Aarhus Universitets segl

No. 112: Literature review of remote sensing technologies for coastal chlorophyll-a observations and vegetation coverage

Harvey ET, Krause-Jensen D, Stæhr PA, Groom GB & Hansen LB. 2018. Literature review of remote sensing technologies for coastal chlorophyll-a observations and vegetation coverage. Part of ReSTEK (Brug af Remote Sensing teknologier til opgørelse af klorofylkoncentrationer og vegetationsudbredelse i danske kystvande) and DCE Remote sensing in coastal area projects. Aarhus University, DCE – Danish Centre for Environment and Energy, 47 pp. - Technical Report from DCE - Danish Centre for Environment and Energy No. 112. http://dce2.au.dk/pub/TR112.pdf

Summary

This report reviews the possibilities of using data from different remote sens-ing (RS) techniques to supplement the conventional national NOVANA pro-gramme for monitoring of water quality (Chlorophyll-a; Chl-a) and submerged aquatic vegetation (SAV; seagrasses and macroalgae) in Danish coastal wa-ters. Strengths, weaknesses and knowledge gaps are discussed and recom-mendations on future steps for integration of RS techniques for monitoring in Danish coastal waters are provided.

The complex bio-optical characteristics of the Danish coastal waters demand RS sensors that can distinguish between Chl-a from the commonly high levels of coloured dissolved organic matter (CDOM) and total suspended solids (TSS), as well as from sediment reflection and SAV in the nearshore zone.
While most shallow Danish coastal waters are well mixed all year round, areas like the Kattegat, the Belt Sea and the deeper parts of e.g. Limfjorden experi-ence summer stratification, causing the Chl-a signal sensed by RS to only rep-resent the upper part of the water column. As EU’s Water and Marine Strategy Framework Directives only require summer data on the mixed layer, stratifi-cation is a minor problem. However, near shore bottom reflection is a major issue for most sensors aiming at determining Chl-a in the coastal zone. To overcome this, either high spectral resolution to discriminate this noise is needed, and/or RS data of high spatial resolution, which diminishes the edge effects in the shore zone.

Experiences with RS monitoring of Chl-a have in recent times mostly been based on the European Space Agency’s sensor MERIS and the two MODIS sensors from NASA, which were especially adapted to water. With the re-cently launched and now operational Sentinel 2 MSI and Sentinel 3 OLCI sen-sors from the European Commission, new possibilities are emerging as these sensors provide higher spatial and temporal resolution and better retrievals in the coastal zone.

The shallow littoral zone around Denmark comprises a complex mixture of seagrasses, macroalgae, benthic filter feeders and soft/hard bottom. To dis-criminate these benthic features from each other, the data must have a combi-nation of both high spectrally and spatial resolution. While there has been a fast development of RS techniques capable of mapping the distribution and abundance/biomass of seagrasses over the last two decades, no single tech-nology can be recommended to monitor the different parameters and their changes over time. Optimal monitoring of submerged aquatic vegetation re-quires integration of field observations and different RS techniques.

Vegetation mapping in Denmark has shown promising results using airplane-derived summer orthophotos. The orthophotos are available biennially, and are ordered by the Ministry for other purposes. The optimal use of these or-thophotos for nationwide mapping of eelgrass meadows requires some ad-justments regarding image processing. At the smaller scale, experience with Red-Green-Blue colour maps (RGB) obtained from drones also provide prom-ising results for mapping eelgrass meadows and floating macro algal mats and is a technique that potentially can supplement validation of larger areas. By contrast, the relatively coarse resolution of freely available satellite data poses larger challenges for mapping patchy and mixed vegetation, but can be used on a regional scale.

Test and validation of RS derived Chl-a and submerged aquatic vegetation data can be done using data from the existing NOVANA monitoring pro-gramme when there is a good match in time and space between ground truth and RS data. Unfortunately, with the current parameters included in the NO-VANA programme (2017-2021), it is not possible to validate CDOM and TSS. However, additional Chl-a data for areas less spatial and temporal covered are of high value for water body assessment. RS-derived and NOVANA-de-rived data on marine vegetation supplement each other nicely in terms of the monitoring parameters they deliver: The NOVANA programme delivers de-tailed information on the lower depth limit of eelgrass meadows, which is highly sensitive to changes in water clarity and quality and therefore a key indicator of quality status. RS-based assessment of area distribution provides an important supplement to the understanding of the many ecosystem functions and services, provided by eelgrass meadows and other marine forests.