Project Title: Using Hyperspectral Remote Sensing Techniques to Determine Physiological, Phenological and Compositional Changes in Eastern Australian Grasslands
Supervisors: Professor Alfredo Huete (UTS)
PhD conferred: 2017
Australian temperate grasslands are among the most modified natural ecosystems on the continent, due to their position in agricultural and urban areas. These grasslands not only provide ecosystem services such as carbon sequestration, erosion control, and critical habitat, but native grasses are drought-tolerant and represent alternative pasture sources for grazing livestock. There has been recent interest in better understanding grassland ecology and physiology under a changing climate, as climate-based alterations to these ecosystems are likely impact both food security and conservation interests.
Remote sensing offers a cost-effective and accurate method to gather information on vegetation characteristics in inaccessible areas. Sensors mounted on satellites, aircraft, or at ground-level, can detect the light wavelengths that are reflected from the land surface, which in turn can be related to vegetation characteristics. Analysis of these spectral signals can provide information on biophysical variables such as leaf area index (LAI), fractional cover, biomass, and absorbed radiation; all factors that indicate plant growth, health and productivity. When taken over time, these measures can show seasonal patterns (phenology) that represents growth and senescence for most temperate species. These spectral fingerprints can also be used to differentiate one vegetation type from another if their spectral characteristics are sufficiently different.
Remote sensing of Australian grasslands has been used to investigate pasture growth rates, bushfire risk, and land cover mapping, however, the ability to identify grasslands of different quality and species composition remains a challenge. Hyperspectral remote sensing offers a great advantage over traditional methods, as it detects hundreds of discrete light bands across the visible and infrared spectrum. This greater spectral resolution can improve the sensitivity and discriminatory powers of remote sensing methods.
This project aims to determine the relationships between native grassland biophysical parameters and spectral vegetation indices using hyperspectral instrumentation. Specifically;
1. Differentiate temperate grassland communities using hyperspectral sensors, through spectral separation and time-series data, so that high-quality communities can be identified;
2. Establish a relationship between biophysical characteristics and vegetation indices for these ecological communities, so that grassland condition can be estimated; and
3. Identify potential climatic drivers that influence the structure and condition of temperate grasslands, and relate these to potential climate change impacts.