natural internal drainage basin can be distinguished from a manmade quarry, borrow pit, or stock pond with several criteria: 1) The morphology of the depression can indicate if it is a man-made excavation or natural sinkhole. Surface topography gradually slopes toward large internal drainage sinkholes. Collapsed sinkholes can have vertical or under-hanging walls. Quarries have steep, freshly cut walls. Man-made stock ponds are constructed in tributary drainages and have berms to retain water. 2) Large sinkhole basins typically contain siliceous remnant rock fragments. These rocks result from extensive dissolution of the previously overlying and upgradient limestone rock. 3) Large internal drainage basins become concave through surface solution in a sufficiently sized catchment area. Solution sinkhole bowls form by solution of the rock around the sinkhole opening, which increases as more water is available to dissolve the rock (Hauwert et al., 2005; Hauwert, 2009). A sinkhole catchment area is roughly 300 times the bowl volume in same units (ie: ft3 vs ft2 or m3 vs m2) unless the relation is altered by certain less soluble rock types, sinkhole fill, or catchment area cut off. 4) Excavated material present that is equal the volume of the hole. The excavated material for a stock pond is sometimes used to construct the berm. 5) Historical aerial photographs and topographic maps can reveal the history of the depression. If the depression is not present in older aerial photographs then the feature could be a non karst closed depression. 6) The hydrostratigraphic rock unit where the depression occurs may point to its origin. The Regional Dense member is a softer, nodular 30 ft thick unit that was historically quarried for roadbase but rarely shows large solution features. Solution sinkholes are common in the Leached and Collapsed, Grainstone, Kirschberg, and Dolomitic members of the Edwards Group. 7) The presence and type of vegetation present in a sinkhole bowl may provide clues about the catchment size, the duration of water retention, and the infiltration rate. The prevalence and density of the plant community should be evaluated to determine if hydrophytic wetland vegetation is present. The National List of Plant Species that Occur in Wetlands is a publication that categorizes plant species according to their affinity for occurrence in wetlands. The indicator categories are obligate wetland plants, facultative wetland plants, facultative plants, and obligate upland plants. Training in wetland delineation may be useful in assessing water retention and infiltration rates.
12
Views of Quad Border Sink in the field (left) and on aerial photograph. Note the darker tone, circular shape, and contour depression that distinguishes the sinkhole on aerial view. The northern end of Quad Border Sink is dominated by three wetland plant indicator species, shown on the following photos, including Iva annua (Marsh Elder), Eleocharis montevidensis (Sand Spikerush), and Carex cherokeensis (Cherokee Sedge), which form an abrupt boundary between upland plants and wetland plants. Solution sinkholes with a wetland habitat are common found in southwestern Travis County. 1. Iva annua (Marsh Elder) – Facultative wetland plant indicator category. 2. Eleocharis montevidensis (Sand Spikerush) – Facultative wet plant indicator category. 3. Carex cherokeensis (Cherokee Sedge) – Facultative wet plant indicator category. 1.)
2.)
3.)
Generally, the most commonly occurring wetland indicator plant species found in wetlands in Travis County is spikerush (Eleocharis sp.). Being able to identify commonly occurring wetland plants, such as spikerush, is beneficial to recognizing wetlands and ponded internal drainage basin sinkholes. The ponding of water within a sinkhole should be evaluated with care. Note that in Appendix C, one assessment for Hwy 45 and the Tabor site suggested that broad soilfilled “features” ponding water recharged little or none. In contrast, TCEQ (2004) states that “soil filling in a sinkhole, solution cavity, solution-enlarged fracture or cave does not indicate that rapid infiltration cannot occur at the feature.” Slow drainage within ponded internal drainage sinkholes have not been documented and periodic observations from COA lands suggests these sinkholes completely drain in less than a few days after rains. Plugged drains within sinkholes can be temporary and may be reopened by natural 13
or anthropogenic activities (White, 1988). The TCEQ’s instruction to geologist states that the soil-filled ponded features “cannot be ranked as having a low relative infiltration rate unless it is further investigated.” Another reason why ponded internal drainage sinkholes should be considered to have high potential infiltration rates is because of their large catchment areas (as Ford, 2000 remarked for Quad Border Sink above). TCEQ (2004) states: “The size of the catchment area is a factor that should be considered in assessing the probability that a feature is capable of rapid infiltration. Features that drain a catchment area >1.6 acres and, therefore, have potential to serve as a significant recharge feature should be given a high probability (>35 points) of rapid infiltration. This assumption is based on the concept of feedback within a karst system—that is, when a flow path carries a lot of water, the openings that comprise that flow path are more likely to become enlarged and better connected through time than the openings on a flow path that has moved smaller volumes of water.” Some questions about ponded internal drainage basins, such as the speed water infiltrates in these sinkholes, the amount of the inflow that evaporates versus much recharges, and the source and age of the sediment filling the sinks, remain to be answered. Solution Sinkholes Solution of limestone or dolomite is most intensive where the water first contacts the rock surface. Aggressive dissolution also occurs where flow is focused in pre-existing openings in the rock, such as along joints, fractures, and bedding planes, focusing in the zone of water-table fluctuation where ground water is in contact with the atmosphere. Fractures expand into linear, open fissures from a combination of dissolution and pressure release near the surface. Some fissures and conduits develop well-integrated plumbing with the deeper subsurface and aquifer. Collapsed sinkholes, which already have well-developed hydraulic connection to the subsurface, may begin to capture surface flows and mature from surface dissolution. Features such as fissures, solution cavities, and collapsed sinkholes that have well-developed subsurface connection may develop circular and concave sinkhole bowls as they pirate larger catchment areas. The volume of a concave solution bowls is generally proportional to its catchment area (see page 7-8).
14
Collapse Sinkholes Collapsed sinkholes are examples of karst features that are not created by surface dissolution. Most collapsed sinkholes in the Austin area formed as the land surface eroded down to the roof of pre-existing caves. These caves probably formed near an old water table or by vadose flows moving from the surface to an earlier water table. Collapse sinkholes form when the stress field generated by a subsurface void reaches the surface (Newton, 1984). This depends on the width of the void and the thickness and competence of rock above the void to the surface. Collapse sinkholes form in many parts of the world due to human activities, such as pumping (loss of buoyant support for inherently unstable rocks), placing heavy loads (ponds, building structures, etc.) over shallow voids, or directing surface flow that speeds local dissolution (Aley et al., 1972; Newton, 1984). Because water tables are 200 feet deep or more in the Barton Springs Segment (as opposed to Florida where the water table is generally much shallower), catastrophic collapse is rare in this area. Collapsed sinkhole development accounts for the abundance of caves high on hillsides or hilltops where little catchment area is present. Over time, collapsed sinkholes may pirate larger catchment areas and transform into a concave-shaped bowl solution sinkhole. Barker Ranch Cave #1 is a typical example of a collapse sinkhole, as evidenced by its overhanging rim, convex rather than concave cross section, and large chamber volume compared to its relatively small catchment area.
Soil-piping sinkholes “Suffosion” or soil piping sinkholes occurs when unconsolidated overburden sediments infill preexisting cavities below them. This downward erosion of unconsolidated material into a preexisting cavity is also called “raveling” and describes both the catastrophic sinkhole and the more gradual formation of a collapse sinkhole. Erosion begins at the top of the carbonate bedrock and develops upward through the overlying sediments toward the land surface. A good example of a soil-piping sinkhole is visible about 500 feet south of Flint Ridge near the edge of the TxDOT right-of-way. 15
Stop 2 – Sandbur Cave Sandbur is a sinkhole and cave developed within the Kirschberg Member of Edwards Aquifer. In 2007, The COA used Sandbur Sink as a dye injection site, injecting rhodamine WT into the feature. Three nearby sites (Wildflower Center Cave, Hangtree Cave at Hwy 45 west of Mopac, and Bear Creek on the Tabor tract) were also injected with different dyes. All four tracers arrived at Barton Springs within two to four days of injection. Barton Springs is 10 to 14 miles from the four injection sites along the estimated flow paths. No tracer was detected in the Drip Pit of Flint Ridge Cave following injection of one tracer in nearby Bear Creek. Based on the tracing results, Bear Creek does not contribute to the Drip Pit in Flint Ridge, even though the Drip Pit has a lower elevation than Bear Creek. Also we learned that travel time from this area to Barton Springs is considerably faster than the three-year travel time originally estimated from Highway 45 (Appendix B).
