Imagery for Impact Mangroves on the move

Thirty years of Landsat data maps the nationwide movement and migration of Australia’s mangrove forests

Published:13 December 2019

Aerial view of deep green mangrove forest and curving creek

Using a newly developed satellite data classification technique, researchers from Geoscience Australia’s Digital Earth Australia (DEA) program have created a series of maps that show how mangrove locations have shifted year-to-year in response to storm events and a changing climate.

Australian mangroves are dispersed geographically across the nation’s long coastline — over 35,900 km, not counting islands. Until this study, mapping mangroves across Australia in a nationally consistent way had been confounded by the forests’ remoteness and inaccessibility.

The decades-long United States Geological Survey (USGS) Landsat data archive offered a solution. The new Landsat-based mangrove maps are nationally consistent and show the extent of Australian mangrove forests every year between 1987 and 2016. The maps are now informing reports like the Great Barrier Reef Outlook that describe changing mangrove ecosystems and entities like the Queensland Department of Environment and Science are using the maps to better manage mangroves.

Mangroves on the move

In 2016, a massive mangrove die-back occurred in northern Australia around the Gulf of Carpentaria where mangroves along 1,000 km of coast died.

Green and blue scientific illustrations of coastal landscape showing wiggling river features

In April 2006, Cyclone Monica tore ashore as a category 5 hurricane and caused a major mangrove decline in Junction Bay, but over the next 5 to 6 years, those mangroves were able to re-establish themselves, as this series of Landsat-based mangrove maps show

The cause of the dieback event was not immediately understood, but now it is clear that the combination of a years-long drought, high temperatures, and a sea-level decrease caused by a very strong El Niño, combined into a potent mangrove killer. Additionally, this research has shown that the die-off actually started in 2014, earlier than originally recognised.

“The mangrove dieback can be seen in the data,” explains the study lead author, Director of Digital Earth Australia Product Development and Landsat Science Team Member, Dr Leo Lymburner.

Surprisingly, despite the dire 2016 environmental death-trap for northern Australian mangroves, the Landsat-derived maps reveal that mangrove forest area nationally for that year was still larger than it was in 1992 when the maps show mangrove forest extent at its lowest point in the recent past. From that 1992 minimum, mangrove forests steadily grew until peaking in 2011 with an additional 600 sq km compared to 1992.

This new study puts the massive 2016 dieback it into a broader context. Many of the mangroves that were lost during that event had recently expanded into that region — during the 1992 minimum, many of them were not there. These mangrove victims were relatively ‘recent colonists’ that had followed the higher sea levels to new portions of the coastline.

The study also showed that the 2016 dieback did not just affect the Gulf of Carpentaria. The declines were echoed in Western Australia’s Cambridge Gulf and the Alligator Rivers Region of the Kakadu National Park in Australia’s Northern Territory.

The new Landsat-based maps provide insights such as these around the country.

Changes in mangrove extent are cyclical and heavily influenced by sea-level and cyclones, the study found. As the climate continues to change, the magnitude of sea-level change and extreme weather will grow, having serious implications for mangroves.

Muddy mangrove tidal creek with leafless mangroves

Making the maps possible

The 2008 public release of the entire Landsat archive by USGS has made massive, long-term change studies like this one possible. This study used every image collected over Australia by Landsats 5, 7, and 8 to create its yearly maps.

Mangroves were classified using a method made possible by the data-density. Vegetation seasonality along with canopy traits were used together with previous Landsat-derived surface water maps to identify mangrove forests with high accuracy.

“The long-term acquisition plan of the Landsat mission provides a unique and invaluable dataset for tracking multi-decadal changes in the density and distribution of mangroves at continental scales,” says Dr Lymburner. “This allows us to understand the long-term effects of sea level fluctuations and the multi-year damage and recovery cycles of mangroves affected by severe tropical cyclones.”

It’s not just Australian mangroves that can be better understood. This newly developed mangrove mapping approach could lead to improved mangrove mapping worldwide.

Anyone can view the mangrove data on the DEA Maps website, where a virtual mangrove flipbook is at your fingertips. Watch mangroves forests as they expand and contract in response to local environmental conditions.


Lymburner, Leo, Peter Bunting, Richard Lucas, Peter Scarth, Imam Alam, Claire Phillips, Catherine Ticehurst, and Alex Held. 2019.  “Mapping the multi-decadal mangrove dynamics of the Australian coastline.” Remote Sensing of Environment:111185. doi:

Rogers, Kerrylee, Paul I. Boon, Simon Branigan, Norman C. Duke, Colin D. Field, James A. Fitzsimons, Hugh Kirkman, Jock R. Mackenzie, and Neil Saintilan. 2016.  “The state of legislation and policy protecting Australia’s mangrove and salt marsh and their ecosystem services.” Marine Policy 72:139-155. doi:

Original source: Landsat Science. Republished with permission.
Author: Laura E.P. Rocchio, NASA Landsat Communication and Public Engagement