AEGC 2019: From Data to Discovery – Perth, Australia 1

A statewide 3D geological model for New South Wales Edward Stolz Giovanni Spampinato John Davidson Geological Survey of NSW Geological Survey of NSW Geological Survey of NSW 516 High St, Maitland NSW 2320 516 High St, Maitland NSW 2320 516 High St, Maitland NSW 2320 ned.stolz@planning.nsw.gov.au giovanni.spampinato@planning.nsw.gov.au john.davidson@planning.nsw.gov.au

INTRODUCTION

The Geological Survey of New South Wales (GSNSW) is creating a statewide 3D geological model. The model is being developed in conjunction with the NSW Seamless Geology geodatabase (Colquhoun et al. 2018), which combines the best mapping across New South Wales (NSW) geological information and presents the geology with the present-day topography, as well as at major geological time interfaces. The statewide 3D model delineates major basins, orogens, deep-crustal faults and fault networks. These features are represented as interlocking sets of 3D surfaces and volumes located within a common reference frame. More detailed smaller scale models such as hydrocarbon, coal, mineral, geothermal and groundwater resource models will be embedded into the regional scale features of the statewide 3D model.

METHOD AND RESULTS The statewide 3D model comprises a series of interlocking province-scale models and a 3D statewide fault network within a common reference frame. Models at this scale will represent the current topography, basin and basement interfaces, major tectonic subdivisions, crustal-scale structures, major stratigraphic horizons and geological period boundaries. An example of a regional scale model for the Sydney–Gunnedah Basin is shown in Figure 1. Data Sources Base data for 3D modelling are sourced from the NSW Seamless Geology geodatabase (Colquhoun et al. 2018), seismic sections, drillholes (petroleum, minerals, coal and water bores) structural measurements and geophysical images

(gravity, magnetics and airborne electromagnetic). The models are dynamic and are updated whenever relevant new data are collected. The spatial distribution and resolution of data varies across the state from very dense to very sparse. Data comes primarily from GSNSW databases, with some data from Geoscience Australia (topography and geophysics) and NSW Land and Property Information (topography and satellite imagery). Full 3D models are usually based primarily upon a small number of key datasets; the NSW Seamless Geology, drillholes, seismic sections and interpreted geological cross-sections. An example of seismic and well data used to constrain the Eromanga–Surat Basin 3D model surfaces (Oliveira in prep.) is shown in Figure 2. Clean datasets often do not exist for modelling projects and considerable effort is needed to digitise, translate, clean, validate and interpret source data before modelling commences. Data extend across UTM projection zones and location information must be carefully verified before being integrated into a model. Modelling Process The 3D modelling process developed by GSNSW is described in Robinson (2016). The workflow recognises the scalar and interlocking nature of orogenic provinces and basins, as well as the structures and stratigraphy contained therein. It prioritises large-scale geological features and then works down in scale to infill models with increasing detail. All models are constructed using implicit modelling that relies on relatively sparse constraining data. The exact methods selected for each model depends on the type, quality and spatial distribution of the constraining data. There are three basic end-member modelling methods: (1) Compile a series of key datasets wherever they occur within the model, then use stratigraphic tables and surface mapping to link the datasets and create surfaces which span the areas with sparse or no data. (2) Work directly from seismic sections and drillholes to interpret horizons and structures, then link these together guided by surface mapping line-work. (3) Use the surface mapping line-work as the primary reference dataset, and constrain the dips of horizons and structures with information from structural measurements, drillholes, seismic sections and potential field modelling. A combination of these methods may be used in a single model depending on the density and distribution of data.

DISCUSSION It is important that the lithology and structure in the statewide 3D model match the surface geology shown in the NSW Seamless Geology geodatabase (Colquhoun et al. 2018). In areas where the Seamless Geology was compiled from limited information (i.e. poor outcrop) then the modelled 3D surfaces

SUMMARY The Geological Survey of New South Wales is building a statewide 3D geological model comprising a series of interlocking province-scale models and a 3D statewide fault network. The model is being developed in conjunction with the New South Wales Seamless Geology geodatabase and is constrained by mapped geology, interpreted geology, drillhole intersections, seismic sections and potential-field imagery and modelling. Detailed 3D models of mineral, hydrocarbon, coal and groundwater resources will be embedded within the framework provided by the statewide 3D model. The model will support land use management, mineral and energy exploration, scientific research, groundwater management, civil engineering and other applications. Key words: NSW, 3D model, statewide

A 3D statewide geological model for New South Wales Stolz, Spampinato and Davidson

