A reasearch and development project utilising drone LiDAR and Photogrammetry for coastal detection and monitoring
In 2019 Diodrone commenced a 5 year research and development project into the use of aerial survey for the purposes of coastal erosion, change detection and monitoring. Stanwell Park was chosen as the study site due to the relatively diverse range of coastal habitats existing within close proximity to one another and including ocean beach, aeolian sand dunes, two lagoons, coastal lagoons and steep slopes of littoral rainforest.
The research and development project received funding for a 5 year period and includes the use of drone LiDAR and photogrammetry data capture techniques.
To identify the most ideal reality capture tools and data processing techniques for coastal erosion detection and monitoring.
On a 6 monthly basis, and following significant weather events, Diodrone completes drone LiDAR and photogrammetry surveys over the beach front and lagoon areas.
Data is processed and analysed for precision and accuracy before undergoing a change detection analysis and reporting process involving assessment of LiDAR derived Digital Terrain Model (DTM), LiDAR derived Digital Surface Model (DSM), photogrammetry derived DSM, and photogrammetry derived point cloud.
Combined LiDAR and Photogrammetry survey over Stanwell Park Beach has proven to best exploit the benefits of each data capture technique while offsetting their relative limitations. LiDAR provides reliable survey of terrain where vegetation cover exists and photogrammetry provides a precise survey and high quality visualisation over sand and rocks.
RESULTS – LiDAR OR PHOTOGRAMMETRY?
LiDAR provides a more reliable point cloud of the terrain surface beneath vegetation than does photogrammetry, though even the most precise of LiDAR systems lack the precision that a well executed photogrammetry surface is able to achieve. For this reason, where precision and accuracy are of the highest importance, the best practice is to capture both LiDAR and photogrammetry data sets.
The two data sets enable quality assurance and quality control by comparing one to the other and analysing for gross differences—controlling for geolocation errors and erroneous anomalies that may otherwise be undetected from the ground control network alone.
- precise and accurate in areas of no vegetation and high contrast
- rich visualisation due to true to life colours
- economically achieves accuracies of <20mm RMSE
- accurate and reliable in areas of vegetation cover
- works in low contrast environments like uniform sand and snow surface covering
- economically achieve accuracies of <50mm RMSE
True 3D analysis of changes and point cloud outputs colourised by difference provide immediate and intuitive understanding of changes evident in the data and insight into the likely processes responsible.
The photogrammetry derived DSM and LiDAR derived DTM provide ideal 2.5D data sets for detecting and analysing variances in the z axis (elevation). Differences in surface elevation are colourised and mapped over the survey area to identify key areas of change. Volume changes to elemental zones of the beach (north, south, middle, berm, foredune and hind dune areas are analysed to development a sand movement model of the site and assess for net loss or gain across the survey area.
Areas of prominent difference area then assessed by cross section analysis across multiple epochs to gain insight into trends and seasonal fluctuations. The Stanwell Park Beach north lagoon outlet has displayed surface elevation changes of up to 2m during the study period.
Evident in the analysis above, significant changes due to aeolian processes on the frontal dune and coastal erosion following a significant east coast low weather events in early 2020.
A multi-temporal cross section through the front dune on Stanwell Park Beach. Losses of material on the windward side of the dune and approximately equivalent to gains of material on the leeward side of the dune are typical of aeolian transportation.