Mineral and compositional mapping using airborne hyperspectral and geophysical products, North Queensland

ABSTRACT SUBMISSION

20th International Geophysical Conference and Exhibition Adelaide, South Australia

22-25 February 2009 Presenter Dr Rob Hewson

CSIRO Exploration and Mining

26 Dick Perry Ave, Kensington, WA 6151, Australia

Phone: 08-64368689

Facsimile: 08-64368555

email: [email protected]

Second Author Dr Thomas Cudahy

CSIRO Exploration and Mining

26 Dick Perry Ave, Kensington, WA 6151, Australia

Phone: 08-64368689

Facsimile: 08-64368555

email: [email protected]

Third Author Mal Jones

Geological Survey of Queensland

Block A 80 Meiers Road Indooroopilly Qld 4068:

Phone: 07 3362 9348

email : [email protected]

Fourth Author Matilda Thomas

Geoscience Australia

GPO Box 378 Canberra ACT 2601 Australia

Phone: 02 6249 9216

Facsimile: 02 6249 9972

email: [email protected]

Fifth Author Dr Carlsten Laukamp

James Cook University

Townsville, QLD 4811, Australia

Phone: 07 4781 4942

Facsimile: 07 4781 6116

Email: [email protected]

Sixth Author Fitriani Agustin

Curtin University

Townsville, QLD 4811, Australia

Phone: 07 4781 4942

Facsimile: 07 4781 6116

Email: [email protected] Technical Area: Minerals – Hyperspectral techniques (oral) PRESENTER PROFILE : Rob Hewson is a Research Scientist at CSIRO – Exploration and Mining, having previously studied geophysics at Melbourne University, worked as a geologist/geophysicist in Shell for 7 years, and completed a MSc(Hons) in geophysics at Macquarie University. He undertook a PhD at University of New South Wales on geological remote sensing and has since worked for CSIRO as a remote sensing geoscientist since 1998. Email: [email protected]

MINERAL AND COMPOSITIONAL MAPPING USING AIRBORNE HYPERSPECTRAL AND GEOPHYSICAL PRODUCTS, NORTH

QUEENSLAND

Rob Hewson1*, Tom Cudahy1, Mal Jones2, Maltilda Thomas3, Carlsten Laukamp,4, Fitriani Agustin5

1 CSIRO Division of Exploration & Mining;

[email protected], [email protected]; 2 Geological Survey of Queensland,[email protected]

3 Geoscience Australia, [email protected] 4 James Cook University, [email protected],

5 Curtin University, [email protected]

Key Words: Hyperspectral, radiometrics, Mt Isa, mineralogy, soil mapping

INTRODUCTION

Large areas of prospective North and North-East Queensland have been surveyed by airborne hyperspectral sensor, HyMap, and airborne geophysics as part of the “Smart” exploration initiative by the Geological Survey of Queensland. In particular, 25000 km2 of hyperspectral mineral and compositional map products, at 4.5 m spatial resolution, have been generated and made available via the internet (http://www.em.csiro.au/NGMM/; http://www.dme.qld.gov.au/mines/hyperspectral.cfm) (Figure 1). In addition, more than 130 ASTER scenes were processed and merged to produce broad scale mapping of mineral groups (Figure 1) (Thomas et al, 2008). Province-scale, accurate maps of mineral abundances and minerals chemistries were generated for North Queensland as a result of a 2 year project starting in July 2006 which involved CSIRO Exploration and Mining (www.csiro.au/science/ps16a.html), the Geological Survey of Queensland (GSQ) (www.dme.qld.gov.au/mines/hyperspectral.cfm), Geoscience Australia (www.ga.gov.au), James Cook University, and Curtin University. This project also involved CSIRO Minerals Down Under (MDU - www.csiro.au/science/MineralsDownUnder.html), the Cooperative Research Centres for Predictive Mineral Discovery (pmd*CRC) Landscape Environment and Mineral Exploration (LEME) and HyVista Corporation (www.hyvista.com).

Figure 1 : Location of the Smart sponsored North Queensland hyperspectral survey areas (orange and yellow blocks) and the ASTER mosaic (brown boundary).

