Hidden water in remote areas – using innovative exploration to uncover the past in the APY Lands
Reliable water availability is critical to sustaining community water supplies and determining economic development opportunities. In many cases, particularly in remote and arid areas such as in the Anangu Pitjantjatjara Yankunytjatjara (APY) Lands in the far northwest of South Australia, groundwater is the only viable source of water. However, there is limited knowledge of the groundwater resources in these remote regions and the Musgrave Province, where the APY Lands is located, is no exception. Consequently, there is a need to identify and determine the potential of groundwater resources in regions – such as the APY Lands – to supplement their community water supplies and to provide water for economic development which leads to employment opportunities.
The Goyder Institute for Water Research’s Facilitating Long-term Outback Water Solutions (G-FLOWS) suite of research projects has developed new techniques to interpret airborne electromagnetic (AEM) geophysical data, coupled with hydrogeological techniques, to identify groundwater resources buried by deep sedimentary cover which is a major constraint to identifying water sources in the northern parts of South Australia.
In its third stage (Stage 3), G-FLOWS is utilising new AEM data collected in 2016 under the South Australian government’s PACE initiative (co-funded by the Department for Environment and Water, DEW) to undertake a targeted program of data acquisition, interpretation and mapping of groundwater resources in the Musgrave Province. The research, a partnership between DEW, CSIRO, Flinders University and the Geological Survey of South Australia, is applying new and innovative geophysical and hydrogeological techniques developed in the previous two G-FLOWS projects (Stage 1 and Stage 2), combined with a variety of field evaluation techniques, to map the groundwater resources in the APY Lands.
Part of the G-FLOWS Stage 3 work program is to validate the AEM geophysical interpretation and investigate groundwater potential in the palaeovalley drainage systems imaged in the AEM dataset by establishing hydrogeological control sites as a part of a drilling program. These sites include a number of newly constructed water wells with the aim of reducing uncertainty in the interpretation of AEM data and thereby identifying potential deep groundwater resources in the palaeovalley systems.
This article focuses on the hydrogeological aspects of the drilling program, whereas the complementary article by Krapf et al. (2019, same MESA journal issue) focuses on the stratigraphy and evolution of one of the palaeovalley systems, in particular the Lindsay East Palaeovalley.
Study area and hydrogeological setting
The regional study area for G-FLOWS Stage 3 is centred on the Indigenous APY Lands located ~1,100 km northwest of Adelaide, in the far northwestern corner of South Australia. The study area covers ~26,600 km2 within the central region of the APY Lands and encompasses a number of communities and homelands including Amata, Pukatja (Ernabella), Yunyarinyi (Kenmore Park), Kaltjiti (Fregon), Mimili and the administrative centre of Umuwa (Fig. 1).
It is estimated that an Indigenous population between 2,000 and 2,500 currently live in the APY Lands (Australian Bureau of Statistics, ABS, 2016). According to the ABS (2017), the main industry and largest employer in the APY Lands is education and training, although with respect to income generation, the pastoral industry (i.e. cattle) is of particular importance. Retail trade, arts and recreation are also notable employers and generators of income. In particular, the arts industry is expanding at a rapid rate with galleries being established in Sydney and Adelaide.
The climate of the region is semi-arid to arid, with mean annual rainfall generally below 300 mm/year. The community of Pukatja reports an annual mean rainfall of 279 mm (Bureau of Meteorology, BOM, 2019), and a mean annual temperature of 27 °C (BOM 2018) with most of the rain occurring in the summer (December–January) months (BOM 2018). Vegetation cover comprises predominantly grassland, shrubland and open woodlands.
Topography in the study area varies considerably. The northern part is dominated by the Musgrave Ranges which host the highest elevation point in South Australia, Mount Woodroffe, at 1,435 m AHD (Australian Height Datum, i.e. sea level) which is located ~40 km west of Pukatja (Fig. 1). The Everard Ranges dominate the southern margins of the study area in close proximity to Mimili. Lying between the Musgrave and Everard ranges are extensive plains and rangelands (~550 m AHD), dominated by aeolian sand dunes and dunefields, sandplains and alluvial plains. A number of creeks drain the Musgrave Ranges to the south and through the centre of the study area. Officer Creek is the most prominent watercourse into which smaller tributaries such as Currie Creek and Ernabella Creek feed into (Fig. 1). For most of the year, these creeks are dry and only flow under episodic high rainfall conditions. Currently, flow in these creeks and watercourses is not being monitored.