Sandbur Cave
16
Small Sinkhole/Karst Depressions Small, soil-filled sinkholes or karst depressions (a term used by City of Austin geologists), such as those visible across the Tabor tract, are very common in the Recharge Zone of the Barton Springs Segment. Many of these are small, immature sinkholes are filled with soil, while others may be created by soil piping into a subsurface void. Typical ring infiltrometer installation used in Lindley (2005) study
A study by Adrien Lindley (2005) of the University of Texas tested the infiltration of small soil-filled sinkholes, some random features, and background soil using the ring infiltrometer shown in the photo above. Lindley (2005) found that small soil-filled and excavated sinkholes in the Edwards Aquifer Recharge Zone showed similar infiltration as background soil, generally at a rate of one inch within 6 to 16 hours (see results figure below). This background infiltration rate is not necessarily low, but is lower than U.S. Dept. of Agriculture (2007) estimated hydraulic conductivity of 0.5 to 5 hrs per 1 inch for Speck soils up to 20 inches deep. Two soil-lined small sinkholes showed slight improvement in infiltration rate after some excavation, although the sinkholes were not all completely excavated to open cavity. Lindley (2005) also found that many of the soilfilled small sinkholes/karst depressions were associated with bedrock bowls, suggesting that they were immature surface solution sinkholes. Photos on pages 19 to 21 illustrate an investigation of a karst depression where soil is piped down to a previously unknown subsurface cave. Much of the upland recharge occurs through the soil between large sinkholes (Hauwert, 2009). Runoff is not evenly distributed across the landscape but focuses in even poorly defined drainages and depressions. Infiltration does not necessarily occur evenly through the soil as a front, but may focus at macropores (Hillel, 1998). Surprisingly, Lindley (2005) measured the highest infiltration along a macropore created by juniper roots over the Trinity Aquifer.
17
Infiltration Rates Measured by Lindley (2005)
18
Example Examination of a Small Sinkhole or Karst Depression in Hays County
Photo by Brian Hunt, BS/EACD
Walking in transects across a site, a small karst depression (or small soil-filled sinkhole) in the soil is flagged, measured, and GPS located.
Photo by Brian Hunt, BS/EACD
Initial hand-excavation reveals solution cavity blowing air strongly. Photo of Nathaniel Banda, Barton Springs/Edwards Aquifer Conservation District (BS/EACD).
19
Photo by Bill Larsen of Zara Environmental
Further investigation reveals a cave that can be entered. Photo of Nathaniel Banda.
Photo by Nathaniel Banda of BS/EACD
Peering at a coral snake in the cave entrance.
20
Photo by Nathaniel Banda, BS/EACD
Bill Larsen of Zara Environmental looking into the cave. Note the classic convex (bell shaped) cross section of the collapsed sinkhole entrance.
21
22
Stop 3 – Barker Ranch #1 Barker Ranch #1 is a classic collapsed sinkhole, with a roughly 1 acre catchment and very subdued solution (concave) bowl, but underhanging entrance. A soil tracing investigation is being conducted on the Tabor tract by Brian Cowan, a University of Texas Department of Geological Sciences graduate assistant under the direction of Dr. Jay Banner, and in cooperation with the COA Watershed Protection Department (Cowan, et. al., 2007). This investigation tests connections between the surface and subsurface cave drips and examines soil and epikarst attenuation using both injected tracers and natural runoff chemistry. Barker Ranch #1 Cave was selected as a subsurface monitoring point because of its persistent drips (see figure and cave map below). By utilizing natural cave drips for monitoring, the natural flow paths through the soil and epikarst can be examined without disturbance. The study may be expanded to include simultaneous tracing to other cave drips. Injection sites for tracers are selected in low areas where runoff accumulates and recharge naturally focuses over soil-covered areas. These sites include often subtle macropores such as small sinkholes (karst depressions), solution cavities, burrows, as well as soil sites where no macropores are present. Moderate sinkholes and caves are not used for injection sites in this study since they quickly bypass the soil and epikarst zones. At this time, six injection sites have been tested within 350 feet of Barker Ranch #1. Tracers connect cave drips to specific surface areas contributing to its subsurface catchment area. Tracers, such as potassium bromide, ammonium carbonate, and iron standards, are injected under natural rain conditions. The attenuation of the solute and travel times from specific sites is measured. With sufficient traces the subsurface catchment area can be defined. When a trace does not clearly establish a hit or miss relative to background, repeated traces can increase the confidence of a result. An example of tracer breakthrough from a karst depression about 100 feet from Barker Ranch #1 is shown below. The chemistry of natural runoff is compared to cave drip chemistry to determine the ability of soil and epikarst to attenuate low (natural) levels of natural water-quality constituents across the drip catchment (subsurface catchment) area. Furthermore, by monitoring cave drip rate, the relative proportion of macropore/conduit flow can be distinguished from slower drainage from matrix flow in soil and epikarst using hydrograph separation. An example of Barker Ranch #1 drip rate response to rainfall is shown below. Magnesium (Mg), strontium (Sr), and calcium (Ca) provide relative indicators of groundwater residence time. Waters with relatively low Mg/Ca and Sr/Ca ratios (ie., after initial tracer breakthrough) have short residence time, whereas waters with relatively high Mg/Ca and Sr/Ca ratios have longer residence time (ie., such as baseflow conditions). A comparison of dripwater trace element concentrations with injection tracing results indicates that groundwater residence time is reflected by local geochemistry, even on short timescales of days to weeks. These findings are consistent with drip-rate data, which shows that high Mg/Ca and high Sr/Ca dripwaters are correlated with relatively low drip rates. The rapid infiltration of water through soils and significant variability in dripwater chemistry have important implications for decisions on water resource 23
management and interpretations of trace element variations in speleothems for paleoclimate indicators
Diagram illustrating the soil and epikarst in Barker Ranch #1 Cave. The cave drips are about 20 feet below the ground surface. Illustration by Brian Cowan from Krejca et al., 2000 Breakthrough of 500g KBr in Barker Ranch #1 Drip from Site 1 0.4
Concentration in BR#1 Cave Drip (mg/l ) .
0.35
KBr injected
0.3 Br
K
rainfall (inches/10)
0.25
0.2
3 to 7 hr breakthrough
0.15
0.1
0.05
0 03/13/07
24
03/15/07
03/17/07
03/19/07
03/21/07
03/23/07
03/25/07
03/27/07
03/29/07
25
Cave Map for Barker Ranch #1
26
Stop 4 – Flint Ridge Cave and Sinkhole Flint Ridge is a cave that descends about 150 feet below the surface on the Tabor Tract. It has a large internal drainage catchment of about 69 acres in area. Note that upland drainages across the site change from well-defined to undefined over short distances (see Tabor Study Site map). Since 2003, the flow into Flint Ridge Cave has been measured using flumes and weir. Periodically water samples are collected for baseline runoff quality. A highway is proposed to pass over portions of Flint Ridge and its right-of-way is about 150 feet from the cave entrance. Baseline collection of water quality and flow are used to measure recharge components and will be used to evaluate changes in future conditions. Site assessments have provided conflicting opinions regarding recharge of Flint Ridge Cave and the Tabor tract (Appendix B). Flint Ridge was originally estimated to have a catchment area of only 0.75 acres. Water infiltrating on the Tabor tract was thought in one assessment to discharge from local seeps along nearby Bear Creek rather than recharge the aquifer. Recharge from this area was stated to require 1,035 days to travel 10 miles to Barton Springs at an estimated rate of 51 feet per day. Presentations have been recently made that hypothesized the largest drip in Flint Ridge Cave, the Drip Pit is actually fed by Bear Creek flows rather than from rainfall on the surface over the drip location. This was surmised since the elevation of the Drip Pit (located beneath the proposed Highway 45) was lower than nearby Bear Creek. A number of recent studies have tested the hypotheses presented in the assessments and found large errors in these statements. The surface catchment area to Flint Ridge Cave is actually closer to 69 acres rather than 0.75 acres, based on field gps survey, and a professional survey of a large portion of the catchment (Hauwert, 2009). Furthermore a 0.75 acre catchment could not support the measured flood flows generated by rainfall when the soils were relatively saturated. In 2007, dye was injected at two sites of Tabor tract and two other nearby caves arrived at Barton Springs much faster than 1,035 days, instead all four arrived within two to four days (Hauwert, 2009). One of the four tracers was injected along a one-mile stretch of Bear Creek closest to Flint Ridge Cave. The tracers were not detected anywhere inside Flint Ridge Cave including the Flint Ridge Drip Pit. Shallow cave drips and standing water are present throughout Flint Ridge above Bear Creek elevation, particularly in the Balcony Room of Flint Ridge. A report of the 2007 traces is being prepared by the City of Austin and Barton Springs/Edwards Aquifer Conservation District. This report should clarify some of the errors made in site assessments of the Tabor tract.
