b'Education matters of the Cook Gap Schist. High-gradeFurthermore, the erosional history wasIt aims to provide prospectivity maps graphitic samples exhibited greaterextracted from the preferred LEM to:of lateritic Ni-Co by integrating various conductivity compared to low-gradeestimate the depth of emplacement forexploration data layers using machine graphitic samples but there is littleknown porphyry deposits; and re\x1dnelearning algorithms and designing a di\x1eerence in the resistivities of hosta prospectivity map explicitly estimatework\x1bow to process available public rock and mineralised lithologies duewhether probable deposits would be atexploration datasets in NSW and to signi\x1dcant clay, magnesite, andthe near-surface. turn them into features to determine silici\x1dcation of graphite. the probability of \x1dnding target Nathan Wake, The University of Sydney:mineralisation types in potential Based on these \x1dndings, a systematicMachine-learning in lateritic Ni-Coregions. In addition to the prospectivity work\x1bow for graphite exploration in thisprospectivity mapping by utilising publicmaps, the most important features in region has been established: geological and geophysical datasets in theexploring the targeted mineralisation Lachlan Orogen of NSW. type and its relationship with the mineral 1.Aeromagnetic interpretation to locatesystem model will be investigated. areas where the Cook Gap SchistThis interdisciplinary project will use occurs as a shallow horizon and tothe power of data science for mineral map structure. exploration and tackle the challenges 2.Airborne Electromagnetic (AEM)that the industry will face in the coming survey to identify potential graphiticdecades, particularly in the scope of lenses. critical minerals needed for the clean 3.Ground TEM survey to re\x1dne graphiteenergy transition.targets and establish deposit boundaries. The main technical challenge is the relatively low number of known lateritic Addison Tu, University of Sydney:Ni-Co mineral occurrences (positive Miocene to present landscape evolutiontraining samples) and choosing negative models and implications for porphyrysamples in barren regions. This study copper preservation. presents a machine learning-based framework for generating prospectivity maps of lateritic Ni-Co in the East-Central Lachlan Orogen that will address this problem. There are some limitations associated with available exploration datasets of the Lachlan Orogen, such as the irregularity of crucial basement outcrops, variations in the preservation and thickness of lateritic weathering pro\x1dles, and the presence of widespread Cenozoic sedimentary cover and complex regolith. This can impede The state government is aiming tounderstanding the geological setting and develop a Critical Minerals productionlandform evolution that leads to large-and downstream processing industrialscale mineralisation events. However, hub in the central-western region ofemploying geophysical data such as NSW due to the potentially rich mineralmagnetic, radiometrics, and spectral endowment of the Lachlan Orogen.remote sensing along with other data Nickel and cobalt have been identi\x1dedtypes can aid in imaging the basement as being particularly prospective insource rocks beneath sedimentary cover laterite deposits by relatively recentand highly potentially prospective and exploration initiatives in the eastern andwell-preserved lateritic weathering central sub provinces of the Lachlanpro\x1dles.Orogen. There is a knowledge gap in identifying and representing theMarc Young, Flinders University: potential for lateritic Ni-Co and thisExperimental evaluation of protocols for project focusses on machine learningthe separation of Younger Dryas magnetic to generate prospectivity maps for newmicrospherules.resources in the East-Central Lachlan Orogen. Machine learning utilisesSince 2007, a growing body of evidence Development of highly calibrateddata-driven algorithms and techniqueshas provided support for the notion Landscape Evolution Models (LEMs)that automate clustering, classi\x1dcationof a major cosmic impact event at the has allowed for the \x1drst-orderand prediction of data (Rad, 2018).onset of the Younger Dryas (YD), an simulation of the New Guinea marginSupervised machine learning will beabrupt cooling event that occurred response to external forcings. Resultsused in this project, which trains abetween ~12 800 and ~11 600 BP. Rather include constraints for uplift rates,model on known input and output datathan a large single-impact event akin denudation rates and palaeo-elevations.so that it can predict future outputs. to the Cretaceous-Paleogene impact, DECEMBER 2023 PREVIEW 34'