Photo: Martina Bauchrowitz
"We are really lucky that we have the chance to participate in this year's LakeLab experiment," says Carmen Cillero Castro and adds: "Here we can examine 25 water bodies with different water qualities in a confined space and may furthermore access all data collected by other teams during the experiment." With the 25 water bodies, the Spanish remote sensing specialist means Lake Stechlin plus the enclosures of the LakeLab, literally 24 lakes in the lake. She and her colleague, Federico Cheda, characterize the various water bodies from above with a drone (see photo above) which is particularly advantageous when clouds prevent data acquisition from satellites. Their drone is loaded with 3 different multispectral and hyperspectral sensors and enables them, for example, to detect and quantify harmful algal blooms in drinking water reservoirs at an early stage.
Based on the water colour, remote sensing methodologies are used to classify the quality of water bodies. Pure water appears blue to us because blue light is absorbed weaker by the water molecules than red light. The water colour, however, changes depending on the amount and type of dissolved and particulate substances and organisms present in the water. In particular, it is influenced by phytoplankton, i.e. cyanobacteria and algae. These organisms contain pigments required for photosynthesis, mainly chlorophyll a, which makes them appear green.
Light sensors designed to accurately resolve the components of light, so-called spectrometers are aboard various satellites and allow recording the signal of these pigments in lakes. By corroborating these measurements with bio-physical models, the amount of algal biomass present in the lake can be estimated. "However, to what depth the sensors 'see' the pigments is variable," says Australian scientist Janet Anstee. Are they able to measure down to the chlorophyll maximum, i.e. the maximum of the algal biomass? "This is why it is important to parameterise the models with a set of realistic values which have been measured directly at the relevant water depths. Only then it is possible to make reliable assessments of water quality in lakes by means of remote sensing, without ever having sampled the lake itself, simply by interpreting the colour of the water."
Klaus Joehnk, who arrived from Australia together with his colleague Janet Anstee, is especially interested in blue-green algae and whether their growth changes during the mixing of lakes. If, for example, an algal bloom occurs after a severe storm at a lake which is used as a drinking water reservoir, it may harm those who drink the water because of the toxins produced by some blue-green algal species. As their name suggests, these algae make the water appear blue-green. They can also be detected by means of a pigment that they need in photosynthesis, known as phycocyanin. "Unfortunately, however, the satellites that currently provide images of the Earth from space lack the wavelength bands that can be used to quantify blue-green phycocyanin," says Klaus Joehnk. At present, such so-called hyperspectral sensors suitable for high spatial resolution are predominantly available as hand-held or underwater devices. At the LakeLab, Joehnk and Anstee use a series of instruments to measure the underwater light spectrum at different depths.
Yet not only phytoplankton has an effect on water colour. Due to the large distance, satellites mainly measure the light scattered in the atmosphere between the satellite and the water surface. This accounts for up to 98 % of the signal observed in space. The reflections of the sun and the sky on the surface of the water also play a role. "All these interfering factors have to be accounted for to get to the actual colour of the water," says Philipp Grötsch of the City College New York. For this purpose, he developed an algorithm that allows subtracting these reflections, which depend on the illumination conditions (i.e. clear, cloudy). Grötsch wants to incorporate measurement data into his algorithm, differentiating the direct sun light from the diffuse light of the sky and clouds. For this reason, Grötsch started working with John Wood. The English engineer has developed an instrument that measures these two light components separately. Together, Grötsch and Wood are now also using it at the LakeLab.
The so-called diffuse attenuation coefficient Kd is another factor used in remote sensing models. It is a measure of the transparency of a water body and describes the attenuation of the incident light throughout the water column. "Here at the LakeLab we measure Kd with the KdUinoPro, our self-constructed and open-source device. It consists of waterproof boxes in which we have installed light sensors," explain Raul Bardaji Benach and Carlos Rodero Garcia. But because the transparency of the water can vary at different depths – e.g. because phytoplankton may accumulate in distinct layers of the water column – the two Spanish researchers have attached their measuring boxes to a cord at intervals of 50 cm and thus measure the diffuse attenuation coefficient Kd for the entire water column. Their instrument design is freely available online and has been incorporated in educational programs and citizen science efforts around the globe.
These different questions and approaches brought together the remote sensing experts at the LakeLab. "The direct exchange with my colleagues was an important reason for me to come to Lake Stechlin," says Philipp Grötsch. His colleagues agree: It is an extremely productive time for them, allowing to discuss their measurement methodologies and experiences in a relaxed environment, and to jointly develop ideas to further advance water remote sensing.
The scientists
The environmental engineers Dr. Carmen Castro Cillero and Federico Cheda work at 3edata in Lugo, northern Spain. This remote sensing company was founded by Carmen Cillero Castro and two colleagues as a spin-off of the University of Santiago de Compostela.
Dr. Janet Anstee and Dr. Klaus Joehnk are group leaders at the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Canberra, Australia.
Engineer Raul Benach Bardaji and PhD student Carlos Garcia Rodero work together at the Marine Science Institute of the Supreme Council for Scientific Research – CSIC (Consejo Superior de Investigaciones Científicas) in Barcelona, Spain.
After his doctoral thesis in the field of remote sensing, physicist Dr. Philipp Grötsch has been working as a postdoctoral fellow at the City College in New York since the beginning of 2019. John Wood is an engineer and equipment designer specializing in solar radiation and runs his own company, Peak Design Ltd. in Winster/Derbyshire, England.
The stay of the researchers at the LakeLab is financed by the EU project AQUACOSM, the network of leading European AQUAtic MesoCOSMen facilities, which is managed by IGB.