27
28
Identifying Caves The following criteria may help identify caves: 1) Enterable voids. A subsurface extent that can be accessed and mapped. 2) Interpreted karst origin. Rounded solution cavities and fissures, with widths generally one foot or more, as signs of focused rock dissolution. Large, well-defined solution or collapsed bedrock bowls are good indicators of likely caves. 3) Karst features with airflow. Airflow can have daily cycles of blowing in or out. Consequently, depending on the time of day (and perhaps seasonal cycles) a feature that may have strong airflow might not show strong airflow when examined. When strong airflow is observed blowing out of a feature, the feature can be distinguished as having a large volume of open subsurface extent. 4) Drainage patterns. The termination of a well-defined surface drainage, points of surface flow loss in a drainage channel, and large topographic depressions are good cave and sinkhole indicators. Accumulation of debris washed to a feature or a vortex over a creek-channel swallet suggest both significant catchments and subsurface connections associated with caves. 5) Presence of cave-adapted species. The appearance of common cave-adapted troglophiles, trogloxenes, and troglobites such as Plethodon albagula (Slimy salamander), Ceuthophilus sp. (cave crickets), Syrrophus marnockii (Cliff chirping frog), bats, etc. are also good cave indicators (USFW, 2006). 6) Fill extending into the subsurface. Rock, sediment, trash, or debris piles that have subsurface extent, and openings between boulders where fines washed through, particularly were airflow blows into or out of the pile, may indicate filled caves. Historically local ranches have covered caves, perhaps to prevent injury to livestock or residents and to dispose of waste (Hauwert, 2009 Appendix A). 7) Geophysical techniques may be used to indicate possible caves. These techniques are discussed in submitted papers by Dr. Mustafa Saribudak, Dr. Jack Holt, and student papers in Appendix I. Cave Maps Cave mapping today generally use compasses, inclinometer, and tape measure to measure from station to station within sight of each other. The height and depth of the room at each station location, as well as details of the passage are documented to include on the cave map. From the cave map we can examine the subsurface extent. The cave map can be scanned and georeferenced so that its outline can be plotted in relation to surface features in a mapping program such as ArcGIS. After excavating rocks and debris from the entrance, members of the University of Texas Grotto (Mark Minton, Dale Pate, Bill Russell, and Nancy Weaver) explored and mapped Flint Ridge Cave in 1984 and 1985. Flint Ridge Cave now serves as a preserve for rare cave species of concern as part of the Balcones Canyonland Preserve.
29
Cave Map of Flint Ridge Cave
Cave Radio Location One technique to accurately locate the surface location and depth to points into a enterable cave passage is to use a cave radio. A transmitter is taken into the cave, while a receiver is used on the surface to locate the receiver location and depth. Points in Flint Ridge Cave were radio located in 2006 for the City of Austin. Cave specialist Keith Heuss constructed the instruments and measured angles from the transmitter using a receiver. The transmitter was placed and sketched at predesignated locations in the caves by William Russell, Julie Jenkins (both Texas Cave Management Assoc.) and Mark Sanders (City of Austin). Concrete monuments were placed on the surface over seven designated locations in the cave by Park Smith and Environmental Youth Corps staff. The project was coordinated and monument sites were GPS located by Nico Hauwert (City of Austin). From cave radio data, we estimate the elevation of the Flint Ridge Drip Pit flows are about 653 feet mean sea level, compared to elevations of about 730 to 750 feet msl for Bear Creek, which is located about 1,500 feet south and west of the Drip Pit in Flint Ridge Cave. While cave radio locations can be located within a foot of transmitter locations, there is greater error with depth measured with cave radio. Variations in depth measurements varied within about 15 feet between 7 different readings for Flint Ridge Drip Pit. 30
Geophysical Surveys Geophysical surveys conducted over Flint Ridge Cave are presented as submitted papers and in Appendix I. The geophysical surveys relied on cave radio location of known cave passages.
31
Photo Tour of the Upper Flint Ridge Cave
Photo by Mark Sanders.
Near the entrance, the drip room, about 25 feet below the surface, contains many active soda straws, stalactites, stalagmites, and columns that are signs of drainage through the overlying soil. Cave passage is developed within the Kirschberg Mbr and trends southeast.
Photo by Mark Sanders.
In the photo above, Kathleen O’Connor looks at ceiling formations in the Balcony Room, about 150 feet from the entrance and about 40 feet below the surface. Cave drips provide signs of storage and infiltration in the overlying soils and epikarst rock. Although drips occur throughout this and other local caves, major drip areas are localized to a few locations where drips continue far into drought intervals, even in the shallowest portion of the cave. The focusing of drips suggests vadose flow is localized.
32
Photo by Nico Hauwert
This photo shows David Johns entering the Breakdown Room, lying about 50 feet below the surface. Large bedload gravels demonstrate the high flows entering from the entrance during some storms. Soft, white, pulverulitic beds can be seen here, consisting of poorly cemented euhedral dolomitic grains which are characteristic of the Kirschberg Member.
Photo by Nico Hauwert.
The Culvert Crawl passage is developed within the upper Dolomitic Member and resembles a smoothly scalloped storm sewer. The passage abruptly rotates perpendicular to a trend southwest and is highly fracture-controlled.
33
Photo by Mark Sanders.
Kathleen O’Connor crawls in a Flint Ridge Cave Culvert Crawl passage developed within the Dolomitic Member. This passage is highly localized along a fracture. Pooled water is always present on the floor while formations above drip into the passage.
34
Submitted Papers This field trip was fortunate to have submitted papers and student research reports (Appendix I) that discuss concepts presented here in further detail. The author is solely responsible for content of the article. These do not necessarily represent the views of the City of Austin or the Texas Commission on Environmental Quality.
35
Does Rainfall Infiltrate Through Upland Soils To Recharge The Edwards Aquifer? Nico M. Hauwert, Ph.D., PG (Excerpt summary from Hauwert, 2009, modified Feb. 2010) DeCook (1957) made the following observations about recharge of groundwater to the Edwards Aquifer in Hays County: “Of the precipitation that falls on the outcrop, a part may run off on the surface to other areas, but this amount is relatively small, as the topography is subdued in profile and stream-cut valleys are widely spaced. An undetermined amount of the rainfall is returned to the atmosphere by evaporation, intercepted or transpired by vegetation, or absorbed as replenishment to the soil moisture. The remainder infiltrates to the underground reservoir. Although data are not at hand to illustrate how much of the precipitation ultimately reaches the Edwards ground water, it is generally believed that it is a relatively high percentage in limestones such as the Edwards as compared with that in other rock types. In the San Marcos area, the soil veneer on the Edwards is relatively thin, and differential solution has produced many open spaces in the rock for the reception of water.” The difficult task of quantifying the amount of recharge to the Barton Springs Segment was begun in the 1980’s, when recharge was measured from each watershed based on creek flow and measured discharge from Barton Springs (Slade et al., 1986). It was estimated that 0.89% (up to 2.6% depending on interpretation) of the direct rainfall over the recharge zone became authogenic recharge to the Barton Springs (Woodruff, 1984). Slade et al. (1986) stated that the contribution of Barton Creek to Barton Springs could potentially be found to be less with further investigation. Groundwater tracing since 1996 has suggested that most of the Barton Creek channel on the Recharge Zone does not contribute recharge to Barton Springs, consequently the total creek recharge contribution may be significantly less. Another method of estimating reasonable recharge values is to examine other areas of similar rock type (limestone) as well as different rock types across the world. A recharge value of 1% of rainfall was found to recharge the non-karst Eagle Ford Shale in the Waco area (Harrison, 1996) over a one year water balance. In karst areas composed of limestone rock, the amount of recharge to aquifers from rainfall is much higher, generally greater than 20% (table 1). One limit of comparing recharge values from different areas is that previous investigations do not necessarily distinguish authogenic (recharge from direct rainfall on the recharge zone) from allogenic recharge (upgradient runoff sources such as the contributing zone). Large upgradient allogenic recharge sources will tend to increase the total recharge derived from rainfall. Furthermore, for some of the same karst areas (such as the adjacent Trinity limestone aquifer) recent estimates of recharge from rainfall tended to be higher than older values (Table 1). The variety of studies also utilized different methods, such as recharge estimates based solely on mass balance comparison of chloride between rainfall and spring discharge (such as Jones et al., 2000), that have the potential to greatly underestimate recharge if chloride is enriched by processes other than evapotranspiration (Hauwert, 2009). 36
Table 1 % Recharge/
% Upland Recharge/
Direct Precip.
% Major Creek Recharge
28-32% 0.9 to 2.6%*
15% / 85%
>10% ---
56% / 44%
Location
Aquifer, Basin,
Source
or Region Hauwert, 2009
Barton Springs Seg.