Airborne radiometric data acquired over the same North Queensland Mt Isa – Cloncurry areas as the hyperspectral surveys (Figure 2), had been acquired at flight line spacing of 200 metre. Such geophysical radiometric data provides a useful opportunity to compare the mineral mapping potential of both techniques, for a wide range of geological and vegetated environments. In this study, examples are described of soil mapping within the Tick Hill area, and geological / exploration mapping within the Mt Henry and Suicide Ridge prospects of North Queensland (Figure 2).

Figure 2 : Location of the Mt Isa hyperspectral surveys (black) draped over a Ternary image of the Mt Isa GSQ radiometric data. DATA ACQUISITION AND PROCESSING The HyMap® system is flown by HyVista Corporation Pty. Ltd. (http://www.hyvista.com/) on a fixed wing aircraft typically at an altitude of about 2.5 km. The sensor collects reflected solar radiation in 128 bands covering the 0.44- to 2.50-µm wavelength range, including the visible near infrared (VNIR) and shortwave infrared (SWIR) regions of the electromagnetic spectrum (Cocks et al., 1998). These particular North Queensland airborne HyMap surveys were flown in the spring of 2006 and 2007 and consisted of approximately 200 flight-lines, at 5 m pixel resolution, from the Mount Isa. Georgetown, Hodgkinson and Pajingo areas. The flight lines were collected in blocks of 12-20 for each target area including overlap for later mosaicing of image products. Combined, the total volume of HyMap data (radiance at sensor and ground reflectance) consisted of approximately 750 Gigabytes.

Several suites of software were developed by CSIRO to cross-calibrate and process the multi-scene reflectance data to pre-competitive geoscience products. In particular the generation of seamless mineral maps data required the extraction of various products (“indices”) related to the depth and shifts of diagnostic spectral absorption features associated with mineral content and changes to their possible chemical composition or crystallinity. Such mineral product indices

were generated from hyperspectral reflectance data for each flight line and then mosaiced into a single block image. The thresholds of the index value, ascribed to the mineral content or composition, were chosen by checking the spectral signatures of the resulting image anomalies for each index product. Masking was implemented of overlapping mineral/vegetation absorption features, as well as areas of water and deep shade. For each pixel mapped as a given mineral in the final product, a number of diagnostic spectral criteria had to be satisfied, including the determination of its zero abundance. In addition, 335 field samples were collected for the purpose of ground-truthing the accuracy of the HyMap products during four field campaigns in 2007/2008.

The airborne geophysics radiometric data acquired as part of GSQ’s Smart and previous projects, encompasses the Mt Isa Inlier and has been levelled between surveys and processed into grids of 50 metre cell size. It is available via Geoscience Australia (www.geoscience.gov.au/gadds) and GSQ (http://www.dme.qld.gov.au/mines/smartex.cfm).

COMPARISON STUDY RESULTS

Mt Henry The Mt Henry Mine and prospect,(19o 46’S, 139 o53’E) contains a series of Proterozoic metamorphic and sedimentary units of variable feldspathic and micaceous composition (Figure 3f). A significant radiometric response is apparent for the Potassium and Thorium channels and to a lesser extent, Uranium (Figures 3a-c). However the comparison of the combined Ternary radiometric product (Figure 3e) with the hyperspectral white mica (e.g. muscovite and illite) content product (Figure 3d) shows the absence of significant mica content within the PLrb micaceous feldspathic sandstone unit of the Surprise Creek Formation (Figure 3f). Although potassium is present within muscovite, greater concentration may occur within K-feldspars (e.g. orthocloase) (Figure 3). A consideration possibly relevant here is that at very high temperatures muscovite may also breakdown in the presence of quartz to produce K-feldspar (Gribble, 1988). It is also interesting to note that the Mt Henry Mine is present along this contact between the mica and radiometric anomalies (384000E, 7813000N, Figure 2e) of PLrb. This absence of an micaceous anomaly within areas of a elevated potassium radiometric response for the PLrb unit could be related to its increased K-feldspar content. However the issue of the increased Thorium and Uranium response for the PLrb unit remains to be understood. The feldspathic quartzite and clayey sandstone unit, PLra, shows neither a radiometric or mica response (Figures 3d and e). Further investigations are needed to interpret these mineral and geochemical anomalies (e.g. via XRD and XRF analyses) identified by these hyperspectral versus radiometric techniques. Overall however this comparison indicates that there appears an opportunity to further discriminate mineral and geochemical trends using both techniques, either related to host rock mineralogy or possible alteration. A detailed description of these and other hyperspectral products and processing for the Mt Henry Study Area is described in Cudahy et al. (2008a).

