High-Resolution Satellites Detect Harmful Algae in Inland Waterbodies

Remote detection of harmful algae blooms (HAB) has become increasingly important due to their detrimental effects on surrounding living organisms. Using spectroscopic data gathered by three satellites, Professor Fiona Johnson’s research group at the Water Research Center, University of New South Wales, has analyzed HAB in small and medium inlet water bodies. They presented a pre-proof edition of their findings in the recent issue of Science of The Total Environment. Data from Landsat 8, Sentinel 2, and Planetscope satellites were used.

What are Harmful Algae Blooms?

Algae are primary plants that make up the foundation of food colonies. They range in size from microscopic to single-celled botanical elements to hefty seaweeds. Under unmonitored, conducive conditions, algae can grow out of control. Some of these webs of algae produce toxins hazardous to fish, animals, and birds that inhabit the same space. In some extreme cases, it can lead to human fatality. Regions where such algae have accumulated are referred to as harmful algae blooms.

Knowledge of the evolution of HAB in a particular region is critical to taking informed actions to manage the threats posed. The dynamics of HAB can change spatially with time, depending on its exposed conditions.

Testing HAB in Water

Remote sensing has emerged as the most suitable detection tool for HAB. Remote sensing relies on optical interrogation to map out the pollutants in a sample from a distance. By matching the spectroscopic details of the acquired data to known spectral bands of toxins, an image of HAB at particular locations can be analyzed.

The urgency of the climate crisis has initiated many collaborative government efforts to monitor earth's atmosphere and surface using satellites. Data recorded by satellites provide spectral information on many samples, including HAB on water bodies. Satellites have calibrated spectrometers onboard that image the spectral characteristics of various regions on earth as they orbit.

Depending on the wavelengths associated with HAB images, algorithms are developed to study the data. Semi-empirical algorithms and semi-analytical approaches have been widely applied to large water bodies such as lakes. Machine learning methods have also been implemented to analyze large water body images. 

Small and medium-sized water reservoirs still need in-depth investigations regarding HAB.

Detecting HAB using Satellite Data

Professor Johnson and her colleagues set out to find the feasibility of accurately detecting HAB in smaller water bodies by analyzing satellite imagery. These images are from three satellites, including Landsat 8, Sentinel 2, and Planetscope.

Publicly available spectral data were deliberately chosen so that the results would generally apply to all small or medium reservoirs.

The research group aimed to answer the following questions through its analysis of the satellite images:

• The measure of Sentinel 2, Landsat 8 and Planetscope satellites’ ability to detect HAB
• How the image resolution affected their ability to image HAB

In this study, small- to medium-sized water bodies had a surface area between 0.001 and 5 km². Six different-sized segments of Grahamstown Dam, a mesotrophic freshwater lake in Ferodale New South Wales, were used as representative water bodies for the project.

A machine learning approach and 20 existing algorithms helped analyze the spectral data obtained from Sentinel 2, Landsat 8, and Planetscope satellites.

Each satellite offered images at different spectral ranges and with different resolutions. Attention was paid to key features determining the relationship between wavelength ranges and the impact of resolution on detecting HAB.

Feasibility of HAB Detection

Two algorithms returned the most accurate results among the mathematical approaches applied to analyze the images. These were NIR minus Red (NmR) and Curvature Around Red (CAR), respectively.

The most decisive property for HAB sensing was the spectral ranges afforded by the satellites. Images from Landsat 8 and Sentinel 2 proved the best technology to detect HAB in small and medium reservoirs.

The detection results were found to be independent of the size of the waterbody. This is an important outcome suggesting that this methodology can be applied generally to any small or medium bodies of water to detect HAB.

Challenges and Outlook

Several limitations can be improved when using satellite data for HAB sensing. A key attribute is correcting the time lag between the time the satellite passes over the specific region and sampling calibrations.

Weather influences on imaging also have to be accounted for during analysis. Moreover, the algae dynamics and influences from peripheral contaminants require further scrutinization.

Future research can improve HAB detection in small- to medium-sized water bodies by developing an effective strategy for integrating satellite imaging with water monitoring initiatives.

Reference

S. Liu, W. Glamore, B. Tamburic, A. Morrow, F. Johnson. (2022) Remote sensing to detect harmful algal blooms in inland waterbodies, Science of The Total Environment, 158096, ISSN 0048-9697, https://www.sciencedirect.com/science/article/pii/S0048969722051956

Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.