1,2

(1)

10 20 30 40 50 21.4

21.6 21.8 22 22.2 22.4 22.6 22.8 23

Length(cm)

Elevation(m)

20 40 60 80 100

10 10.5 11 11.5 12

Length(cm)

Elevation(m)

Site 1

Site 2

Stalactites Length vs Elevation

Elevation (m)Elevation (m)

(a) Site 1 Cave ceiling

(c) Site 2 Cave ceiling

Kashif Mahmud

¹

, Gregoire Mariethoz

1 Connected Waters Initiative Research Centre, UNSW Australia, Sydney, NSW, Australia

Background

Lidar Investigation of Infiltration Water Heterogeneity in the Tamala Limestone,

EP41B-3514

Contact: [email protected]

www.civeng.unsw.edu.au/current-students/research/kashif_mahmud

1,2

, Pauline C. Treble

¹

Approach and Methodology

Reference: Treble, P. C., C. Bradley, A. Wood, A. Baker, C. N. Jex, I. J. Fairchild, M. K. Gagan, J. Cowley, and C. Azcurra (2013), An isotopic and modelling study of flow paths and storage in Quaternary calcarenite, SW Australia: implications for speleothem paleoclimate records, Quaternary Science Reviews, 64, 90-103.

Site description

Understanding matrix, fracture and conduit flow from 3-D information enhances our knowledge of the karst

aquifer. Dense 3-D point clouds are increasingly used as highly detailed input datasets. Caves make it possible

to enter the aquifer to directly capture 3-D point clouds to study a part of karst flow system. Analysis of caves using Lidar has been growing worldwide in recent years.

However, so far Lidar data has not been used to perform quantitative morphological analysis of karstic features.

There is a need to investigate the use of Lidar data, as well as the methods needed to process this kind of information.

By performing a morphological analysis of karstic features based on Lidar data, we try to understand groundwater

flow processes through saturated conduits, fractures and the matrix, and how they are expressed in the geological structures.

Our field site, Golgotha Cave (36.10°S 115.05°E, Fig 1A), is in aeolianites of Quaternary age: wind-blown calcareous sands that have deposited widely around the coast of SW Australia. The cave is 200m long and up to 25m wide, and the dune limestone is 20-30m thick over the monitoring

sites in the cave (Fig 1B). The long cave chamber was

originally formed by vadose zone waters and subsequent widened by ceiling collapse. Site 1 is located approx. 60m into the cave and sites 2 and 3 are located approx. 20m

further into the cave within a second large chamber which appears to be less stable than site 1, as evidenced by roof- collapse, dense rubble on the floors and small breakdown chambers in the walls and ceilings.

Objectives

1. Integrate Lidar data to build up a stalactite

morphological model based on stalactite size and shape.

2. Relate stalactite density variations with the topographic elevation of the cave ceiling, potentially indicating the

groundwater flow distribution governed by hydraulic gradient deviations.

3. Develop a relationship between stalactites length and diameter.

Fig 2: Different sites ceilings with stalactite distribution.

Only showing data from site 1-2 to illustrate method.

Ceiling sizes: site 1 = 19.0m by 6.4m, site 2 = 4.5m by 5.0m (Red specifies the higher ceiling elevation).

Fig 3: (a) Cave ceiling topography in 2D. (b) Moving averages with a window of 40x40 pixels grid. (c) Topographic anomaly maps. (d) Locations of stalactites. Isolate all pixels of topography anomalies with a deviation from the moving surfaces

above a threshold. Colour scales represent elevations in meter of the ceiling points (Red specifies the higher elevation).

1. For scanning the geological pattern we placed the lidar scanner at a few selected points at all sites.

, Andy Baker

3

Southwest Western Australia

3 Institute for Environmental Research, Australian Nuclear Science and Technology Organisation, Lucas Heights, NSW, Australia

2 Institute of Earth Surface Dynamics, University of Lausanne, Switzerland

Table 1: Stalactites properties from Lidar data analysis

Fig 4: Histogram plots of the topography anomalies

Distance(cm)

(b) Moving averages

200 400 600 200

400 600 800 1000 1200 1400 1600

1800 21.8

22 22.2 22.4 22.6 22.8 23 23.2

Distance(cm)

(c) Topography anomalies

200 400 600 200

400 600 800 1000 1200 1400 1600 1800

−0.1 0 0.1 0.2 0.3 0.4 0.5

Distance(cm)

(d) Locations of Stalactites

200 400 600 200

400 600 800 1000 1200 1400 1600 1800

Distance(cm)

100 200 300 400 500

50 100 150 200 250 300 350 400

450 −0.2

0 0.2 0.4 0.6 0.8 1 1.2

Distance(cm)

100 200 300 400 500

50 100 150 200 250 300 350 400 450

Site 1

Site 2

Distance(cm)

100 200 300 400 500

50 100 150 200 250 300 350 400

450 10

11 12

10.5 11.5 12.5 200 400 600

200 400 600 800 1000 1200 1400 1600

1800 21.4

21.6 21.8 22 22.2 22.4 22.6 22.8 23 23.2

100 200 300 400 500

100

200

300

400 10

10.5 11 11.5 12 12.5

(a) Cave ceiling

Distance(cm)

Fig 5: Stalactite length vs diameter plot with the application of kernel smoothing (Left). Colour scales represent density of points. Stalactite length vs ceiling elevation plots (Right).

Longer stalactites tend to occur at comparatively lower

ceiling elevations, which represent greater hydraulic gradients.