The hydrogeology of the region is complex in terms of both the hydrostratigraphy and the groundwater flow systems. Data is limited as it is a remote area which is difficult to access (special permits and permissions are required to enter the APY Lands), and as such the understanding of hydrogeological processes are general and assumptive at best. Most of the hydrogeological information comes from basic investigations into water supplies for the communities, road building and special research projects such as G-FLOWS. The occurrence and distribution of water wells is quite sparse throughout the vast area of the APY Lands (Fig. 1) and monitoring and maintenance of wells in this remote area is challenging.
Broad-scale geology and hydrogeology of the region has received some recent attention through investigations focused on small-scale localised water supplies for road building (Pawley and Krapf 2016), which has also delivered new insight into the hydrostratigraphy. Prior to this G-FLOWS project three regional hydrostratigraphic units had been identified generally across the region, namely, an upper shallow unconfined sand aquifer consisting of colluvial and fluvial sand, silts and clay sediments followed by a variably weathered rock (saprolite) grading into a fresh fractured rock aquifer.
Groundwater level data, some acquired recently as a part of G-FLOWS Stage 3, helped to build the newly interpreted watertable surface (Fig. 2). Examination of the watertable surface contours indicate that groundwater flow direction generally follows topography. At a study area scale, contours indicate that groundwater is flowing in a southeasterly direction originating from the Amata area. However, directly north of Kaltjiti the flow direction is almost north–south and parallel to Ernabella Creek (Keppel et al. in press). In this location the upper sand aquifer is more continuous.
Groundwater levels vary from ~800 m AHD in fractured rock aquifers near Mount Woodroffe in the Musgrave Ranges, to ~320 m AHD in sedimentary aquifers south of the Everard Ranges. The topographic highs (Musgrave and Everard ranges) appear to be the largest influence on the watertable surface which closely follows the topography (Keppel et al. in press).
The Lindsay East Palaeovalley (Fig. 2) also appears to have an important influence on groundwater flow since watertable contours display a low regional hydraulic gradient between the two ranges, forming a groundwater divide separating the Lindsay East and Hamilton palaeovalleys (Hou et al. 2012).
Previous studies have suggested that the Musgrave Ranges and the headwaters of the drainage channels originating in the ranges, are important recharge areas (Leaney 2013). A number of flow reversals can be observed throughout the study region, for example, the Everard Ranges may also form a zone of recharge.
Faulting is widespread across the APY Lands and the region is known to be tectonically active with evidence of many small-scale seismic events (Pawley and Krapf 2016). The specific impact of faults on localised groundwater flow patterns within the study area (apart from the influences of secondary porosity and permeability), is difficult to evaluate given the limited groundwater data available. However, given the structural and tectonic history of the Musgrave Province, their influence is likely to be important (Keppel et al. in press; Krapf et al. 2019).
Unravelling the subsurface
Approximately two-thirds of the APY Lands was captured in an AEM survey comprising 17,395 line kilometres (Fig. 3) conducted in 2016 (Heath, Wilcox and Davies 2017). AEM is a non-invasive, fast and effective method, particularly in remote areas where ground access can be challenging, for assisting in mapping the location and geometry of aquifer systems including palaeovalleys, which constitute an important groundwater resource for local communities, industry and the environment. It presents the opportunity to gain information about the subsurface in an otherwise data poor area (Soerensen et al. in press).
The survey was flown with a line spacing of 2 km in a north–south direction and employed two time domain AEM systems. The western part of the area was flown with fixed wing aircraft employing the TEMPEST system, while the eastern part used the SkyTEM system flown by helicopter (Soerensen et al. in press).
The AEM data acquired across the area (Fig. 4) uncovered an extensive palaeovalley drainage system (Soerensen et al. in press; Costar et al. 2018). The conductivity depth sections inverted from AEM data revealed a complex, well-defined and relatively narrow set of palaeovalleys that contrast with those depicted in the contemporary landscape of today (Munday et al. 2013; Soerensen et al. in press). Although these palaeovalleys have been previously recognised (Rogers 1995; Magee 2009) and mapped (Bell et al. 2012; Hou et al. 2012), the AEM data enabled the mapping of the palaeovalley network in more detail and at higher spatial accuracy (Krapf et al. 2019) in comparison to previous interpretations (Fig. 5).
