Machine learning in the Australian critical zone

Abstract This chapter presents examples of integrative research from Australia’s terrestrial surface environment where the pedosphere, atmosphere, hydrosphere, and biosphere interact, the so-called ‘Critical Zone’. In Australia, for around 25 years, national environmental data layers available through Geographical Information Systems software have been combined with field-based measurements and observations at sparsely sampled points, using machine learning tools for pattern recognition, to produce spatially explicit predictive models for mapping soils and soil properties, species distributions, and other features in the Critical Zone. The availability of spatially extensive datasets representing different factors of landscape evolution and their exploration with machine learning and rule induction techniques allow the evaluation of emergent patterns against existing domain knowledge, which in turn can lead to new insights into Earth surface processes.