Edwards
Barton Springs Seg.
Edwards
Woodruff, 1984
San Antonio Seg.
Edwards
Huang and Wilcox, 2005 Ockerman, 2002
Bexar County TX
Edwards
1.5%
Central TX
Trinity
Muller and Price, 1979
4%
Central TX
Trinity
Ashworth, 1983
11%
Central TX
Trinity
Kuniansky, 1989
6.7%
Central TX
Trinity
Bluntzer, 1992
7%
Central TX
Trinity
30%
Uvalde Cty TX
Trinity Carboniferous Limestone
54%
100%/0%
Mendip Hills, England
Mace et al., 2000 Dugas, 1998 Atkinson, 1977
33%
Crimea, Ukraine
15-20%
Barbados
Uplands
Jones et al., 2000
Dublyanskii et al., 1984
25-30%
Barbados
Below 2nd Cliff
Jones et al., 2000
60%
Guam
Mink and Vacher, 1997
67%
Guam
14-19.5%
Morocco
Jockson et al., 2002
23-42%
Tunisia
36-91%
Tunisia
Eocene Ls
10-45%
Israel
W. Galilee
53%
Israel
Na'aman Spr
41%
Syria
Damascus
45%
Syria
Ghab
Voute, 1961
52%
Greece
Lilaia
51%
Greece
Parnassos-Ghiona
12-124%
Hungary
Tettye
Aronis et al, 1961 Burdon and Papakis, 1961 Kessler, 1957
Plio-Plest.
Bolelli, 1951 Tixeront et al., 1951 Schoeller, 1955 Goldschmidt, 1960 Mere, 1958 Burdon, 1960
*upland direct recharge only, total recharge estm. to be 5.95% which includes rainfall and recharge of contributing area flows
Modified from Hauwert, 2009; see for complete literature citings.
From 2004 to 2005, upland recharge was measured more directly on the Barton Springs Segment (Hauwert, 2009) using a water-balance method within an internal drainage microbasin over a 1.4-year interval. Evapotranspiration, measured using eddy covariance and bowen ratio methods using a 45-ft high tower, was approximately 68% of rainfall. Discrete recharge or runoff directly entering the cave drain, measured using flumes at both HQ Flat and Flint Ridge sinkholes, was 5% of rainfall over the 1.4 year interval, decreasing to 3% of rainfall over a longer interval of average rainfall. By subtracting measured ET and discrete recharge from measured rainfall, the diffuse recharge through upland soil-covered areas was measured to be 26% of measured rainfall. This 1.4 year water budget interval had 21% higher than average rainfall and was selected where the measured soil moisture was similar at the start and end to eliminate possible changes in soil-moisture storage. The 1.4 year water balance 37
measured 32% of rainfall recharging within the microbasin. Most of the recharge measured occurred on the slopes of the large sinkhole basin (46 acres) that are similar in appearance to areas with similar slopes and soils across the Recharge Zone. This was the most precise measurement of upland recharge at a single site (averaging about a square mile in area) on the Barton Springs Segment to date. This study is limited in that it provides only a single measurement over a relatively short interval experiencing rainfall 21% higher than average. It is intended that this water balance be extended to about 4 years of historically average rainfall. Since ET is the largest component next to rainfall in the water budget, it is hoped that data from other climate towers can be used to evaluate expected variation between Central Texas sites. A third method of measuring recharge compares chloride concentrations in rainfall to subsurface chloride concentrations and assumes all chloride enrichment is due to evapotranspiration. However, in the Barton Springs Segment, chloride increased over an order of magnitude between cave drips and spring discharge, suggesting chloride enrichment from a source other than ET (Hauwert, 2009). By comparing chloride concentrations in rainfall with concentrations in cave drips on the Tabor site yielded recharge values of about 50% of rainfall. Because infiltration through the soil may not be evenly distributed but may localize along macropores, it can be difficult to examine soil infiltration. Soil studies may be limited by small scale infiltration rings or trenches focused in areas of thick soil where macropores significant for infiltration may be missed. Soil tracers, particularly organic dyes, may not conservatively penetrate the soil as surface water may. Runoff is not evenly distributed across the land surface of the uplands Edwards outcrop area, but is commonly focused in slight depressions and swales. Examination of soil using smallscale examinations (Wilding and Woodruff, 2008) have suggested that soils and unweathered bedrock surfaces over the Trinity and Edwards Aquifer are incapable of transmitting the amount of recharge estimated through recent aquifer-wide and microwatershed scale water balances (Hauwert, 2009) and cave drip observations at subsurface catchment scale (Cowan et al., 2007). While a random, small scale examination of soil infiltration on karst can provide valuable information about portions of the soil matrix, it is possible that judging infiltration by such small scale tests alone without including the components of macropores and karst features may be analogous to testing the material composing a kitchen colander or bathtub and concluding that these structures do not drain. In karst areas such as the Edwards Aquifer, we are fortunate to be able to physically enter the subsurface to make direct observations about soil infiltration. In shallow caves, formations are common signs that rainwater infiltrates through the soil across the outcrop areas of the Edwards Aquifer. Since most cave drip flows are focused to a few locations within each cave across the Edwards Aquifer, we can see that much of the flow through the epikarst and vadose zones are localized along specific paths. By monitoring cave drip rate, we can distinguish a component of rapid macropore flow as well as a component of slow drainage that can extend far beyond precipitation events. Soil tracing to cave drips, particularly where natural precipitation events are utilized, is a particularly effective method to study soil infiltration and attenuation. It is clear that further research is needed to examine soil infiltration and recharge to the Edwards 38
Aquifer. However, the weight of evidence from studies examined above is that about a third of rainfall directly recharges upland soil-covered areas of the Edwards Aquifer.
Dessication crack (macropore) developed during the 2009 drought on the surface over the Flint Ridge Cave near the Balcony Room. Photo by Brian Cowan, UT Geological Sciences.
References Cowan, B.C., Banner, J.L, Hauwert, N.M., and Musgrove, M.L., 2007, Geochemical and physical tracing of rapid response in the vadose zone of the Edwards karst aquifer: Geological Society of America Annual Meeting Paper no. 69-3. DeCook, K. J., 1957, Geology of San Marcos Quadrangle, Hays County, Texas: M.A. thesis, University of Texas at Austin, Texas, 90 p. Harrison, A.D., 1996, Recharge mechanisms of swelling clays and shales, Central Texas: M.S. thesis, Baylor Univ., Waco. 187 p. Hauwert, Nico M., 2009, Groundwater Flow and Recharge within the Barton Springs Segment of the Edwards Aquifer, Southern Travis County and Northern Hays Counties, Texas: Ph.D. Diss., University of Texas at Austin, Texas. 328 p. Slade, R. Jr., Dorsey, M., and Stewart, S.,1986, Hydrology and water quality of the Edwards Aquifer associated with Barton Springs in the Austin Area, Texas: U.S. Woodruff, C.M. Jr., 1984, Water budget analysis for the area contributing recharge to the Edwards Aquifer, Barton Springs Segment: in C. Woodruff and R. Slade, ed., Hydrogeology of the Edwards Aquifer-Barton Springs Segment, Travis and Hays Counties, Texas, Austin Geological Society Guidebook 6, p. 36-42. Woodruff, Jr. C.M. and Larry P. Wilding, 2007. Bedrock, soils, and hillslope hydrology in the Central Texas Hill Country, USA—Implications on environmental management in a carbonaterock terrain, Environmental Geology.v. 55, no. 3, p. 605-618.
