20/20 Vision – A revolution in coastline visualisation and monitoring is here!
By Principal Coastal Engineer Matt Blacka
Coastal imaging kicked off in Australia just over 2 decades ago, and at the time was the most sophisticated technique available to remotely measure coastlines. Fixed-position cameras were mounted on high rise buildings to take still photographs of the beach for quantification of shoreline change. This technique was often synonymous with the Argus system, which was the coastal imaging platform adopted almost universally by researchers around the world.
2020 sees us amidst another revolution in remote coastline monitoring and visualisation. Some of the most significant advances in over two decades are now coming to the fore, and “coastal imaging” is beginning to look like a very different beast – one that is finally lifting its head out of the sand (pun intended ;-) Gone are the times of the (retrospectively) clunky research-focused systems of old. Gone are the times of delayed static images of our coast, and data locked into inaccessible files on a distant server. In 2020 we are now bringing more to the table.
Don’t get me wrong, it’s still about high quality data for our beaches, but these days as coastline managers we want, need and can have so much more. Our tools are more dynamic and systems more sophisticated; It’s all about satellites, smart phones, social media, crowd-sourced and machine learning. Dynamic visualisation of our beaches in real time and accessible data sets with which we can easily interact via dashboards and toolboxes - and all functioning via the cloud.
Over the past couple of years my colleagues and I have been plugging away on a whole bunch of interesting and clever projects, all part of WRL’s advanced coastal monitoring bundle. When I reflect and consider the collective progress that the group has made, it’s easy to appreciate that coastal imaging no longer looks anything like it did 20 years ago, or even 10 or 5 years ago. We’ve come a long way on this journey and worked with some great project partners along the way.
CoastSnap community beach monitoring was coined by Mitch Harley and Mike Kinsela with humble beginnings in a partnership between WRL, OEH and Northern Beaches Council. The simple system provides a platform for community members to snap images of a beach from a fixed position using their smartphones and share the images via social media. Mitch has gone on to develop a package of tools to analyse the crowd-sourced imagery, such that we can now quantify erosion and accretion of monitored beaches using nothing more than photos posted by members of our community on social media. In just a few years the CoastSnap network has now grown to over 25 photo monitoring points in Australia and 14 other countries around the world, with a new CoastSnap app currently in the pipeline.
The CoastSat toolbox was developed by Kilian Vos in 2019 and provides a streamlined package for mapping long-term sandy-beach shoreline positions using freely available satellite imagery. Satellite remote sensing provides a low-cost long-term shoreline data set where no in-situ field measurements are available. The CoastSat system enables users to extract shorelines from Landsat 5, Landsat 7, Landsat 8 and Sentinel-2 satellite imagery, providing analysis of long-term shoreline positional change and insights into beachface slope. All from your desk and completely for free!
Over the past 2 years Chris Drummond has been busy working on BeachStat, a system to automatically track beach users and shoreline positions using low-cost remote camera systems, and machine learning algorithms. This economical system uses fully stand-alone customised trail cameras deployed at remote locations to record beach activity, with imagery sent and stored via the cloud. We are also partnering with Spectur camera systems for applications where real-time viewing of the beach is required. The output from both systems is the same – BeachStat provides long term continuous data that allows us to quantify patterns of beach use and understand the value of beaches to our communities; data that is becoming increasingly important as we attempt to evaluate the site-specific costs and benefits of future management strategies for our beaches.
In 2019 we also launched our first wave overtopping risk forecast and monitoring system at Fairy Bower in Manly, with our project partners Northern Beaches Council. The system developed by Ian Coghlan and Dan Howe comprises a fixed position Spectur video camera for real-time and post event viewing of overtopping events via the cloud, coupled with an automated forecast tool for overtopping risk. The forecasts are made on the basis of ocean and weather conditions, including site specific modelling of wave transformation, runup and overtopping of the promenade. Hazardous conditions above a risk threshold trigger an automated warning email to Council and WRL staff ahead of time. This neat system provides a “heads-up” that enables staff to manage risks at the site through pedestrian control, and to monitor conditions remotely from their desk. The camera system also allows us to fine tune the forecasts and improve reliability of overtopping predictions.
Lastly – there have been significant advances to our mainstream Coastal Imaging program. Over the past 12 months our systems have come forward in leaps and bounds with a focus on making the systems more practical and integrated for operational beach management (as well as maintaining high-accuracy beach quantification as always). We are again working with our long-term partners the City of Gold Coast and Tweed Sand Bypassing to monitor the Gold Coast beaches over a stretch now spanning from Letitia Spit to South Stradbroke. Once complete, this camera network will have over 50 fixed cameras providing continuous coverage of an astounding 35 km of coastline, making it the largest Coastal Imaging network in the world. These days our cameras capture imagery of the beaches every 15 minutes, and we are now also collecting time-accelerated videos every half hour – all stored and moved around via the cloud. The combination of static and dynamic footage is proving a significant enhancement for visualising the beaches, while maintaining the ability for high-accuracy quantification of coastal processes.
The cameras are only one part of the development to our traditional Coastal Imaging program though - we are now using our system to automatically map the shoreline every hour of the day, to count beach visitors and map their locations, to pick surfers on the line-up and to measure their rides. We have also partnered with geospatial data specialists Geoplex to develop a world-class operational dashboard system that makes the imagery and long term data sets much more accessible, and our own image analysis system (Cyclops) developed by Dan Howe and others, will soon supersede our use of Argus after 20 years. These changes now allow the Coastal Imaging system to provide a truly integrated data set for management of the beaches.
It has been a transformative couple of years in the world of image-based coastal monitoring systems and Coastal Imaging in 2020 only slightly resembles Coastal Imaging of 2000. What’s pleasing to see is that Australia and my colleagues at WRL in particular, continue to innovate in this space, making significant contributions to the growth, development and integration of these systems into practical coastline monitoring solutions.