Optimising Recovery of Vine Performance Following Irrigation During Extended Periods of Water Deficitnd ASTER remote sensing data

Abstract

The study involved investigating the use of ASTER and HyMap datasets from the Tick Hill area, Mount Isa region in Queensland, to make mineral maps and validate them, and subsequently to use the mineral maps to map and study the regolith-landforms. The processing techniques used on the multispectral and hyperspectral data were relative band depth (RBD) and Spectral Indices, and a sequence of masking procedures to minimize spectral overlap effects from other materials such as vegetation, atmospheric particles and minerals. The ASTER datasets allowed mapping of mineral groups such as Al-OH, Mg-OH and iron oxides rather than individual mineral maps. HyMap data, due to its better spectral resolution was able to map kaolinite, kaolinite crystallinity, iron oxides (hematite and goethite), white mica, Mg-OH + carbonate minerals, and to an extent silica and Al-smectite.The application of the mineral maps applied to map regolith-landforms allowed better characterization of regolith materials as compared to traditional band combination methods. Surface mineralogy could be linked to specific surface regolith materials such as kaolinite and iron oxides representing ferruginous materials (duricrusts, soils or mottled zones), abundance of well crystalline kaolinite equated to saprolite and mottled saprolite and high Mg-OH equated to slightly weathered (saprock) exposures of Pre-Mesozoic basement rocks. Spatial variation in mineralogy permitted interpreting changes in surface regolith and refining regolith-landform units as mapped from simple Red-Green-Blue band combinations. Kaolinite crystallinity maps were effective in highlighting in situ regolith from transported regolith, and allowing interpretation of the presence of deep weathering profiles being capped by ferruginous materials (duricrusts) or saprolite exposures on the hills, rises and erosional plain landforms of the regions.