Figure 3 : Mt Henry Study Area a) Potassium radiometrics; b) Thorium; c) Uranium; d) hyperspectral white mica content product; e) Ternary radiometric product; f) 1:100,000 published geological mapping (Courtesy GSQ).

a) b)

c) d)

e) f)

Tick Hill

The Tick Hill area (21o35’S, 139o55’E) is located in the southern part of the Mount Isa Inlier (Figure 2) and experiences a dry tropical climate of less than 500 mm per annum falling. The topographic relief is subdued (< 200 m) and vegetation cover is limited (< 25%) comprising narrow belts of green vegetation along ephemeral drainage and more pervasive dry vegetation cover elsewhere. The geology comprises folded and metamorphosed Proterozoic mafic and felsic rocks with Phanerozoic cover sequences including sandstones and calcareous sediments. The Australian Soil Resource Information System (ARSIS) soil map of Australia (www.asris.csiro.au/index_ie.html) describes the soil regime over the Tick Hill region as comprising: highly weathered bedrock (> 50%); soil on bedrock (20 - 50%); and alluvial sediments (< 20%). The soil types of the Tick Hill Study Area have been broadly classified to include: Ferrosols (FE), Vertosols (VE), and Tenosols (TE), as described by ASRIS, as shown in Figure 4b within the Ternary radiometrics image. A false colour image of the area (e.g. visible-near infrared image, highlighting green vegetation as red) shows some of the variability due to vegetation cover, geology and landforms. The Ternary radiometrics broadly maps the boundary between Ferrosols, described as containing high free iron oxide, and the areas of Vertosols (clay soils rich in smectites) and Tenosols (soils of poorly developed horizons). It should be noted that the Vertosols were in part also likely to be defined with the aid of digital elevation data in areas of drainage and low relief. An examination of the radiometrics shows that the potassium (Figure 4d) indicates large areas of K-rich materials (e.g. muscovite, illite and feldspar) developed over the both the Proterozoic basement rocks and associated alluvium/colluvium while high levels of Thorium (Figure 4e) are apparent within the Tertiary laterite and associated rubble (e.g. ironstone lags) (Figure 4c). The Uranium imagery appears to highlight distinct north-south anomalies within the Cambrian outcropping sandstones in the south-eastern part of the Study Area (Figure 4f). In summary it appears that the radiometrics data highlights both soil and outcrop related units. The white mica and Al smectite clay hyperspectral products also map a variety of clay rich alluvium and mica rich outcrops (Figures 4g and h). The Al smectite content however appears to indicate discrepancies with the boundaries of the broadly classified smectite bearing Vertosol. The ferric iron hyperspectral product (Figure 4i) approximately coincides with areas of elevated Thorium and Tertiary laterites shown in the geology (Figure 4c). The association of thorium with ferric rich laterites and ironstones has been previously noted in hyperspectral - radiometric comparisons at Woodie Woodie (Hewson et al., 2005). Cudahy et al. (2008b) outlines a detailed study of these hyperspectral derived soil mineral products at Tick Hill, describing its ability to discriminate Al-clays and also provide indirect indicators of soil pH, Cation Exchange Capacity and permeability.

a) b) c)

d) e) f)

Figure 4 : Tick Hill Study Area a) False colour VNIR HyMap image; b) Ternary radiometric image with ASRIS soil classification; c) published geological map (Courtesy GSQ); d) Potassium radiometrics; e) Thorium; f) Uranium; g) hyperspectral data derived white mica content; h) Al smectite content; and i) ferric iron content. Suicide Ridge The Suicide Ridge area, 21oS 140o 40’E, located approximately 30 kilometres south of Cloncurry and is dominated by the NNW-SSE striking Soldiers Cap Group, composed of metasedimentary rocks, interlayered amphibolites and the Saxby Granite. Substantial areas are also covered by Mesozoic sandstones of the Gilbert River Formation. The radiometrics acquired within this area is dominated by the potassium channel (Figures 5a and b) with some NNW trending Th and U anomalies apparent (Figures 5e and d). The discrimination of several of the major rock units within the one image was achieved by collating three different mineral

