Ready for Summer – Remote Sensing of Bathing Water Quality

Ad:

It is summertime and bathing activities are amongst the most popular summer attractions in many inland lakes and coastal cities. However, effective control and monitoring of the quality of bathing waters is often flawed, especially regarding near real-time assessments. A variety of factors can significantly compromise the water quality, most of all by sewer influents, biological reasons such as bacteria and algal bloom, and organic as well as inorganic chemical compounds [1]. Traditionally, water quality estimation on recreational beaches is based on in-situ measurements at local scales as highlighted by the World Health Organisation (WHO) [2] and required by the US Beaches Environmental Assessment and Coastal Health (BEACH) Act, the revised European Water Framework Directive (EU 2006/7/EC) or the Australian Guidelines for Managing Risks in Recreational Water [3]. However, an in-situ sampling of parameters is often inaccurate and furthermore associated with high costs, often leading to a time delay until public information resulting from incubation requirements before laboratory analysis. As a result, beaches have often been closed too late or unnecessarily [4].

As an emerging field, remote sensing applications have been successfully trialled in the last years and near real-time satellite data have proved to derive current water quality conditions on local and regional scales. With a proceeding advancement in satellite sensor technology, a variety of water quality variables, including chlorophyll-a (Chl-a) and cyanobacteria indicators, total suspended solids and turbidity, Secchi disk depth, colour indices, and even enterococci levels and heavy metal pollution, can be monitored and assessed in long-term and near-daily views [1,3].

Figure 1: Spatial and temporal dynamics of rainfall, TSS and Tb near Borghetto stations between 22 and 25 July 2004 [4]

Figure 1: Spatial and temporal dynamics of rainfall, TSS and Tb near Borghetto stations between 22 and 25 July 2004 [4]

Particulate matter and water clarity

Events of strong precipitation cause not only overflowing of sewers but also soil spills, thus compromising bathing water quality. Remote sensing of turbidity (Tb) and total suspended solids (TSS) is widely applied in this context due to low cost and high spatial coverage. Earth observation data, frequently derived from MODIS Terra and Aqua satellites, has shown robust evidence for the relation of reflectance measures in a variety of electromagnetic spectra to the concentration of particulate matter as well as parameters of water column sediment. The scattering from suspended particles influences reflectance spectra, and Secchi disk depth (SDD), usually measured in-situ, has been found to be strongly correlated to Tb and TSS, thus allowing for reliable modelling of the effects of intense rainfalls on bathing water quality [4]. In this context, Bugnot and colleagues highlighted in a study from 2018 that water clarity is generally negatively correlated with red bands [5]. Moreover, recent studies highlighted the potential of remotely sensed data from the Terra and Aqua satellites for a near real-time update of water quality information [3]. Especially in the case of sewer overflows due to the associated bacterial contamination, high-temporal evaluation of Tb, TSS and SDD plays a crucial role to determine risk-free bathing options.