Figure 5 Palaeovalley distribution map across the G-FLOWS Stage 3 study area (after Bell et al. 2012).
Discovering hidden water reserves
In mid 2018, a drilling program led by DEW was undertaken primarily to validate the AEM geophysical model of Soerensen et al. (in press). The underlying objective – if indeed drilling revealed new groundwater sources – was to examine the potential of using the AEM data as a blueprint for water targets across the APY Lands.
The drilling program targeted the Lindsay East Palaeovalley located near the community of Kaltjiti in the eastern part of the APY Lands (Fig. 4). This site was chosen due to its close proximity to the Kaltjiti–Mimili road and the availability of groundwater information for the area due to its close proximity to the community and their water supply wells. The drilling program established a number of hydrogeological control sites to provide observations of the subsurface and install groundwater wells to monitor level (and sample), in order to gain insights into the behaviour of the resource and the potential of the Lindsay East Palaeovalley to provide further supplies (Fig. 6). Furthermore, drilling determined the thickness of the palaeovalley fill and the actual depth to basement as well as providing, for the first time, a detailed stratigraphic record of the palaeovalley fill due to the acquisition of two diamond drill cores (Krapf et al. 2019).
The drilling program employed three drilling techniques (Table 1):
- Shallow holes (<35 m deep) targeting the watertable used compressed air rotary drilling. This method enabled real-time water cut identification and sampling.
- Deeper holes (>35 m deep) employed rotary mud drilling. This technique is an ideal drilling method for unconsolidated and semi-consolidated palaeovalley formations, since the drilling mud stabilises the formation and maintains the integrity of the drillhole for well completion and/or running of downhole geophysical logs. A compressed air rotary technique was employed for the first two drillholes (wells DH1a and DH1d, SA Geodata unit nos 5344-87 and 5344-82). However, issues related to containing airlifted water and collapse of the hole-wall meant that drilling converted to a mud rotary methodology for the remainder of the deep palaeovalley drillholes.
- Finally, a triple tube HQ diamond coring technique was used to collect drill core samples.
Table 1 Nominal well specification types.
|Well type||Casing material||Casing diameter (mm)||Aquifer monitored||Screen type||Screen aperture (mm)||Screen diameter (mm)||Screen length (m)||Sump (m)|
Shallow well |
(<35 m deep)
|C12 PVC||155||Shallow sediments (watertable)||Machine slotted C12 PVC||0.5–0.7||155||3||1|
Deep well |
(>35 m deep)
|C12 PVC||177||Palaeovalley water-bearing zone sediments||Wire-wound stainless steel||0.5||141||3–6||1|
Site DH1 is located ~5 km southeast of Kaltjiti in close proximity to the Kaltjiti–Mimili road (Fig. 6). This drillsite is located in the centre of the Lindsay East Palaeovalley, which is orientated north–south and perpendicular (potentially) to the drainage path (Figs 4, 6). The 2016 AEM survey data was used to select the drillsite where the palaeovalley fill was expected to be at its thickest (AEM flight line 503401).
Once the diamond drill core was acquired through the entire palaeovalley sequence at site DH1, seven wells were drilled and constructed at this site (Fig. 6). Wells were designed to provide water samples to determine groundwater quality and age in a number of different units within the saturated palaeovalley fill and completed to allow aquifer tests which were conducted at a later date (March 2019).
Wells were developed (air and water jetted) and chlorine sterilised upon completion. Development was monitored by the DEW on-site hydrogeologist to ensure the wells were free of fine material and groundwater ran clear of sediment. Development ranged between 1–3 hours per well.
As mentioned earlier, the deep palaeovalley drilling used air circulation initially which was useful since it enabled real-time observations and assessments of the groundwater quality (salinity) and yield, and allowed a quick and accurate determination of the water-bearing zone at a depth of 55–65 m. However, it is a finding of this study and a key recommendation that future drilling in any palaeovalley in the APY Lands (particularly for holes >30 m deep), should use mud circulation drilling to deal with unconsolidated flowing sand intervals which present difficulties and challenges for well completion.