39
Perspectives of the Effectiveness of Structural Runoff Filtering Systems on the Edwards Aquifer Recharge Zone Ineffectiveness of Structural BMP’s on the Recharge Zone Raymond M. Slade, Jr., Hydrologist As part of the review process for the Barton Spring Salamander Recovery Plan, a local consulting engineering company presented a hypothetical scenario for a large, highintensity (44% impervious cover) dwelling development using Best Management Practices (BMPs) consisting of a sediment pond and irrigated runoff. They claim that the development would meet SOS water-quality standards for runoff because the BMP would remove 100% of the urban-related contaminants. The proposed development would occur on the recharge zone of the Edwards aquifer and the scenario assumes runoff to the BMP to represent about 25% of rainfall from the developed area. However, the scenario does not include the water-quality impact from the portion of runoff that is lost as recharge in route to the BMP. That recharge, which represents runoff from land with 44% impervious cover, would contain substantially degraded water quality. Runoff Lost as Recharge on the Recharge Zone Only small volumes of rainfall directly infiltrates to the Edwards aquifer through soils. Most recharge occurs through critical environmental features (such as caves, faults, and fractures) located in streams and in areas with overland flow. The amount of runoff lost as recharge from developed areas depends upon characteristics of the land and development. Recharge would be substantially increased for sites with critical environmental features, especially if such features were contained in streambeds or overland flow areas with substantial drainage basins. Two known studies have documented relative volumes for runoff lost as recharge on the recharge zone. Table 7 of the report "Stormwater pollutant loading characteristics for various land used in the Austin Area" (City of Austin, 1990) shows, for a 40% impervious cover area, runoff to be 15.8% of rainfall for sites on the recharge zone and 22.2% of rainfall for sites not on the recharge zone. These data are based on small basins, generally less than about 150 acres. The difference in runoff (6.4% of rainfall) represents estimated losses to recharge for basins on the Edwards aquifer. Therefore, for the development scenario presented above, 29% of runoff (6.4 / 22.2) would be lost as recharge. Based on long-term gaged data for rainfall and streamflow, Woodruff (1984) made an independent analysis of runoff and recharge that included the entire recharge zone. He found that 40% of runoff is lost as recharge while 60% of runoff is not lost as recharge.
40
Water Quality of Runoff Lost as Recharge Two substantial studies, based on local data, have been conducted to document the relation between impervious cover and water-quality characteristics. Veenhuis and Slade (1990) based their analyses on data from many large watersheds in the Austin area, while the City of Austin (1990) based their analyses on data collected from many small watersheds in Austin. Both studies found that degradation from a basin with 40% impervious cover would cause, for many water-quality constituents, increases in waterquality concentrations of many hundreds or thousands percent. Therefore, for the theoretical development, a large percentage of water-quality loads would be removed for the runoff reaching the BMP, however, about 29% of the runoff would recharge the aquifer with water-quality concentrations many times greater than those for natural conditions. Based on this scenario, the overall effectiveness of the BMP in removing total water-quality loads for the development scenario is reduced by values at least 71 percent (100-29), thus deeming the BMP not to meet SOS water-quality standards. Leakage From BMP Storage Ponds on the Recharge Zone Most structural water-quality BMPs use large storage ponds to impound runoff prior to filtering by sand, wet ponds, or irrigation. Large volumes of water stored in such ponds can be lost as recharge. Most ponds on the recharge zone are designed to be lined with impermeable layers, often consisting of a clay material. However, such layers can be damaged by maintenance, vandalism, or large storms. For example, many years ago a thorough water-budget analysis of rainfall, inflow and outflow was performed for many storms on the largest sand-filtering basin at Barton Creek Square Mall Shopping Center. The data proved that substantial volumes of water were being lost as recharge to the Edwards aquifer from the pond. Apparently, the effectiveness of the clay liner as an impermeable layer was compromised during the process of removing and replacing the sand. Data for such a water-budget analyses generally do not exist for other BMPs on the recharge zone, thus the amounts of recharge lost from other ponds is unknown. As demonstrated above, the water quality for recharge from leaking storage ponds on developed areas would be substantially degraded from that of natural conditions. Additionally, infiltration rates to the aquifer from such ponds likely would be substantially greater than would occur without ponds, because large water depths in the ponds cause increased water pressure over faults and fractures, thus forcing more recharge than would occur naturally. Furthermore, because of small streams and steep slopes, most streams on the recharge zone contain water while rainfall occurs and only for short periods thereafter. However, storage ponds frequently contain water, thus causing recharge to occur for much longer periods than would occur naturally. Recharge loads are calculated by multiplying recharge volumes by water-quality levels. As shown above, recharge volumes from storage ponds likely would be increased, and the water-quality of the recharge would be substantially degraded. Therefore, water-
41
quality loads from a leaking pond likely would be many times greater than would occur without such a pond. Therefore, for a leaking storage pond on the recharge zone, the effectiveness of the BMP in reducing total water-quality loads is further reduced based on the amount of leakage.
Design and Operation of BMPs Structural BMPs are more effective in removing total water-quality loads for areas away from the recharge zone than for areas on the recharge zone. Also, BMPs on the recharge zone probably are more effective for sites on small developed areas with impervious covers approaching 100%. For such sites, only small volumes, if any, of runoff would be loss as recharge in route to the BMP because the high-level impervious cover would prohibit or substantially limit recharge volumes. However, for sites with extreme impervious-cover levels, high performance BMPs would need to be used and maintained in order to effectively lower the high water-quality levels from such developments. The featured development scenario and others typically assume that all runoff from the development enters a single pond and further assumes that no runoff from outside the development boundaries enter the pond. Such circumstances rarely occur in the environment. Also, most ponds are designed to capture about 0.5 to about 1.5 inches of runoff and are designed to be drained in one to many days. Any rainfall depths that exceed the designed depth before the pond is drained would outflow the pond without being filtered. Many storms per year typically exceed these rainfall limits, and thus large-water-quality loads from such storms are not filtered. The removal efficiencies also assume that all filter media are operating at peak performance and equipment failures do not occur. For example, most water-quality BMPs in the Austin area represent sand filters, many or most of which receive little if any maintenance. The effectiveness of these systems in reducing water-quality levels becomes extremely limited if maintenance is not performed because the sands become saturated with organic carbon and other urban-related contaminants. Experience in the operation of irrigation systems has identified several maintenance problems. Pumps for irrigation systems often are deemed inoperative because intakes become clogged with debris or sediment and sprinkler heads become clogged with sediment. Also, the entire system is susceptible to becoming inoperative due to vandalism. Irrigation systems require frequent inspections and maintenance to remain functional, however, not all irrigation systems receive such inspections and maintenance.
42
Environmental Factors Affecting Irrigation Systems Preliminary data and information indicate that irrigation of urban runoff can be very effective in lowering water-quality levels. However, many factors influence the effectiveness of these systems. Soil particles on irrigated areas should be small enough to attenuate irrigated water—very permeable soils with large amounts of rock or sand allow contaminated water to quickly move to the underlying Edwards aquifer with minimal filtering. However, soils with large amounts of clay will not allow irrigated water to enter the soil zone. For such conditions, irrigated water could be retained on the surface as unfiltered, and then runoff or be flushed away by rainfall. Greater soil depths allow for greater attenuation of irrigated contaminants. Also, high density of vegetation provides greater removal of contaminants through direct absorption of irrigation and uptake and attenuation of contaminants through roots in the soils. Flat land slopes retain irrigated water on site for greater durations, which causes less runoff from the irrigated site and thus greater filtering through infiltration. Finally, irrigation on soils wet from rainfall could cause reduced or no infiltration, thus non-filtered water could runoff from the site. Holding ponds have limited storage capacity thus irrigation has often occurred during rainfall events in order to prevent ponds from overflowing. During such conditions, it is likely that much or most irrigated waters would runoff from the site. A Non-Conventional BMP A runoff-filtering system manufactured by AquaLogic Inc., in San Antonio, Texas is being used on the Edwards Aquifer recharge zone in the San Antonio area. The system contains a sediment-settling basin and standpipes containing 10-micron filtering media, designed to filter all received runoff. AquaLogic Inc. provides frequent inspection and maintenance via contract with property owners, thus assuring that the system probably retains peak or near peak performance. Maintenance includes removal of all material from the sediment pond and replacement of filter media in the standpipes. Although the effectiveness of this system has not yet been tested for most urban-related contaminants, it is likely as effective or more effective than sand filters. Also, it might be superior to other BMPs simply because it receives scheduled and mandated inspections and maintenance.
43
Resistivity and Natural Potential Anomalies over Flint Ridge Cave Mustafa Saribudak, Ph.D., Environmental Geophysics Associates, Austin, TX
Objective and Geology
Geophysical surveys [resistivity and natural potential (NP)] were conducted over a known cave (Flint Ridge Cave) in order to determine its geophysical signature. This study was part of a larger project for the Lower Colorado River Authority (LCRA). Flint Ridge cave occurs in Kainer Formation of Kirschberg and Dolomitic members. The cave was studied by George Veni and Associates in 1999. Depth and map views of the cave are shown in Figure 1, which are reproduced from Jenkins et al. al., June 1984December 1999. The cave starts with an entrance-sinkhole and trends in multiple directions with multiple rooms.
