PhD Project Proposal Bushfire and Natural Hazards CRC
Remote Sensing of Savanna Tree Structure and Biomass
A key focus of the ‘Northern Hub’ program of the BNH CRC based at CDU is to investigate, and where possible help develop, sustainable economic options for building the resilience of remote communities across the north. In many parts of the northern savannas, currently there are few mainstream employment options available (pastoral and mining industries, Defence, service industries), and even fewer that meet the aspirations of Indigenous people. Outside of town centres, Indigenous people make up the majority of the population in many northern regions. One of the emerging opportunities for Indigenous communities has been the rapid rise in community-based Ranger groups—focusing on land and sea management activities. While to date these programs have been largely publicly funded, there is a keen desire on the part of the community groups themselves to develop sustainable environmental service enterprises which can offer culturally appropriate employment opportunities, including on part-time and casual bases. Landscape-scale fire management offers one of those opportunities, especially through the provision of greenhouse gas emission offsets and carbon sequestration in living vegetation / biomass, in so-called ‘savanna burning’ projects. The research project outlined below aims to develop a robust approach for measuring carbon stocks in savanna vegetation which, in the future, would have direct application in savanna burning and related biodiversity assessment projects.
Fires in Australian savannas are common, cause severe damage to the environment and emit large amounts of greenhouse gases. It is widely recognised that biomass burning is a globally significant driver of carbon CO2 cycling and an important source of greenhouse gases. As above ground biomass (AGB) is approximately 48% carbon, there is a clear need for techniques to efficiently and reliably quantify 3D above ground biomass structure and biomass changes related to changes in fire regime (frequency, timing and intensity of fires). In recent years, there has been an increasing emphasis to use active sensors like Radar and airborne laser scanning (LiDAR) systems to estimate various 3D characteristics of vegetation structure such as crown biomass, bulk density and height. Laser scanning, combined with up-to-date advanced data processing methods, has the potential to overcome the disadvantages and weak points of passive sensors to deliver very precise and reliable full 3D biomass structure information, which can be used to estimate carbon storage.
This research will:
a) develop a methodology to estimate vegetation biomass using an existing airborne LiDAR dataset collected by the AusCover facility of the Terrestrial Ecosystem Research Network (TERN) in Litchfield National Park in 2013.
b) relate the LiDAR derived biomass estimates to different fire regimes.
c) utilise the LiDAR derived biomass estimates to develop a methodology to estimate vegetation biomass from low or medium resolution satellite imagery (e.g. MODIS, Landsat).
d) apply the methodology developed in c) to derive large area estimate of vegetation biomass in Australian savannas from low or medium or resolution satellite imagery and relate these to different fire regimes.
The PhD candidate undertaking this research, Grigorijs Goldbergs, has received BSc, Engineering and MSc degrees from Riga Technical University, Latvia. He has worked for government institutions and companies in Latvia and Denmark where he gained extensive experience in extraction of 3D information from imagery and LiDAR data. He provided evidence of this expertise by advising CDU staff on processing of above mentioned LiDAR and other datasets. This expertise has proven to be difficult to find in Australia.
Until recently he was employed by the University of Twenty, Netherlands as lecturer in photogrammetry and remote sensing. He therefore does not only have the expertise required to undertake this project. He has also shown the ability to transfer it to others.
|2018||Journal Article||Efficiency of Individual Tree Detection Approaches Based on Light-Weight and Low-Cost UAS Imagery in Australian Savannas. Remote Sensing 10, (2018).|
|2018||Journal Article||Hierarchical integration of individual tree and area-based approaches for savanna biomass uncertainty estimation from airborne LiDAR. Remote Sensing of Environment 205, 141-150 (2018).|
This PhD research aims to develop and assess methods, using stereo satelitte imagery and laser scanning data, to extract 3D tree biophysical structural parameters for the purposes of accurately estimating biomass/carbon stocks in NT mesic savannas.
This PhD research aims to develop and assess methods, using stereo satellite imagery and laser scanning data, to extract 3d tree structural parameters for the accurately estimating biomass / carbon stocks in NT mesic savannas.
The main goal of this study is to determine the optimal procedure for the estimation of above-ground biomass in north Australian mesic savannas by using LIDAR remote sensing based methods.