Cyanobacteria and algal blooms

Phytoplankton is an essential indicator of water ecology and quality and hence regularly monitored in-situ [6]. However, in correlation to global warming and shifts in water ecology, especially toxic cyanobacterial blooms, often referred to as cyanobacterial harmful algal blooms (CyanoHAB) are an increasing health threat to the majority of species – including humans [7]. Remote sensing thus gained increasing importance in monitoring and assessing phytoplankton as well as cyanobacteria distributions in freshwater lakes and coastal waters. However, there are several challenges associated with quantifying spatial patterns of cyanobacterial toxins, including the need for surrogate pigments – since cyanotoxins cannot be directly detected by remote sensing and “the variability in the relationship between the pigments and cyanotoxins – especially microcystins (MC)” [7]. Regarding microcystins, the primary challenge is associated with their inability to absorb visible or near-infrared (NIR) light as they are not pigments. This fact implies that a dual-model strategy for mapping is required, including the determination of a relation between microcystins and a reliably detectable surrogate pigment, and the establishment of the relationship between satellite data and this appropriate surrogate [7}. According to Stumpf and colleagues (2016), data derived from Medium Resolution Imaging Spectrometer (MERIS) has shown the most accurate outcomes regarding the required chlorophyll-a and phytocyanin quantification in this process. The inclusion of recent laboratory measurements of remote sensing reflectance for selected phytoplankton species from Poland [9] has the potential to be a significant aid in the calibration of input data into radiative transfer models. Furthermore, in a most recent study, researchers from Germany compared freshwater lake chlorophyll-a from five different sensors (MODIS-Terra, MODIS-Aqua, Landsat 8, Landsat 7 and Sentinel-2A) with in-situ data. They found that satellite chlorophyll-a maps allowed to follow the spatial distributions during bloom events and that harmful algal blooms were in high correlation with biomass development of cyanobacteria. However, they highlighted the varying performances in the comparison (Landsat 8: RMSE: 3.6 and 19.7 mg•m−3, Landsat 7: RMSE: 6.2 mg•m−3, Sentinel-2A: RMSE: 5.1 mg•m−3, and MODIS: RMSE: 12.8 mg•m−3) due to uncertainties of up to 48% for the in-situ derived data [6]. This is contrary to other compared results from MERIS imagery, which were processed using the National Oceanic and Atmospheric Administration’s (NOAA) satellite automated processing system (SAPS), with in-situ SDD measurements. In these results, quantification of CyanoHAB from different spatial areas in Florida, Ohio and California resulted in an 80% correspondence with in-situ samples. [8].

Figure 2: Main pigments occurring in cyanobacteria and diatoms with the location of main absorption peaks (based on Roy et al., 1989). The “+” sign indicates the presence of the chosen pigment [8]

Figure 2: Main pigments occurring in cyanobacteria and diatoms with the location of main absorption peaks (based on Roy et al., 1989). The “+” sign indicates the presence of the chosen pigment [8]

Contamination by pollutants and pathogens

The provision of detailed spatial information on water contamination by pollutants (e.g. pesticides and heavy metals) and pathogens (e.g. Enterococci, E. coli, viruses) is the most critical factor to ensure public health. However, these usually don’t affect remote sensing reflectance and thus can’t be detected by satellite. Nonetheless, a variety of studies has highlighted the correlation between those factors with factors such as Tb and TSS, as sediment-loaded waters often carry significant amounts of contaminants and pathogens due to surface runoff – especially from agricultural used sites. This fact has been widely accepted and some countries, such as Mozambique, use satellite-derived data of TSS and Chl-a as signals for potential outbreaks of cholera [10].  Regarding pollutants, handheld hyperspectral remote sensing applications have shown the ability to gather temporal and spatial information about heavy metal contamination of lead (Pb), nickel (Ni) and chromium (Cr) in bodies of water at low cost [11]. However, correlations between spaceborne earth observation applications and in-situ measurements are still weak. There is more work needed to develop robust optical proxies and indicators for relevant pathogens and pollutants to assess these by using satellite imagery [10].

Outlook

Successful management of bathing waters – coastal or inland – can only be derived by robust environmental assessment of the associated drivers. Combining modelling and analysis of temporal and spatial patterns under the use of historical and up-to-date remote sensing imagery is a promising option to ensure near real-time information and thus public health. The improvement of strategies and applications for the collection of pigment and toxin measurements is likely to improve the monitoring of toxins, pollutants and pathogens. The recent launch of Sentinel-3B will hopefully extend the data available and at the same time provide higher accuracy. The generation of earth observation satellite sensors currently under development is likely to improve this even further, especially in the context of monitoring the states of water bodies. Especially the European Space Agency’s (ESA) Felyx project, being developed by IFREMER, PML and Pelamis, is a highly promising future approach for the monitoring and forecasting of bathing water quality. Until then, we need to rely on current measures taken to ensure low-risk bathing pleasures at coastal as well as inland lake beaches.