AEGC 2019: From Data to Discovery – Perth, Australia 2

may improve the geological interpretation and justify modification of the Seamless Geology database. In some areas, the large scale of the model may create a mismatch in resolution between the 3D surface elements and the line segments in the more detailed NSW Seamless Geology. This is rectified by creating smaller model surface elements near the intersection with surface mapped geology. Geological cross-sections are used to link interpretations to seismic sections, establish structural and stratigraphic relationships, and to assess the interpretations from gravity and magnetic forward models. Figure 3 shows cross-sections used to constrain the location and geometry of faults in the Eastern Lachlan Orogen (Spampinato 2018). Seismic horizon picks are combined with Seamless Geology and drillhole data to interpret cross-sections. In the absence of seismic data, the geometry of horizons or faults is constrained using mapped relationships and magnetic and gravity images and modelling. Horizons can also be constrained by applying previously defined stratigraphic relationships. An initial statewide framework model of the basins and orogens of NSW will be finished by the end of 2019. After this, the project team will focus on resource scale models that will be embedded in the statewide model. Future modelling is planned to incorporate soil character and thickness, small-scale faulting, and other geotechnical data collected prior to and during major construction projects. These models will support informed decision making about the location and form of major infrastructure and reduce the risk in future civil engineering works. Over the next 10 years as part of its participation in the MinEX CRC, GSNSW will conduct a major program of drilling through deep post-Permian sediments into underlying basement in five areas of the state. This work will generate new geological data that will be incorporated into model updates and constrain ongoing re-interpretation of surfaces.

CONCLUSIONS The NSW statewide 3D model has applications for land use management, mineral and energy exploration, scientific research, water resource management, civil engineering and waste management. For the resources industry, benefits of the 3D modelling include an improved understanding of regional architecture and broad

scale relationships to resource distribution. An improved understanding of crustal structure, the distribution of key source units at depth and location of large-scale fluid pathways will assist in reducing exploration search space and recognising areas of potential not previously identified. Constraints on depth to target horizons is also a key outcome and will include depth of sedimentary cover overlying potential metalliferous resources, depth to aquifers for water resources and depth to target stratigraphy for energy resources. The statewide 3D geological model of NSW will provide a versatile framework for data integration and visualisation. It will support the development of new geological concepts and inform decisions about land use and resources.

ACKNOWLEDGEMENTS

The author publishes with the permission of the Executive Director, Geological Survey of NSW, Department of Planning & Environment.

REFERENCES Colquhoun G.P., Hughes K.S., Deyssing L., Ballard J.C., Phillips G.,Troedson A.L., Folkes C.B. & Fitzherbert J.A. 2018. New South Wales Seamless Geology dataset, version 1 [Digital Dataset], Geological Survey of New South Wales, NSW Department of Planning and Environment, Maitland. Oliveira F. and Davidson J. C. 2018, 3D Geological Model of the Sydney-Gunnedah Basin, NSW Australia, Geological Survey Report No: GS2018/0160. Oliveira F. in prep. 3D Geological Model of the Eromanga-Surat Basin, NSW – Australia, Geological Survey Report. Robinson J. 2016, Geological Survey of New South Wales 3D map development workflow, Geological Survey Report No: GS2016/0608. Robinson, J. 2017, Statewide post-Carboniferous cover thickness model of New South Wales, Geological Survey Report No: GS2017/0126. Spampinato G. 2018, 3D fault architecture of the Eastern Lachlan Orogen, New South Wales, Geological Survey Report No: GS2018/0576.

A 3D statewide geological model for New South Wales Stolz, Spampinato and Davidson

AEGC 2019: From Data to Discovery – Perth, Australia 3

Figure 1. Perspective view of the Sydney–Gunnedah Basin 3D model (Oliveira and Davidson 2018). The view is 900 km from south to north and has a 10x vertical exaggeration. Light blue colours show surfaces of the Permo-Triassic stratigraphy. The top of crystalline basement is shown in grey, while major faults are shown in dark blue.

Figure 2. Perspective view of the Eromanga–Surat Basin 3D model showing the wells and seismic sections used to constrain the modelled surfaces. The view has a 50x vertical exaggeration (after Oliveira in prep.).

A 3D statewide geological model for New South Wales Stolz, Spampinato and Davidson

AEGC 2019: From Data to Discovery – Perth, Australia 4

Figure 3. Perspective view of the 3D model of the fault architecture of the Eastern Lachlan Orogen, showing the statewide depth to basement surface (Robinson 2017) and geological cross-sections from 250K mapping used to constrain the modelled faults. The view is 1100 km from south to north and has a 4x vertical exaggeration (after Spampinato 2018).