Geothermal exploration using hyperspectral and thermal remote sensing : inferring shallow hydrology of the Waiotapu Geothermal Field, New Zealand : a thesis presented in partial fulfilment of the requirements of the degree of Doctor of Philosophy in Earth Sciences at Massey University, (Manawatū Campus), Palmerston North, New Zealand
Loading...
Date
2023
DOI
Open Access Location
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Massey University
Rights
The Author
Abstract
Geothermal areas can exhibit a series of surface manifestations (e.g. mineral alteration and deposition, thermal anomalies, hot springs, and characteristic plant species) which can be directly detected with remote sensing techniques within the visible, near-infrared, and thermal ranges. These surface manifestations are, to some extent, a reflection of the subsurface activity. There is a wealth of techniques, including geological, geophysical, and geochemical methods, which can be used to explore and monitor geothermal areas; however, remote sensing techniques from airborne and spaceborne platforms can provide a cost-effective alternative. In many cases, geothermal areas are densely covered by vegetation which can further increase the time and cost of exploration. However, vegetation has the capability to reflect the environment it lives in. Here, we propose vegetation can be utilised as a proxy for subsurface geothermal activity using a combination of hyperspectral (VNIR/SWIR), thermal infrared, and LiDAR imagery with rock/soil and plant elemental concentration values. These techniques are used in geothermal areas but have rarely been employed to analyse plants growing in the area. At Waiotapu Geothermal Field, less than 10% of the surface is directly exposed, areas where the hyperspectral airborne successfully identified three main lithologies and alteration minerals; acid-sulphate alteration, “mixed” alteration, and silica-sinter deposition. While plants cover the remaining 90% of the surface, with kanuka shrub (kunzea ericoides var. microflora) as the dominant species in soils >40 °C. As such, kanuka was selected for our investigation and four geothermally relevant elements were chosen (Ag, As, Ba, and Sb). In areas with near-neutral high-chloride springs with a significant upflow (e.g. Champagne Pool), Ag, As, and Sb are enriched in rock/soil samples and are uptaken by kanuka plants. Whereas high Ba concentrations were found in plants living in peripheral areas where water mixing is taking place. The foliar element concentration zonation maps were successfully developed through classification using Random Forest and regression with Kernel Partial Least Squares. Employing ICPMS data and laboratory, airborne, and satellite hyperspectral (VNIR/SWIR) remote sensing data to create models to predict the foliar element concentrations. The results correspond well with the geology and thermal profile of Waiotapu Geothermal Field. Additionally, thermal anomalies selected from airborne TIR broadband imagery were studied using point pattern analysis such as randomness test-statistics, to map their preferred patterns and orientation, which appear to be controlled by subsurface permeability and water flow. This research opens new opportunities for geothermal exploration and monitoring through plants using hyperspectral imaging, which can overcome the limitations of geothermal exploration methods in densely vegetated areas.
Description
Chapter 4 is reused under a CC Attribution 4.0 International (CC BY 4.0) license. Chapter 5 is subject to a journal embargo. The thesis will be made available in May 2025.
Keywords
Geothermal resources, Hydrology, Remote sensing, Hyperspectral imaging, New Zealand, Waiotapu, Infrared imaging