Figure 3: Example satellite Earth observation data products of the Celtic Sea on the 10 September 2009 from MODIS at 1335 UTC and AVHRR at 0616 UTC. a) MODIS OC5 chlorophyll-a estimates, b) MODIS OC5 chlorophyll-a estimates with regions of dense blooms labelled, c) MODIS K. mikimotoi likelihood using Kurekin et al. (2014), d) MODIS K. mikimotoi likelihood classified into chlorophyll-a concentrations using Kurekin et al. (2014), e) MODIS algal anomalies using Shutler et al. (2012) and f) AVHRR sea surface temperature using Miller et al. (1997). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) [2]

Figure 3: Example satellite Earth observation data products of the Celtic Sea on the 10 September 2009 from MODIS at 1335 UTC and AVHRR at 0616 UTC. a) MODIS OC5 chlorophyll-a estimates, b) MODIS OC5 chlorophyll-a estimates with regions of dense blooms labelled, c) MODIS K. mikimotoi likelihood using Kurekin et al. (2014), d) MODIS K. mikimotoi likelihood classified into chlorophyll-a concentrations using Kurekin et al. (2014), e) MODIS algal anomalies using Shutler et al. (2012) and f) AVHRR sea surface temperature using Miller et al. (1997). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) [2]

References

[1] Usali, N., & Ismail, M. H. (2010). Use of remote sensing and GIS in monitoring water quality. Journal of Sustainable Development, 3(3), 228.

[2] Shutler, J. D., Warren, M. A., Miller, P. I., Barciela, R., Mahdon, R., Land, P. E., … & Roast, S. D. (2015). Operational monitoring and forecasting of bathing water quality through exploiting satellite Earth observation and models: the AlgaRisk demonstration service. Computers & Geosciences, 77, 87-96.

[3] Zhang, Z., Deng, Z., Rusch, K. A., & Walker, N. D. (2015). Modeling system for predicting enterococci levels at Holly Beach. Marine environmental research, 109, 140-147.

[4] Corbari, C., Lassini, F., & Mancini, M. (2016). Effect of intense short rainfall events on coastal water quality parameters from remote sensing data. Continental Shelf Research, 123, 18-28.

[5] Bugnot, A. B., Lyons, M. B., Scanes, P., Clark, G. F., Fyfe, S. K., Lewis, A., & Johnston, E. L. (2018). A novel framework for the use of remote sensing for monitoring catchments at continental scales. Journal of environmental management, 217, 939-950.

[6] Dörnhöfer, K., Klinger, P., Heege, T., & Oppelt, N. (2018). Multi-sensor satellite and in situ monitoring of phytoplankton development in a eutrophic-mesotrophic lake. Science of The Total Environment, 612, 1200-1214.

[7] Stumpf, R. P., Davis, T. W., Wynne, T. T., Graham, J. L., Loftin, K. A., Johengen, T. H., … & Burtner, A. (2016). Challenges for mapping cyanotoxin patterns from remote sensing of cyanobacteria. Harmful algae, 54, 160-173.

[8]Urquhart, E. A., Schaeffer, B. A., Stumpf, R. P., Loftin, K. A., & Werdell, P. J. (2017). A method for examining temporal changes in cyanobacterial harmful algal bloom spatial extent using satellite remote sensing. Harmful algae, 67, 144-152.

[9] Soja-Woźniak, M., Darecki, M., Wojtasiewicz, B., & Bradtke, K. (2017). Laboratory measurements of remote sensing reflectance of selected phytoplankton species from the Baltic Sea. Oceanologia.

[10] Zheng, G., & DiGiacomo, P. M. (2017). Uncertainties and applications of satellite-derived coastal water quality products. Progress in Oceanography, 159, 45-72.

[11] Rostom, N. G., Shalaby, A. A., Issa, Y. M., & Afifi, A. A. (2017). Evaluation of Mariut Lake water quality using Hyperspectral Remote Sensing and laboratory works. The Egyptian Journal of Remote Sensing and Space Science, 20, S39-S48.

See Also


Advertising



Like this article and want more?

Enter your email to receive the weekly GIS Lounge newsletter:

Advertising