GRSG Conference 2022: Orbit to Outcrop
Title: Looking into the effect of weather on geological spectral indices
Author: Harald van der Werff
Geologic remote sensing deals with phenomena that are dynamic in space and time. Sensors with a continuous multi-temporal operation (e.g. Landsat 8 OLI and Sentinel-2 MSI) enable an optimal seasonal acquisition and provide the data volume needed for continental-scale mapping. Lithological mapping is possible in areas with “good exposure” , which typically refers to arid and semi-arid areas. Surface cover that is typically observed in geologic remote sensing is often considered to be invariant (or only changing on a geological timescale), which holds for rocks and minerals and, to a lesser degree, for soils. What of course does change over time is the acquisition environment, driven by seasonal change and the weather.
For small-scale studies, data acquired at a single moment might suffice. Mapping at a continental scale however requires an image collection acquired over a longer time span, and possibly including temperate and cultivated regions. The criterion “in a dry period, and no clouds” is a common approach for data selection, but this is not standardized nor repeatable. An approach to minimize the influence of vegetation is to view the land “at its barest” , but the best view of surface mineralogy does not necessarily lead to the most robust signal. A question that remains is what environmental factors, besides illumination and the state of the atmosphere, influence spectral indices for supposedly stable surface covers most.
We looked at several “geological” spectral indices over time. The potential of the operational Sentinel-2 MSI for geological applications and mineral mapping, particularly iron oxides, has been studied extensively. Sentinel-2 MSI data were therefore chosen to calculate spectral indices (band ratios) for green vegetation, ferric & ferrous iron oxide mineralogy and hydroxyl bearing alteration (clay) mineralogy. The robustness of these indices is assessed over a semi-arid area in southern Spain and compared to changing weather conditions. These data were acquired over a 3-year period with alternating dry and wet periods that resulted in different levels of soil moisture content and vegetation cover.