State-wide assessments of river condition have been undertaken in Victoria since 1999 using the Index of Stream Condition (ISC). The third ISC report provided a snapshot of river health for 29,000km of major rivers and streams in Victoria, using data collected over a six-year period from 2004 – 2010. For the first time the ISC used state of the art remote sensing technology to provide a more accurate and comprehensive assessment of river condition than ever before.
A number of recommendations were identified during the third assessment. This project seeks to address the two main research recommendations relating to the derivation of bank lines and tree canopies.
The first key recommendation was to examine automating the generation of bank lines for as much of the river network as possible. Thus, the first objective of this project is to develop workflows and associated software to automate the mapping of river banks (top and toe of bank) to achieve coverage for at least 90% of the Victorian River Network using the LiDAR and imagery data from the 2010 ISC River Health Assessment. If the project approach and outputs fall short of this objective a full justification will be provided within the final project report detailing the issues preventing the achievement of this objective.
The second key objective is to analyse and report on procedures to identify individual tree crowns within the LiDAR and imagery data from the 2010 River Health Assessment. If previous research is successful and sufficient time permits after completing Objective 1, the preferred algorithm may be tested within this project to extract individual tree canopies, with recommendations for the acquisition included in the final project report.
The developed methodology and algorithms will be used by the Victorian Government within future River Health Assessments. It is the intention of the Victorian Government River Health program to implement this technique in their 2018 fourth ISC assessment. Other agencies involved in assessing river health, such as the Murray-Darling Basin Authority may also potentially benefit from this research.
The automation of the bank line process will produce a more cost effective, timely and reliable product for the ISC assessment. If successful, the individual tree crown identification may enable modified tree structure metrics to be derived to enhance the assessment of river health.
Dr Nathan Quadros
General Manager - Business