Drilling at site DH1 revealed a significant water-bearing zone (referred to as the target-water-bearing palaeovalley zone), based on preliminary on-site assessment of formation material, salinity and estimated yield. For the first time, the nature of the palaeovalley sediment fill and the water supply potential of these systems within the APY Lands was revealed. The drilling program resulted in the discovery of a target-water-bearing palaeovalley zone between 55–65 m below ground.
Hydrogeological control site
The hydrogeological control site DH1 is located ~5 km southeast from Kaltjiti centred on the Lindsay East Palaeovalley (Fig. 6). This was the first time that palaeovalley sediments had been used as water targets and drilled to investigate their potential as a suitable groundwater resource. The sediments within the Lindsay East Palaeovalley can be divided broadly into four major units based on their hydrogeological characteristics (see Fig. 8 for hydrostratigraphy and wells):
- Units 1a and 1b: Unconfined aquifer (well DH1e, unit no. 5344-83). Encompasses the dune and underlying sandplain system (~30 m thick) beneath which lies the hydraulically connected fluvial palaeovalley fill sand deposit with an estimated thickness of 35 m (Krapf et al. 2019). Groundwater is encountered at ~8 m below ground. Salinities in the top 30 m are ~1,000–1,500 mg/L and yields estimated to be <1 L/s. However, for the target-water-bearing palaeovalley zone (55–65 m deep), salinities are lower ~870 mg/L, with much higher yields of 10–18 L/s (wells DH1b, DH1c, DH1d; unit nos 5344-89, 5344-80, 5344-82). Hydraulic conductivities for this zone are estimated at ~5–25 m/day (Costar, Howles and Love in press). Hydraulic parameters were estimated by conducting step drawdown tests and a constant rate discharge test (12 hour continuous pumping).
- Unit 2: Confining bed. Consists of a 20 m thick sequence of silty clay (mud).
- Unit 3: Confined aquifer. Represents the basal palaeovalley fill sediments (wells DH1a, DH1a2, unit nos 5344-87, 5344-78; note DH1a has no screen interval and was replaced by DH1a2) consisting of sand but with a slightly higher salinity range (1,200 mg/L). Yields are <2 L/s which is much less than that of the target-water-bearing palaeovalley zone. Thickness is ~10–15 m, grading into a weathered basement sequence at the bottom (which overlies fractured rock and a consolidated fresh basement).
- Unit 4a: Weathered basement (well DH1f, unit no. 5344-85). Located ~700 m to the west of the centre of the palaeovalley, yields are extremely low (<1 L/s) with salinities ~1,000 mg/L. Fresh basement forms Unit 4b but has not been intersected in the drilling.
Characterising the groundwater
Groundwater characterisation is an important step in conceptualising the hydrogeological processes. While specific groundwater quality assessments are required before it can be used for any specific purpose, they are dependent on the desired use of the water source. Salinity is a useful and preliminary water quality indicator that can be used to determine the potential for groundwater use. The well yield is also an important factor when considering the significance of a groundwater source.
Table 2 provides a summary of the groundwater units and corresponding hydrogeological parameters.
Table 2 Summary of groundwater parameters for the palaeovalley sediments.
|Unit||Aquifer characteristics||Depth (m)||Salinity (mg/L)||Yield (L/s)||Hydraulic conductivity (m/day)||Depth to water (m)||Description|
|1b||Unconfined||55–65||870||10–18||5–25||7.6||Main water-bearing zone within the palaeovalley sediments (i.e. target-water-bearing palaeovalley zone)|
|2||Confining bed||65–85||na||na||na||na||Silty clay (mud)|
|3||Confined||85–108||1,200||<2||na||8||Basal palaeovalley sediments|
|4a||Weathered basement||108–117+||1,000||<1||na||8||Weathered rock (saprolite)|
|4b||Fresh basement||>117||na||na||na||na||Fresh rock with possible fractures|
na Not available
Encountering a groundwater resource with well yields of 10 L/s (at a minimum) and salinities of <1,000 mg/L is a significant find. This water is suitable for many purposes including community water supply and possible economic development such as stock watering and irrigation.