Fig 7: Aspect ratio vs stalactites lengths for all three sites Fig 1: (A) Southwest Western Australia (SWWA) map

showing coastal belt of dune calcarenite [Treble et al., 2013] (inset figure indicates SWWA region). (B) Plan view of Golgotha cave map showing all 3 Sites.

8. We differentiate various types of flow patterns using two stalactites properties i.e. aspect ratio and stalactite equivalent cross-sectional area (Fig 6).

9. Fig 7 shows linearity of stalactite clusters for both sites, as well as at an additional site 3, located 10m from site 2 near the opposite site of the same chamber. The analysis shows more linear clusters of stalactites are present in site 1

compared to other two sites, while site 2 mostly dominated by rounder groups of stalactites.

Site 1 Site 2

36.10°S 115.05°E

Site 3

Topography anomalies (m)

Counts

Site 1 Site 2

Topography anomalies (m)

Counts

−0.10 −0.05 0 0.05 0.1 0.15

0.5 1 1.5

2 2.5

3 3.5

4 x 10⁵

Threshold

−0.20 −0.1 0 0.1 0.2 0.3

1 2 3 4 5

6 x 10⁴

Threshold

2. The Lidar data has 1mm resolution, and we subsampled the resolution to a lower level of 1cm to reduce the number of data points, to minimize the data storage and

computational effort.

3. From the cloud of scan points we further crop the

relevant portion of cave ceiling comprising the stalactites for sites 1, 2 and 3. Fig 2 illustrates this in 3D with the

stalactite distribution.

4. Stalactites were identified in the Lidar data by first subtract- ing a smoothed surface (created with a moving average of

40x40 cm; Fig 3b) from the ceiling topography (Fig 3a) to identify topographic anomalies (Fig 3c). A 98 percentile

threshold (vertical red lines shown in Fig 4) was then applied to the histograms of these anomalies (Fig 4) to locate the

stalactites (Fig 3d).

5. Stalactite positions were validated with field photographs.

6. Statistical properties of stalactites were calculated (Table 1).

7. We then plot stalactite length vs diameter, which signifies the most common geometrical properties of stalactites in a particular site (Fig 5, left). Lastly, stalactite lengths are

plotted against ceiling elevation for all three sites in order to develop a relationship between stalactite length and the

hydraulic gradient (Fig 5, right).

Stalactites properties Site 1 Site 2 Site 3

Ceiling dimension 19 m by 6.4 m 4.5 m by 5.0 m 8.0 m by 6.9 m

Threshold (m) 0.0508 0.1365 0.1397

Total number of stalactites 2010 372 465

Density of stalactites per m² 17 17 8

Maximum diameter of one stalactite (cm) 34.4 26.6 51.9

Minimum diameter of one stalactite (cm) 1.1 1.1 1.1

Average length of stalactites (cm) 4.93 10.43 10.57

Maximum aspect ratio of stalactites 24.0 24.0 16.0

Minimum aspect ratio of stalactites 1.0 1.0 1.0

Average aspect ratio of stalactites 7.6 4.5 4.4

Fig 6: Flow patterns at different sites. (a) Definition of various flow type classification based on aspect ratio and cross-

sectional area. Locations of flow Type 1 (shown in column b) are based on actual stalactite positions shown in Figure 3(d) using 98 percentile of the topography anomalies histogram (Figure 4). And we have used 94 percentile of the topography anomalies histogram to locate possible flow areas (c) to define flow Type 2 (d) and flow Type 3 (e).

(c) Locations of possible flow areas

Distance(cm)

200 400 600

200

400

600

800

1000

1200

1400

1600

1800

(d) Locations of flow Type 2

Distance(cm)

200 400 600

200

400

600

800

1000

1200

1400

1600

1800

(b) Locations of flow Type 1

Distance(cm)

200 400 600

200

400

600

800

1000

1200

1400

1600

1800

(e) Locations of flow Type 3

Distance(cm)

200 400 600

200

400

600

800

1000

1200

1400

1600

1800

Distance(cm)

100 200 300 400 500

50 100 150 200 250 300 350 400 450

Distance(cm)

100 200 300 400 500

50 100 150 200 250 300 350 400 450

Distance(cm)

100 200 300 400 500

50 100 150 200 250 300 350 400 450

Site 1 Flow pattern

Site 2 Flow pattern

Distance(cm)

100 200 300 400 500

50 100 150 200 250 300 350 400 450

(a) Flow type classification

1 5

0 100 200 300 400 500 600 700

Type 2

Aspect ratio

Cross-section area

10

Threshold 94

Type 3

1 5

0 100 200 300 400 500 600 700

Type 1

Aspect ratio

Cross-section area

10

Threshold 98

Distance(cm)

Type 2 and 3

Type 1: Fracture plus matrix flow probably form soda straw stalactite.

Type 2: Pure fracture flow with a possibility to form a flowstone

or a curtain shaped stalactite.

Type 3: Combination of conduit, fracture and matrix flow with

round and large stalactite group.

Classified as Type 1

Work in progress

The physical and chemical properties of dripwater emerging from five of the stalactites have been monitored for the last 8 years as part of a cave monitoring program and more recently continuous drip rate measurements at 32 of these stalactites

for past two years to further analyze other statistical tests and future exploration with stalagmate data.

0 20 40 60 80 100 120 140

0 5 10 15 20

25 Aspect ratio vs stalactites length

Stalactites length (cm)

Aspect Ratio

Site 1 Site 2 Site 3

Higher aspect ratio indicates more linear features hence fractures

Compact clusters have longer stalactites, indicating more matrix flow