IES develops satellite data-based method for lake water quality monitoring

Researchers and Estonian post-doctorate Tuuli Soomets from the Institute for Environmental Solutions (IES) has developed new method for satellite data-based lake water quality monitoring.

The importance of inland water bodies including lakes has never been greater because they help to regulate the carbon cycle, thus impacting local and regional climate. Moreover, lakes are economically important – they provide drinking water, fishing and recreation opportunities. Therefore, the monitoring of lakes and water quality has become a global concern. Responding to it, EU implements the Water Framework Directive to reach good water quality status in all lakes in Europe that are larger than 0.5 km2. To be able to contribute Latvia has to improve the methods and fill the gaps of lake monitoring.

IES leading researcher Dr. Dainis Jakovels and post-doctorate Tuuli Soomets showcased the results of the research.

IES: Why is was necessary to develop remote sensing and satellite data-based lake monitoring method?

Dr. Dainis Jakovels (D.J.): Till now lake water quality in Latvia (Latvia’s national water monitoring program) was done with standard methods – scientist goes to the lake, takes samples, and brings them to laboratory for analysis. It is expensive and time-consuming process. Moreover, it does not allow to monitor all the lakes every year. Currently, researchers can have obtain data from each lake only once every 4 years. During this research we developed satellite data – based remote sensing method.

Tuuli Soomets (T.S.): Remote sensing method is giving a lot of advantages to the traditional lake monitoring methods in terms of spatial and temporal coverage. With traditional methods you do sampling in all lakes each month. But satellites fly over Latvia at least every 5 days. If there are no clouds and we can see the lake in satellite images we have free access to the data that allow us access high accuracy information about chlorophyll concentration (if the lake is blooming or not) and other parameters.

IES: Does this newly developed lake water quality monitoring method can replace in-situ methods?

D.J.: I wouldn’t say that. Rather, this method complements and support standard water quality monitoring approaches. . Our goal is to develop method that will give an overall understanding of the lake water quality status in Latvia. This information will allow to observe the development of water quality status in lakes. It will also support the planning process for water quality monitoring campaigns using standard methods.

T.S.: Sentinel satellites were launched in 2015 and since then data has been gathered regularly. We can have a retrospective data about certain lakes – assess anomalies and understand how the eutrophication of lakes historically have happened. For example, in case of lake Burtnieks, we can see that eutrophication have happened at the very beginning of 21st century. We can also compare consistency of satellite data and information coming from  standard monitoring methods.

IES: For development of this water quality method the Sentinell-2 and Sentinell-3 satellite data was used. What kind of data gathering, processing and analysing approaches you used to obtain information on lake water quality in Latvia?

D.J.: We used Sentinel-2 and Sentinel-3 satellite optical data to get water quality parameters – chlorophyll concentration, transparency, total suspended matter, for example, colour dissolved organic matter. Chlorophyll concentration is the blooming of water plants in lakes, something that everyone understands. It can be a natural process but usually it is caused by human activities, such as strong agricultural practices or intensive decline of fish population doing angling or other practices.

T.S.: At the start, we receive spectral satellite data sets – images at different wavelengths. Before further data processing, we can already visually assess the condition of the lakes, but it can be inaccurate, because it is influenced by the atmosphere and other factors. Therefore, we developed standardised satellite data processing algorithms, that provides accurate information on the water quality parameters of the lakes.

All lakes are not the same, therefore, primarily we developed multiple-step algorithm that determines lake optical water type. Further, we developed another algorithm that interprets each of the pixel in satellite image to get the parameters that we were interested in – chlorophyll concentration, transparency, total suspended matter, for example, colour dissolved organic matter, etc. In short, we developed a mathematical formula (algorithm) that expresses water quality factors from spectral satellite data. For information validation we conducted sample gathering campaigns in lakes using standard methods.

IES: What is the main result of this research?

T.S.: The result of this research is 3 internationally recognized research papers that are published in “Remote Sensing, An Open Access Journal from MDPI”. The knowledge and methods we created can be used for further lake research. It is an open detailed information that describes (satellite data processing, algorithm development etc.) how to determine lake water quality parameters. A specialist with knowledge and experience in remote sensing and Earth observation is required to use these materials.



IES: Remote sensing aproaches are widely used in the word. What is our contribution and innovation of this research?

D.J.: Of course, we are not the only ones working with remote sensing-based freshwater exploration. Almost every country in the world have interest in water quality monitoring with satellite data. But lake types in each region are different, so it is necessary to adapt the method. There are universal algorithms available but as you know – common solutions do not work as precisely for specific cases. Solutions that are adapted for specific region works more accurately.

T.S.: In my opinion, lake productivity studies were the most special part of this research – determining the biomass of water plants, such as algae, that estimates lake overgrowing level. Remote sensing-based primary production level estimation in lakes is complex and rarely used method.

D.J.: Another innovation in this research was that we used combination of Sentinel-2 and Sentinel-3 satellite data. Usual practice is to concentrate on one of the satellites. In this research we combined both satellite information that provides us with more frequent data flow. Sentinel-3 satellite data has lower resolution and is more suited for  larger lakes. However, in the case when there is a cloud cover over Latvia during Sentinel-2 overflight, Sentinel-3 can provide backup information highlighting quality status of water bodies.


IES: What are the IES’s future plans regarding remote sensing-based water quality research in lakes?

D.J.: IES have always had interest in water research using remote sensing approaches. We continuously are improving our knowledge in both areas – water research and remote sensing. Currently we have reached a certain knowledge level and ready to share our expertise in Latvia and globally. We already are leading another remote sensing-based freshwater research with similar approach.

The research is a part of the project “ICT-based remote sensing tool for retrieving water quality products for lake research and monitoring” (FLUID), No. Agreement with State Education Development Agency of the Republic of Latvia No. The project is funded by the European Regional Development Fund and partly funded by State budget and the Institute for Environmental Solutions.

Find more about the FLUID project here.