abundance maps generated by the HyMap imagery; "white mica abundance", “Fe2+

(e.g. ferrous) associated with MgOH", "Kaolin abundance", shown as a Red-Green-Blue composite image (Figure 5f). In particular, the meta-psammites of the Llewellyn Creek Formation and Mount Norna Quartzites are shown in deep red colours, due to their high white mica content. Ferrous rich amphibolites, highlighted by bright green, are apparent within the dolerite dykes of the Llewellyn Creek Formation and the strata discordant Suicide Ridge Breccia Pipe (Figure 5f). Although there is an order of magnitude difference in cell/pixel spatial resolution between the radiometrics (50 m) and the hyperspectral data (5 m), there are only diffuse indications of the dykes and breccia by the Uranium radiometrics (Figure 5d). The Potassium radiometric data (Figure 5b) appeared to highlight both the Llewellyn and Gilbert River Formations but which are clearly discriminated by the kaolinite (blue) and white mica (red) hyperspectral products (Figure 5f). The dramatic improvement in geological discrimination using the hyperspectral data, incorporating VNIR-SWIR wavelengths, compared to visible (“aerophotograhic”) imagery can also be seen in the comparison of Figure 5e with Figure 5f.

Figure 5 : Suicide Ridge Study Area a) Ternary radiometric image; b) Potassium radiometrics; c) Thorium; d) Uranium; e) False colour VNIR HyMap image; f) integrated Red-Green-Blue composite of hyperspectral data derived white mica (red), ferrous iron rich MgOH minerals (green) and kaolinite (blue) contents.

f) e)

e) f) d) c)

b) a)

SUMMARY

This study comparing radiometric and hyperspectral data acquired from the recent GSQ Smart Programme over the Mt Isa area, reveals the improved capability in the extraction of geological and soil information from their combined use rather than in their isolated application. Integration and detailed comparisons between the mineralogical and geochemical products from these two techniques offers the opportunity to refine and update traditional geological and soil maps leading to possible insights to areas of geological alteration or soil degradation.

REFERENCES

Cocks, T., Jenssen, R., Stewart, W. I., and Shields, T. 1998, The HyMap airborne hyperspectral

sensor: The system, calibration, and performance, In Proc. of 1st EARSEL Workshop on Imaging Spectroscopy, Zurich, October 1998, 7pp.

Cudahy T.J., Caccetta, M., Cornelius, A., Hewson, R.D., Wells, M., Skwarnecki, M., Halley, S., Hausknecht, P., Mason, P. and Quigley, M.A., 2005, Regolith geology and alteration mineral maps from new generation airborne and satellite remote sensing technologies and Explanatory Notes for the Kalgoorlie-Kanowna 1:100,000 scale map sheet, remote sensing mineral maps. MERIWA Report No. 252, 114 pages. Cudahya, T., Mal Jones, M., Thomas M., Laukamp ., C, Hewson, R., Caccetta, M., Rodger, A., Verrall, M., and Agustin, F., 2008, Towards global hyperspectral mapping of surface mineralogy: Fundamental information for understanding Earth’s soils and geology, 25th June 2008, Canberra, In Proc. of MINEX – CRC LEME Symposium, Mineral Exploration through cover. Cudahyb, T., Jones, M., Thomas, M., Cocks, P., Agustin, F., Caccetta, M., Hewson, R., Verrall, M., and Rodger, A., 2008, Soil mineral mapping using airborne hyperspectral data, Tick Hill, Mt isa region, Queensland, Proc. 1st Global Workshop on High Resolution Digital Soil Sensing and Mapping - IUSS Digital Soil Mapping Working Group, Sydney, February 5-8. Gribble, C.D., 1988, Rutley’s elements of mineralogy, 27th edition, Unwin HyMan, Boston. Hewson, R.D., Cudahy, T.J., Drake-Brockman, J., Meyers., J., and Hashemi, A., 2006. Mapping geology associated with manganese mineralization using spectral sensing techniques at Woodie Woodie, East Pilbara. Exploration Geophysics, 37, 389-400. Thomas, M., Laukamp, C., Cudahy, T.J., and Jones, M., 2008, Exploration advances: New developments in spectral remote sensing in the Mount Isa region, Australia. Proceedings of the 33rd International Geological Congress, Oslo, Norway.