Day 4

Detailed paper information

Back to list

Paper title Exploring Sentinel-3 based water monitoring for assessing seasonal and spatio-temporal dynamics of the largest lake in the Caucasus region, Lake Sevan
  1. Azatuhi Hovsepyan Center for Ecological-Noosphere Studies, National Academy of Sciences Speaker
  2. Shushanik Asmaryan Center for Ecological-Noosphere Studies, National Academy of Sciences
  3. Gor Gevorgyan Institute of Hydroecology and Ichthyology Scientific Centre of Zoology and Hydroecology, National Academy of Sciences
  4. Vahagn Muradyan Center for Ecological-Noosphere Studies, National Academy of Sciences
  5. Anahit Khlghatyan CENS
  6. Lilit Sahakyan Center for Ecological-Noosphere Studies, National Academy of Sciences
  7. Armine Hayrapetyan Institute of Hydroecology and Ichthyology Scientific Centre of Zoology and Hydroecology National Academy of Sciences
  8. Bardukh Gabrielyan Institute of Hydroecology and Ichthyology Scientific Centre of Zoology and Hydroecology National Academy of Sciences
  9. Thomas Heege EOMAP GMBH & CO. KG
  10. Hendrik Bernert EOMAP GMBH & CO. KG
  11. Martin Schultze Helmholtz Centre for Environmental Research (UFZ)
  12. Karsten Rinke Helmholtz-Centre for Environmental Research
Form of presentation Poster
  • A7. Hydrology and Water Cycle
    • A7.06 EO for monitoring water quality and ecological status in inland waters
Abstract text Lake water quality is a key factor for human wellbeing and environmental health being affected by climate change, anthropogenic activities such as urban and domestic wastewater charges into in inflowing streams, as well as agricultural activities. Generally, in situ measurements are traditionally conducted and widely accepted as instruments for water quality monitoring. However, in many regions, classical monitoring capacities are limited and in case of large water bodies they lack monitoring at the required spatial and temporal scales. In this context remote sensing has great potential and can be used for assessing spatio-temporal dynamics of water quality in a cost- effective and informative manner.
Copernicus Sentinel-3 OLCI instrument launched in February 2016 covering 21 spectral bands (400-1200nm). It has been providing accessible products since 2017, which enables monitoring water bodies at 300m resolution and almost daily scales.
Lake Sevan (40°23‘N, 45°21‘E) is located in Gegharkunik province in Armenia at an altitude of 1900 m a. s. l.. Lake Sevan is Armenia’s largest water body and the largest freshwater resource for the whole Caucasian region. At present, lake water quality is progressively deteriorating due to eutrophication, water level fluctuations, and climate warming and suffers from massive cyanobacterial blooms. Hence, it is very important to study key water-quality variables (e.g. Chl-a, turbidity, harmful algae bloom etc.) describing the ecological status of the lake, and their annual and seasonal dynamics. We emphasize that remote sensing can significantly contribute to this monitoring demand.
In situ measurements having been implemented since 2018 in the frame of German-Armenian joint projects (SEVAMOD and SEVAMOD2) providing detailed information on water quality of the lake on a local scale. These data are, on the one hand, well suited for a basic characterization of the problem but do not provide enough data for tracking the qualitative changes of the water in space and time.
Hence, our research aimed at assessing the seasonal and spatial water quality dynamics in Lake Sevan over 5 years covering 2017-2021 by using Copernicus Sentinel 3 products.
The satellite data processing engine for water quality assessment eoLytics was used for Sentinel 3 data processing. EoLytics is based upon the MIP Inversion and Processing System, that was initially developed at DLR from 1996 and has been continued since 2006 by EOMAP. The fully physics based sensor-generic algorithm in MIP do not require in-situ data for calibration.
The full range of daily incoming time series for Lake Sevan that is free of clouds includes 474 scenes. Using EoLytics algorithms the data were processed for the water quality parameters: Chl-a, total suspended matter (TSM) and harmful algae bloom (HAB-Indicator). It is noteworthy that Chl-a and TSM processing algorithms provide fully quantitative outputs while the HAB algorithm provides a semi-quantitative indicator.
Field campaigns of in situ measurements have been conducted since 2018 on a monthly basis. andIn situ measurements are available at two locations for 2018-2020 years and at three locations for 2021.
This study envisaged the following steps: (i) the analysis of the annul and seasonal remotely sensed data in order to reveal the spatial-temporal characteristics of water quality; (ii) the comparison of in situ measured and remotely retrieved data via regression analysis in order to understand the relationship of the data received from the different sources.
The spatial-temporal analysis of the seasonal characteristics of Chl-a follows typical plankton succession dynamics in large lakes and usually shows maximum values in summer (June and July). However, the seasonal dynamics of Chl-a differ between years and are most likely driven by meteorological dynamics. This link between meteorological variables and plankton dynamics is a key aspect for climate impact assessments and are hardly visible in traditional monitoring data but well observable in remotely sensed data due to its higher temporal resolution. The spatial patterns in the lake point to a large influence from external nutrient inputs as high Chl-a values are repeatedly observed in the vicinity of polluted inflows.
The HAB repeats the overall seasonal trend we have for chlorophyll-a with the highest appearance in 2018, then a bit less in 2019. It follows with much more low appearance in 2020 and 2021. 2017 gives the lowest picture of the HAB occurrence.
An initial comparison of in situ measured and remotely sensed Chlorophyl-a via a linear regression revealed a significant relationship with a relatively low error (RMSE=0.403).
At this stage it can be concluded that the maximum of the Chl-a content during these 5-year period slightly shifted from August to June and was regularly associated with the formation of HABs. The validation of the remote sensing data needs ongoing efforts in order to facilitate a deeper analysis of seasonal trends and spatial-temporal patterns. The observations based on Sentinel-3 sensors provide extremely valuable information on water quality dynamics in Lake Sevan and complement the results from traditional monitoring. The long-term monitoring strategy should therefore exploit the strengths of both approaches and remote sensing is considered to be a key aspect of the foreseen monitoring program for this highly sensitive and important water body.
This study was supported by following projects:
1. SevaMod - Project ID 01DK17022 - Funding Institution: Federal Ministry for Education and Research of Germany "Development of a model for Lake Sevan for the improvement of the understanding of its ecology and as instrument for the sustainable management and use of its natural recourses"
2. “SevaMod2: Project ID: 01DK20038 - Funding Institutions: Federal Ministry for Education and Research of Germany and 91.5% of planned costs), Ministry of Environment of the Republic of Armenia (8.5% of planned costs) “Building up science-based management instruments for Lake Sevan, Armenia”
3. Project ID - 20TTCG-1F002 - Science Committee of the Ministry of Education and Science, Culture and Sport of RA "The rising problem of blooming cyanobacteria in Lake Sevan: identifying mechanisms, drivers, and new tools for lake monitoring and management"
4. Project ID - 21T-1E252 - Science Committee of the Ministry of Education and Science, Culture and Sport of RA "Assessing spatio-temporal changes of the water quality of mountainous lakes using remote sensing data processing technologies".