Geophysical Methods
Resistivity imaging is a survey technique, which aims to build up a picture of the electrical properties of the subsurface by passing an electrical current along electrodes and measuring the associated voltages. This technique has been used widely in determining kars t featu res, such as voids, and s ubsurface s tructures, su ch as faults and fractures. In this study, we used a dipole-dipole resisivity technique with 28 electrodes, which is more sensitive to horizontal changes in the subsurface, and provides a 2-D electrical image of the near-surface geology. Natural potential method (NP) measures the weak ambient electrical current in the subsurface. Natural electrical currents occur everywhere in the subsurface. In karst investigations we are concerned with the unchanging or slowly varying direct currents (d.c.) that give rise to a surface distribution of natural potentials due to the flow of groundwater within permeable materials. Differences of potential are most commonly in the millivolts range and can be detected using a pair of non-polarizing electrodes and a sensitive measuring device (i.e. a voltmeter). NP anomalies may also be caused by topography, changing soil conditions, and faults and fractures in the subsurface.
Geophysical Work at Flint Ridge Cave Resistivity and NP surveys were performed across the part of the cave called “Formation Pit.” A bench marker for this location was observed on the surface and marked on the geophysical profiles for reference purposes. Resistivity data show high resistivity and low resistivity anomalies along the profile with sharp contacts (Figure 2). The high resistivity values with red and yellow colors (i.e., 1000 to 10000 Ohm-m) indicate a syncline-like structure beneath the cave entrance. In
44
addition, low resistivity rocks breaches into the bedrock to the NW of the Formation Pit. This low resistivity anomaly could be due to a paleokarst feature. The Formation Pit is about 5 meter (16 feet) deep from the surface based on the cave map given in Figure 1. Thus I put this information on the resistivity map to show the relationship between the cave and the resistivity data (see Figure 2). NP data collected along the profile is given in Figure 3. The benchmark location (Formation Pit) is shown on the profile for reference purposes. The NP data indicates significant anomalies starting at between stations 40 and 220 feet. These anomalies are likely caused by cave chambers and/or passages and/or geological contacts. It is important to note that the NP data show a significant gradient in the southeast direction. This could be due to three reasons: First, a fault crossing the geophysical profile; second is the hydraulic gradient; and third is the combination of both above. Resistivity and NP data are shown together in Figure 4. The resistivity changes between low and high resistivity values along the profile correspond to sharp changes on the NP data indicating presence of karstic features.
Acknowledgment I thank you LCRA for permission to make this information accessible to the geologic community in Austin and somewhere else.
45
46
47
URBAN GEOPHYSICS: A MAPPING OF MOUNT BONNELL FAULT AND ITS KARSTIC FEATURES IN AUSTIN, TX Mustafa Saribudak, Environmental Geophysics Associates, Austin, TX
Abstract Although most karstic regions are characterized by caves, collapsed features, and sinkholes, such features often do not have surface expressions, and their presence may go unrecorded. Central Texas and the Greater Austin metropolitan area have been built on the karstic limestone (Lower Cretaceous of Glen Rose Formation and Edwards Aquifer) in the Balcones Fault Zone (BFZ), and their growth is expanding. Near-surface karst features in the Austin area have a profound effect upon geotechnical engineering studies, such as structural foundations (residential buildings, shopping malls), utility excavations, tunnels, pavements and cut slopes. Thus the practice of geotechnical engineering is and has been a challenging proposition in the Austin area. Geophysical methods are sporadically used to estimate the locations and parameters of these karst features prior to any of these above-mentioned geotechnical studies. Opinions concerning the effectiveness of these geophysical surveys are mixed, and geophysical techniques are not generally recognized as primary tools in engineering-scale studies. However, remarkable advances in the manufacturing of geophysical instruments over the last ten years have made geophysics a viable tool for geotechnical studies of these karstic features. Data quality has been increased by the advent of continuous data collection. The data are better processed and interpreted by new and improved software packages, which produce improved sub-surface imaging and mapping. Thus integrated geophysical surveys can provide new insight into the near-surface karstic features in the Glen Rose Formation and Edwards Aquifer. I have conducted geophysical surveys (ground penetrating radar [GPR], resistivity imaging, magnetic [G-858], conductivity [EM-31] and natural potential [NP]) at three locations where the Mount Bonnell fault (MBF) is present, along the northern limiting boundary of the BFZ. Results indicate that all methods successfully imaged significant karst anomalies across the known fault locations. Integration of all these anomalies provides a much better understanding of near-surface geology defined by the caves, voids, collapsed materials, sinkholes and the fault itself.
Introduction A study of the geologic map of Austin by Garner et., al., (1976) shows that normal faults along the BFZ are some of the main features, if not the primary features, that have shaped the geology and physiography of the city and its environs. At the regional scale, faults have positioned the geologic unit into a framework that juxtaposes contrasting rock, soil, and terrain, thereby establishing a major physiographic boundary: the Balcones Escarpment, which extends through west Austin, separates the Edwards Plateau to the west from the Blackland Prairies of the Gulf Coastal Plain to the east (Collins and Woodruff, 2001). The Balcones escarpment, with a topographic relief as great as 300 feet in Austin, is a fault-line scarp, and consists of normal faults, which dip toward the east and southeast. The BFZ’s most prominent fault is the Mount Bonnell fault, which composes the northernmost part of the fault zone
48
with a throw of near 600 feet. The Lower Cretaceous Glen Rose Formation is at the surface to the west of the MBF, while east of the fault zone younger rocks of Edwards Aquifer are at the surface (Figure 1). Geophysical methods have been an important component of effective hydrogeological investigations over the Edwards Aquifer. Geophysical surveys employing variety of electrical and electromagnetic methods have been used to successfully map stratigraphy, geologic structure, and depth to the water table in major aquifer systems (e.g., Fitterman and Stewart, 1985; Connor and Sandberg, 2001). In this study, however, I demonstrate the utility of integrated surveys for the near-surface characterization of the MBF in the Austin area (Figure 2). To my knowledge, this is the first application of integrated geophysical techniques to the characterization of faults, fractures, caves, sinkholes and collapsed features in the metropolitan Austin area. The geophysical surveys were performed at the intersections of: 1) Height Drive and Highway 360; 2) West Park Drive and Highway 360; and 3) Bee Cave Road and Camp Craft Road (Figure 2). Conductivity, magnetic, GPR and NP methods were chosen for their ability to very rapidly map variations of their respective physical attributes (e.g., conductivity, magnetic susceptibility, dielectric contrast and ambient electrical current in miliVolts) within the surface. 2D resistivity imaging surveys were conducted to provide information about variation in electrical resistivity as a function of depth. Results of these surveys are described in the following.
Figure 1: Balcones Fault Zone Portion of the Edwards Aquifer in Central Texas.
49
Figure 2: Surface geology of geophysical survey sites. Geologic interpretation modified from Hauwert, 2009 and Rodda et al. (1970) by Nico Hauwert.
Geophysical Results Height Drive Site at Highway 360 A site map of the study area including the location of geophysical profiles and the MBF is shown in Figure 3. The magnetic and conductivity data are shown in Figure 4. Resistivity and NP data are shown in Figure 5. And 3-D GPR depth slices are given in Figure 6. The magnetic data indicate a high anomaly between the stations at 270 and 290 feet, whereas the conductivity data shows a high and low between the stations at 270 and 310 feet. These anomalies correspond to a very significant resistivity anomaly between stations 280 and 320 feet. The source of the anomaly appears to be the soil-filled material on top of a cave. The rest of the resistivity data between the stations at 320 and 418 feet shows significant resistivity anomalies: Low resistivity values shown in blue branch into higher resistivity values. This type of resistivity anomaly usually is an indicator of cave structures. The NP anomaly 50
shown in Figure 5, along with the resistivity data show a very significant “sine-wave” shaped anomaly where the MBF is located. In addition, the rest of the NP anomaly shows cave-like anomalies. The 3D GPR depth slices in Figure 6 indicate the trend of the MBF and three underground pipes. In summary, NP and the GPR data indicate the location of the MBF, which is consistent with the geological data (Hauwert, 2009). The resistivity, NP, magnetic and conductivity data show cave-like anomalies. GPR, magnetic and conductivity data show location of subsurface pipes across the study area. Findings of geophysical surveys are given in Table 1. There is a patched asphalt area on Height Drive where the fault crosses, and the repair on the site may have been necessary because of the nearsurface deformation due to the karstic features. 3D GPR Survey Area Resistivity, NP, Magnetics and Conductivity Line UP Mount Bonnell Fault DOWN
Figure3: 3. Site showing the location geophysical and the Mount Bonnell fault. TheMount fault location is taken from Hauwert, Figure Sitemap map showing the of location of profiles geophysical profiles and the Bonnell fault. The fault N., 2009. location is taken from Hauwert, N., 2009.
