A comparison of multispectral aerial and satellite imagery for mapping intertidal seaweed communities
Habitat‐forming seaweeds are vital components of marine ecosystems, supporting immense diversity and providing ecosystem services. Reports of major changes in the distribution and abundance of large brown seaweeds in the north‐east Atlantic are an increasing cause for concern, but a lack of consiste...
Main Authors: | , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
John Wiley and Sons
2018
|
Subjects: | |
Online Access: | https://discovery.ucl.ac.uk/id/eprint/10047832/1/BrodieEtAl_AQC_Accepted_20180204.pdf https://discovery.ucl.ac.uk/id/eprint/10047832/ |
Summary: | Habitat‐forming seaweeds are vital components of marine ecosystems, supporting immense diversity and providing ecosystem services. Reports of major changes in the distribution and abundance of large brown seaweeds in the north‐east Atlantic are an increasing cause for concern, but a lack of consistent monitoring over time is a key impediment in obtaining reliable evidence of change. There is an urgent need to recognize change rapidly and efficiently in marine communities, which are increasingly affected by pressures of human population growth, climate change, and ocean acidification. Here, the potential for remote monitoring of seaweed habitats is investigated using freely available, high‐resolution aerial and satellite imagery. Three sources of imagery were used: (i) Channel Coastal Observatory (CCO) aerial imagery; (ii) aerial images from the Bing webmap server; and (iii) RapidEye multispectral satellite data. The study area, the Thanet Coast, is an area of chalk outcrop in south‐east England of high conservation status, and includes three Marine Conservation Zones. Eight habitat classes, including brown, red, and green algal zones, were recognized based on ground‐truthing surveys. A multi‐class classification model was developed to predict habitat classes based on the chromatic signature derived from the aerial images. The model based on the high‐resolution CCO imagery gave the best outcome (with a kappa value of 0.89). Comparing predictions for images in 2001 and 2013 revealed habitat changes, but it is unclear as to what extent these are natural variability or real trends. This study demonstrates the potential value for long‐term monitoring with remote‐sensing data. Repeated, standardized coastal aerial imaging surveys, such as those performed by CCO, permit the rapid assessment and re‐assessment of habitat extent and change. This is of value to the conservation management of protected areas, particularly those defined by the presence or extent of specific habitats. |
---|