According to recent aquifer tests, the target-water-bearing palaeovalley zone is capable of yields up to 18 L/s, but a more conservative yield of 10 L/s ensures a long-term sustainable supply (Costar, Howles and Love in press). This rate equates to a volume of ~1 ML per day.
The AEM conductivity depth slice at 40–50 m below ground provides a spatial conductivity distribution across the entire survey footprint (Fig. 6). Conductivity depth slices of the AEM data at regular intervals is a typical output from processed AEM data; however, data can also be represented as a conductivity depth profile (Fig. 7) which can aid in defining the basic geometric architecture of the palaeovalley. Figure 8 illustrates an interpreted geometry of the palaeovalley fill sediments (blue) over the underlining fractured bedrock basement (grey) across site DH1 in the centre of the Lindsay East Palaeovalley (wells DH1a, 1a2, b, c, d and e) with wells DH1f and DH1g (unit no. 5344-86) located outside the palaeovalley extent. Figure 8 has also been annotated with relevant groundwater information from the recent drilling program. This is a small but crucial step in characterising the groundwater in this area and the first step in verifying AEM data.
From these early findings (Figs 6, 7, 8) it becomes evident how data acquired at the hydrogeological control site (DH1) can be used to verify the AEM data. While drilling provides real observations of the subsurface as a point data source, the validated AEM data can be very useful in upscaling point source groundwater and lithological data to the regional scale (i.e. the entire footprint of the AEM survey) and provide a useful tool for targeting water-bearing zones across the region where AEM data exists.
Conclusion and future directions
The discovery of a new fresh groundwater resource (<1,000 mg/L) in the APY Lands has enormous potential for the future development of this remote region in outback South Australia. Availability of a high yielding groundwater resource within the Lindsay East Palaeovalley could unlock the potential for economic development in the region. However, it is vital to follow up with additional hydrogeological investigations to determine the size and sustainability of this groundwater resource.
The methodology developed in this project included a combination of airborne geophysics (AEM), targeted drilling and hydrogeological investigations, which has proved to be highly successful in locating and characterising the palaeovalley groundwater resources. This methodology has the potential to be applied in other data-poor arid regions in South Australia as well as in other palaeovalley areas in Western Australia and the Northern Territory. It is highly recommended that this methodology and our toolbox approach is adopted in other palaeovalley areas to target water resources.
Future work is required to assess the full potential of these palaeovalley resources, which are highlighted here along with some of our key findings:
- AEM data proved to be a useful tool for locating a significant groundwater target. In combination with drilling, the work provides insights into the subsurface distribution and fill of palaeovalleys in the APY Lands including their groundwater potential. The methods developed around the acquisition and interpretation of AEM data can be used in other remote regions across Australia and the world.
- Discovering groundwater in the palaeovalley system and the significant water find provides a new potential resource. Historically, in the APY Lands, groundwater has been found primarily by drilling in close proximity to modern day creek drainage lines with yields obtained generally <2 L/s. While drilling was only conducted in one location on the Lindsay East Palaeovalley, it appears that this palaeovalley can potentially yield five times the amount of groundwater previously discovered in the APY Lands.
- Longer term aquifer testing (up to seven days pumping) would aid in the assessment of the long-term sustainability of the groundwater resource.
- Further drilling and investigations north and south of the hydrogeological control site (DH1) along the Lindsay East Palaeovalley would help determine how extensive the groundwater resource is.
- Drilling into other palaeovalley systems, such as the Lindsay West Palaeovalley (west of the G-FLOWS Stage 3 study area) would also help determine the feasibility of other palaeovalleys as potential water targets and verify characteristics of the wider palaeovalley drainage distribution.
- Upscaling of the AEM by using the findings from drilling at the hydrogeological control site will provide a useful update of the palaeovalley distribution map (last updated in 2012, WASANT Palaeovalley Map by Bell et al. 2012 and Palaeodrainage and Cenozoic coastal barriers of South Australia by Hou et al. 2012). It will redefine the palaeovalley extent laterally (provide more detail rather than a general smooth extent) and vertically (i.e. centre or thickest part) by showing the locality of the thalweg line. A similar finding was derived by Krapf et al. (2019).