51
Stone Wall
NW
Location of Mount Bonnell Fault
SE
nT
Magnetic Data
mS/m
Conductivity Data
Figure 4: Magnetic and conductivity data across the Mount Bonnell fault. A magnetic high, conductivity high and low anomalies are observed between the stations 270 and 300 feet. Another anomaly on both profiles, caused by a buried pipe, is shown between the stations at 370 and 410 feet. The location of Mount Bonnell is referenced based on the geological data (Hauwert, 2009). Stone Wall NW 250
271
292
313
334
355
Mount Bonnell Fault
376
397
SE 418
439 Ft Ohm-m
Caves with collapse materials
Typical NP fault anomaly
mV
NP cave anomalies
250
300
350
400
450 Ft
Figure 5. Resistivity imaging (above) and NP (below) data across the Mount Bonnell fault. Note the correlation between the two data sets except that the resistivity data does not indicate the fault. 52
Intersection of Hwy. 360 and Height Road Strike of Mount Bonnell Fault
N
Pipe
Pipe Pipe
Figure 6: 3-D GPR depth slices showing the strike of Mount Bonnell fault and subsurface pipes. Table 1: Karstic features located by geophysical surveys at Height Drive. Karstic Features Geophysical Methods Magnetic Conductivity Resistivity NP GPR TOTAL
Cave
Sinkhole
Collapse Materials
Faults and Fractures
Conduit
West Park Drive at Highway 360 The site map of the study area is given in Figure 7. The geophysical fieldwork took two stages: I performed reconnaissance magnetic, conductivity and GPR surveys along the orange line; second, I focused on the observed anomalies from the above-mentioned methods, and collected additional resistivity and NP surveys. Thus, the locations of the surveys are based on the site conditions with respect to geophysical methods. 53
The magnetic and conductivity data are shown in Figure 8. The magnetic values vary between 48,200 and 47,200 nT between the start and end of the profile. The magnetic data indicate two significant changes between the stations at 400 and 520 feet and the stations at 700 and 800 feet, respectively (see Figure 8). The approximate location of the MBF (Hauwert, 2009) is shown on the magnetic data in Figure 8. The conductivity values also show anomalies at these locations. A first vertical derivative of the magnetic data indicates shallow anomalies, which may be caused by cultural features or shallow small-scale faults associated with the MBF (Figure 9). The power-spectrum of the magnetic data indicates that the deepest magnetic sources are within the range of 60 to 80 feet deep. Figure 10 shows the resistivity imaging and NP data. Note that the resistivity data was taken between the stations at 250 and 547 feet whereas the NP data is collected between stations at 0 (zero) and 740 feet. The reason for this was that the resistivity profile was limited by driveways. Both data sets show locations of significant karstic features and faults. In addition, the resistivity data show a fault anomaly at the station 420 feet. A section of the GPR data showing a significant collapsed feature is given in Figure 11. The GPR data also shows a shallow fault at about 425 feet, which correlates well with the resistivity data. The GPR data do not indicate any anomaly over a significant resistivity anomaly, which was observed at station 490 feet at a depth of 18 feet and below. This is probably due to the depth exploration of the GPR data, which is about 7 feet. There are other GPR anomalies (small sinkholes, small caves and collapsed areas) along the profile; but I will examine these in another paper or presentation. In summary, the magnetic data showed several shallow faults and/or fracture zones, which correlate well with the resistivity and NP data. The resistivity data show two faults, which are located about 250 feet to the northwest of the MBF projected by the geological data (Hauwert, 2009). The conductivity data shows high anomalies the between stations at 200 and 450 feet, which is probably due to the shallow faults and fracture zones. The GPR data indicate a significant collapsed area, northwest side of which is controlled by a shallow fault. The location of this fault correlates well with the resistivity data. The NP data show fault and karstic anomalies along the entire length of the profile. There is also a significant gradient along the NP profile towards the southeast. This gradient is probably caused by the ground water movement within the conduits of the Edwards Aquifer. At the site, there is repaired, patched asphalt where the faults are located by the geophysical data. Findings of geophysical surveys are given in Table 2.
54
N
Magnetic, Conductivity and GPR profile NP profile UP DOWN
Resistivity profile Mount Bonnell Fault
Figure 7: Site map showing the location of geophysical profiles and the MBF based on the geological data (Hauwert, 2009).
55
NW
Center of Camp Craft Dr. Magnetic data
nT
MBF
SE
A B
Feet
mS/m
Conductivity data
Feet
Figure 8: Magnetic and conductivity data across the MBF. Location of the MBF is referenced based on the geological data (Hauwert, N., 2009). Note the two locations (A and B) on the magnetic profile where the slopes of the magnetic data change significantly, which correlates with the conductivity data. Center of Camp Craft Dr. SE
NW nT/Ft
Feet
Figure 9: The first vertical derivative of the magnetic anomaly indicates high-gradient zones (locations shown with blue arrows) which may be related to shallow cultural features or small-scale faults/fractures associated with the MBF.
56
NW
250
283
316
349
382
415
448
481
SE 547 Ft Ohm-m
514
Caves, voids, sinkholes and collapsed materials
mV
NP anomalies due to caves
NP anomalies due to a combination of fault and karstic features
NP anomalies due to a combination of fault and karstic features Ft
Figure 10: Resistivity imaging (above) and NP (below) data across the MBF at West Park Drive. Horizontal scales of both data sets are the same. Note the steep NP gradient towards the southeast. NW
SE 420
425
430
435
440
445
450
455
460 Feet
Collapsed Road Ft
Collapsed Wall Boundary
Collapsed Materials
Figure 11: Section of GPR data (400 MHz) across the MBF at West Park Drive. The data indicate a shallow fault and associated collapsed subsurface materials. The width of the collapsed area is about 40 feet.
57
Table 2: Karstic features located by geophysical surveys at West Park Drive. Karstic Features Geophysical Methods
Cave
Magnetic Conductivity Resistivity NP GPR TOTAL
Sinkhole
Collapse Materials
Faults and Fractures
Conduit
Bee Cave Road Site at Camp Craft Road A site map of the study area including the locations of geophysical profiles and the MBF are shown in Figure 12. The geophysical surveys, except the GPR, were all conducted along the grassy area between the two driveways of the West Lake Bible Church and West Lake Animal Hospital. The GPR data was collected along Bee Cave Road adjacent to other profiles (see Figure 12). There is an incipient sinkhole (recharge area) in the study area.
GPR Line Resistivity, NP, Magnetic and Conductivity Line
Up Down
4010 West Lake Bible Church
Mount Bonnell Fault
3930 West Lake Animal Hospital
Figure 12: Site map showing locations of geophysical profiles and the MBF. The magnetic and conductivity data are given in Figure 13 show high magnetic and conductivity anomalies between the stations 279 and 330 feet. The source of these anomalies appears to be subsurface. The incipient sinkhole is located at the station at 265 feet. The resistivity and NP data are given in Figure 14. The resistivity data show karstic anomalies (cave, sinkhole, collapsed materials, etc.) along entire length. The high magnetic and conductivity anomalies correlate well with the locations low
58
resistivity material (≤ 20 Ohm-m). Based on this correlation, the source of the magnetic and conductivity anomalies can be attributed to magnetic soils in the subsurface. The NP data shows a very unique “U” type anomaly along the profile. The NP values range between 10 and -38 mV. The NP anomaly appears to be caused by a conduit flow in the subsurface. Two sections of GPR are given in Figures 15 and 16. Both data sets indicate a sinkhole and a collapsed area, respectively. A change in elevation from high to low towards the southeast is observed on Bee Cave Road where the collapsed area starts. In summary, the magnetic and conductivity data show high amplitude anomalies which correlate well with the low resistivity materials between the surface and 25 feet below. The NP data display a simple but strong amplitude anomaly between the stations at 215 and 350 feet, indicating a conduit flow in the subsurface. The GPR data show a sinkhole and a collapsed area along the profile taken on the road. None of the geophysical methods appears to detect the MBF’s signature. This may be due to the dominating
NW
Magnetic Data
MBF
Observed Incipient Sinkhole
SE
nT
Feet mS/m
Conductivity Data
Feet
Figure 13: Magnetic and conductivity profile across the MBF. There is a ferrous electrical pole near the station at 250 fee, but observed high magnetic and conductivity anomalies between the stations at 260 and 330 appear to be caused by subsurface sources.
59
MBF
NW 160
181
202
223
244
265
Observed Incipient Sinkhole
286
307
SE 328
349
Ohm/m
C o l l a p s e d m a t e r i a l s, s i n k h o l e s a n d c a v e s
10.0 5.0 0.0
mV
-5.0 -10.0 -15.0 -20.0 -25.0 -30.0 -35.0 -40.0
160
200
250
300
350 Feet
Figure 14. Resistivity (above) and NP (below) data across the MBF. Location of the MBF is given by Hauwert, Nico., 2009.
60
NW
235
240
245
250
255
260 Ft
SE
Feet
SINKHOLE
Figure 15. A section of GPR data across the MBF. A sinkhole is located between stations 242 and 252 feet. Note that the sinkhole is very close to the surface at the Bee Cave Road. Location of this sinkhole correlates well with the incipient sinkhole and the “sphere-like” high resistivity anomaly that are observed on the surface and the resistivity data, respectively. NW
285
290
295
300
305
310
315 Feet
SE
Feet
C O L LA P S E D A R E A
Figure 16. A section of GPR data across the MBF. A collapsed area, limited by the blue lines, is observed between stations 298 and 317 feet., which correlates well with the low resistivity anomalies (see Figure 14).
61
Table 3: Karstic features located by geophysical surveys at Bee Cave Road. Karstic Features Geophysical Methods Magnetic Conductivity Resistivity NP GPR TOTAL
Cave
Sinkhole
Collapse Materials
Faults and Fractures
Conduit
Discussion of Results/Conclusions All geophysical data obtained from the three sites across the MBF indicate significant subsurface anomalies. These anomalies appear to be due to caves, voids, collapsed materials, sinkholes, underground pipes, shallow faults and fracture zones. It should be noted that the magnitude of the NP anomalies are much stronger at the West Park Drive and Bee Cave Road than the Height Drive site. This observation may be related to the amount of the conduit flow within the Edwards Aquifer. The GPR data taken along the roads indicate significant near-surface anomalies caused by collapsing soil, sinkholes and caves. It appears that these locations appear to be fixed periodically because of patched, repaired asphalt conditions observed on the roads. In conclusion, data acquired and used to evaluate the effectiveness of geophysical methods in detecting karstic features and faults/fractures in the Austin area allowed correlation of unique and consistent anomalies with a known fault. It is clear from this study that integrated geophysical methods can be used to map Balcones faults and their associated karstic features quickly and inexpensively. Results of this study show the benefit of including as many as geophysical methods (five in this study) to both improve fault and karstic characterization of near-surface geology.
References Collins, E.W., and Woodruff, C.M., 2001, Faults in the Austin, Texas, Area-Defining aspects of local structural grain, Austin Geological Society Guide Book 21, p.15-26. Connor, C.B., and Sandberg, S.K., 2001, Application of Integrated Geophysical Techniques to Characterize the Edwards Aquifer, Texas, STGS Bulletin, March issue, p. 11-25. Garner, L.E., Young, K.P., Rodda, P.U., Dawe, G.L., Rogers, M.A., 1976, Geologic map of the Austin area, Texas, in Garner, L.E., and Young, K.P., 1976, Environmental geology of the Austin area: an aid to urban planning: The University of Texas at Austin, Bureau of Economic Geology, scale 1:65,500. Fitterman, D.V., and Stewart, M.T., 1986, Transient electromagnetic sounding for groundwater, Geophysics, v. 51, p. 995-1005. Hauwert, N.M,. 2009, Groundwater flow and recharge within the Barton Springs Segments of the Edwards Aquifer, Southern Travis and Hay Counties, Doctor of Philosophy, The University of Texas at Austin.
62
Rodda, P.U., Garner, L.E., and Dawe, G.L., 1970, Austin West, Travis County, Texas: The University of Texas at Austin Bureau of Economic Geology, Geologic Quadrangle Map 38, p.11.
Acknowledgements I thank Dr. Nico Hauwert for showing the Mount Bonnell fault in the field and for numerous discussions, and a number of friends and colleagues who helped in the data acquisition. I thank Vsevolod Egorov for his contribution on the interpretation of the magnetic data. And last but not least, I am thankful to Esin Saribudak for her contribution to the graphical improvement and edition of the paper. This research project was funded by Environmental Geophysics Associates.
63
Groundpenetrating
radar
and
other
geophysical
investigations
at
Flint
Ridge
Cave,
Austin,
Texas 18
February,
2010 John
W.
Holt Research
Scientist Institute
for
Geophysics,
Jackson
School
of
Geosciences University
of
Texas
at
Austin
[email protected] Other
contributors:
K.
Chaudhary,
K.
Dlubac,
J.M.
Sharp,
Jr.,
M.
Al‐Johar,
T.
Swanson,
T.
Brothers,
J.
Greenbaum,
J.
Nowinski,
V.
Smith,
R.M.
Klee Overview Through
a
student
research
project
and
a
Hydrogeophysics
class
at
the
University
of
Texas
at
Austin,
we
have
evaluated
multiple
geophysical
methods
for
identifying
caves
in
the
recharge
zone
of
the
Edwards
Aquifer
of
Central
Texas,
with
an
emphasis
on
ground‐ penetrating
radar
(GPR).
For
the
research
project,
a
detailed
GPR
study
over
the
Balcony
Room
of
Flint
Ridge
Cave
was
conducted.
For
the
class
project,
two
sites
were
selected.
Site
1
consisted
of
a
known
cave
(Flint
Ridge
Cave)
while
at
Site
2
a
cave
was
inferred
from
surface
observations.
Electrical
resistivity
(ER)
was
used
to
conduct
line
surveys
along
two
subparallel
transects
crossing
over
the
known
cave
at
Site
1,
and
nearly
orthogonal
North‐ South
and
East‐West
transects
at
site
2.
One
such
location
at
each
site
was
chosen
for
collecting
GPR
data
in
a
dense
grid.
Gravity
data
were
collected
only
at
Site
1
over
the
known
cave
location.
The
results
from
the
dipole‐dipole
ER
surveys
indicate
areas
of
anomalously
high
resistivity,
which
were
interpreted
as
air‐Zilled
caves;
however,
the
resistivity
signal
at
the
known
cave
site
was
small,
presumably
due
to
highly
conductive
moist
cave
sediments
(principle
of
suppression).
The
GPR
data
from
Site
1
showed
a
low
velocity
zone
which
correlate
well
with
the
ER
data
and
hence
provide
conZidence
for
locating
a
cave.
The
GPR
data
also
showed
inZluence
from
conductivity
of
near
surface
soils
which
resulted
in
reduced
resolution
of
near‐surface
features.
The
gravity
data
showed
an
anomaly
consistent
with
the
location
of
the
known
cave
at
Site
1;
however,
the
signal
was
on
the
order
of
the
noise
level,
and
it
was
determined
that
a
gravimeter
with
microgal
accuracy
would
be
needed
to
precisely
locate
the
cave.
Our
conclusion
is
that
Zield
observations
coupled
with
the
ER
method
provides
best
reconnaissance
at
these
sites.
This
should
be
followed
by
GPR
with
3‐d
processing,
and
possibly
gravity
to
verify
and
reZine
the
cavity
position
and
volume.
Research Project (Kathrine Dlubac Senior Honors Thesis, May 2007) In
this
project,
we
acquired
a
grid
of
GPR
data
over
the
Balcony
Room
of
Flint
Ridge
Cave,
at
3
GPR
frequencies,
25
MHz,
50
MHz,
and
100
MHz.
Due
to
the
unknown
attenuation
and
scattering
characteristics
of
the
subsurface,
the
goal
was
to
evaluate
64
penetration
depth
and
resolution
for
the
three
frequencies.
We
concluded
that
a
dense
2‐ dimensional
grid
of
data
is
required
in
order
to
conZidently
identify
anything
in
the
subsurface,
and
that
100
MHz
data
provided
the
most
promising
information,
with
a
possible
identiZication
of
the
Balcony
Room
ceiling
and
Zloor.
A
3‐d
migration
algorithm
was
required
in
order
to
recover
information
from
the
data.
50
MHz
data
was
promising
for
large,
deep
fractures
but
further
veriZication
is
needed.
We
propose
to
conduct
an
interior
lidar
survey
of
the
Balcony
Room,
registered
to
the
surface,
so
that
we
can
accurately
compare
interior
volume
and
structure
to
the
data
set.
Below
are
Zigures
highlighting
the
process
and
basic
results.
65
25 MHz
50 MHz
100 MHz
Examples of data at the different frequencies. Location shown as black line in lower right figure. Note that depth of effective penetration decreases with higher frequency but resolution of reflectors increases.
66
Migration
Processing steps (left) and effect of migration algorithm on example data. Migration essentially moves energy to the correct location along the line by modeling the reflections in the medium with wave velocity determined separately.
67
(left) Perspective view of reflector surfaces created from identifying continuous radar reflector horizons across the entire data volume (purple and yellow horizons in top right figure). Upper reflector is approx. 7m depth from surface, lower is approx. 12 m depth. These may represent the top and bottom of the Balcony room.
68
/#&$,&"&I('H(&0">$J&K