- Drilling methodology for drilling deep (>30 m) palaeovalley wells should incorporate a rotary mud drilling methodology to control unconsolidated sediments.
Being able to uncover the palaeovalley drainage network has proven to locate a significant water target. Finding reliable water resources under cover in arid environments is challenging but by having a suitable water target, such as the palaeovalleys, ensures a greater probability of success.
We would like to acknowledge the traditional owners of the APY Lands, the Pitjantjatjara, Yankunytjatjara and Ngaanyatjarra people. In particular, we would like to thank Mr Witjiti George, Mr Maxi Stevens, Mr Robert Stevens, Mr Bruce, Mr Frank, Mr Lee and many others for undertaking on-country site inspections. We would also like to acknowledge the work undertaken by the APY Consultation, Land and Heritage Unit, including Ms Charmaine Jones, Ms Cecilia Tucker, Mr Noah Pleshet and Mr Andrew Cawthorn, who facilitated the necessary clearance approval to undertake this program of works within the APY Lands. Silver City Drilling is acknowledged for their drilling services to deliver this program as well as the staff of Regional Anangu Services Aboriginal Corporation (RASAC) in Umuwa for logistical support including accommodation. We also thank APY General Manager, Mr Richard King, and the entire APY Executive Board who were supportive of the G-FLOWS Stage 3 project. Finally we would like to thank DEW Principal Hydrogeologist Steve Barnett for his review of this article.
Bell JG, Kilgour PL, English PM, Woodgate MF, Lewis SJ and Wischusen JDH comps 2012. WASANT Palaeovalley Map - Distribution of palaeovalleys in arid and semi-arid WA-SA-NT, 1:4,500,000 scale, Geoscientific thematic map. Geoscience Australia, Canberra.
Costar A, Krapf C, Keppel M, Love A, Inverarity K, Munday T and Soerensen C 2018. G-FLOWS S3 – drilling for community in the APY Lands. In Reid AJ, Geological Survey of South Australia Discovery Day 2018: presentation abstracts and posters, Report Book 2018/00034. Department for Energy and Mining, South Australia, Adelaide.
Costar A, Howles S and Love A in press. G-FLOWS Stage 3, APY Lands Aquifer Testing, north-western South Australia, Technical Report Series. Goyder Institute for Water Research, Adelaide.
Hou B, Zang W, Fabris A, Keeling J, Stoian L, Michaelsen B and Fairclough M comps 2012. Palaeodrainage and Cenozoic coastal barriers of South Australia, 1:2,000,000 Series, DIGIMAP 00002. 2nd edn. Geological Survey of South Australia, Adelaide.
Keppel M, Costar A, Krapf C and Love A in press. G FLOWS Stage 3, APY Lands Drilling Program, north-western South Australia, Technical Report Series. Goyder Institute for Water Research, Adelaide.
Krapf C, Costar A, Stoian L, Keppel M, Gordon G, Inverarity L, Love A and Munday T 2019. A sniff of the ocean in the Miocene at the foothills of the Musgrave Ranges – unravelling the evolution of the Lindsay East Palaeovalley. MESA Journal 90:4–22. Department for Energy and Mining, Adelaide.
Leaney FW, Taylor AR, Jolly ID and Davies PJ 2013. Facilitating Long-Term Out‐Back Water Solutions (G‐FLOWS) Task 6: Groundwater recharge characteristics across key priority areas, Technical Report Series No. 13/6. Goyder Institute for Water Research, Adelaide.
Magee JW 2009. Palaeovalley groundwater resources in arid and semi-arid Australia – a literature review, Record 2009/03. Geoscience Australia, Canberra.
Munday T, Adbat T, Ley-Cooper Y and Gilfedder M 2013. Facilitating Long-Term Outback Water Solutions (G-FLOWS) Stage-1: Hydrogeological framework, Technical Report Series No. 13/12. Goyder Institute for Water Research, Adelaide.
Soerensen CC, Munday TJ, Ibrahimi T, Cahill K and Gilfedder M in press. Musgrave Province, South Australia: Processing and inversion of airborne electromagnetic (AEM) data: preliminary results, Technical Report Series. Goyder Institute for Water Research, Adelaide.
For more information, contact: