European Facility For Airborne Research

COOLAPEX

Towards assessment of water quality of the Curonian Lagoon using hyperspectral APEX sensor: optical water properties, phytoplankton and macrophytes

Scientific project

TA-014. Airborne imaging for environmental science applications.

Satellite calibration/validation

VAICIUTE Diana

Remote sensing, bio-optical water parameters, hydrobiology, plankton ecology, eutrophication processes, water quality assessment. Participation in national and international projects: MAURAKUMA "Distribution of Charophyte in the Curonian Lagoon and impact of environmental factors" (Projects of National Research Council), EEE "A system for the sustainable management of Lithuanian marine resources using novel surveillance, modeling tools and an ecosystem approach" (Norwegian Grants), INFORM "Improved monitoring and forecasting of ecological status of European INland waters by combining Future earth ObseRvation data and Models" (7BP).

Giardino C, Bresciani M, Pilkaitytė R, Bartoli M and Razinkovas A, 2010. In situ measurements and satellite remote sensing of Case 2 waters: Preliminary results from the Curonian Lagoon. Oceanologia 52:197-210.

Bresciani M, Giardino C, Stroppiana D, Pilkaitytė R, Bartoli M and Razinkovas A, 2012. Retrospective analysis of spatial and temporal variability of chlorophyll-a in the Curonian Lagoon. J. Coastal Conserv. 16:511-519.

Adamo M, Matta E, Bresciani M, De Carolis G, Vaiciute D, Giardino C and Pasquariello G, 2013. On the synergistic use of SAR and optical imagery to monitor cyanobacteria blooms: the Curonian Lagoon case study. European Journal of Remote Sensing 46:789-805.

Bresciani M, Adamo M, De Carolis G, Matta E, Pasquariello G, Vaičiūtė D, Giardino C, 2014. Monitoring blooms and surface accumulation of cyanobacteria in the Curonian Lagoon by combining MERIS and ASAR data. Remote Sens. Environ. 146:124-135.

Vaičiūtė D, Bresciani M, Bartoli M, Giardino C, Bučas M, 2015. Spatial and temporal distribution of coloured dissolved organic matter in a hypertrophic freshwater lagoon. Juornal of Limnology 74(3): 572-583.

The Curonian Lagoon is a large, shallow (total area 1584 km2, mean depth 3.8 m), mainly freshwater estuary located in the south-eastern part of the Baltic Sea. It is divided between Lithuania (northern part, 26 % of the lagoon area) and the Russian Federation. The Curonian Lagoon provides significant ecological services to society: aquaculture, fishery, recreation, water supply and transport. This aquatic system plays an important role in carbon and other global biogeochemical cycles. The Curonian Lagoon is naturally productive water basin that has been impacted by the undesired processes of both eutrophication and climate change. Nowadays the lagoon is considered to be hyper-eutrophic with recurring spring diatom blooms followed by summer cyanobacteria blooms. Cyanobacteria blooms are a major concern in this region because the chlorophyll-a concentration can be extremely high (up to about 200 mg m-3) and, under specific climate conditions, can be associated with a surface accumulation of algae. Due to intensive blooms the lagoon suffers from limited light penetration, anoxia, production of toxic substances by toxic cyanobacteria species, unbalanced nutrient cycle. Large biomass of both emerged and submerged macrophytes, mostly confined in the littoral zones, contributes to the evidence of ongoing eutrophication. The connection of the lagoon with the Baltic Sea and Nemunas River defines it as a transitional riverine-like system with a mixing of brackish, lagoon and CDOM-rich riverine waters. Currently, our ability to monitor such a large ecosystem is limited by a number of in situ stations, samples and scarce collaboration with neighboring country. Satellite remote sensing is a valuable asset for monitoring such a large water bodies for various purposes. However, some of the present features of the Curonian Lagoon (presence of cyanobacteria, seasonal succession of different algal groups and functional types) cannot be properly monitored with multispectral sensors currently onboard satellites due to unsuitable spectral resolution. Therefore, hyperspectral sensors have become necessary.

DO228-101 - DLR

The APEX sensor on the DLR aircraft is critical to the success of the project. The choice of the Curonian Lagoon as one of our dedicated study sites is strongly based on the availability of the existing field radiometric measurements, in situ data, available facility for the field campaign that will be funded by FP7 INFORM project. The field, airborne and satellite data collected during this field campaign will produce one of the most extensive, datasets for optically complex waters and an invaluable resource for the scientific community. The level-1b data from APEX will be processed to water-leaving reflectance by VITO. APEX is the only sensor for which water-leaving reflectance products are provided operationally in EUFAR.

The objectives of this project are:
• to test and validate algorithms for retrieval of the different main phytoplankton functional types, the absorption coefficient of yellow matter, phytoplankton primary production, spatial distribution and macrophyte functional groups;
• to evaluate the spatial and temporal dynamics of phytoplankton and correlate this intraday variability with meteorological conditions during the study period, with the final goal of evaluating their influence on phytoplankton growth dynamics;
• to extend ongoing activities of FP7 INFORM project on the development of specific algorithms that will be validated using field radiometric measurements and hyperspectral data simultaneously acquired during field campaign carried out in the Curonian Lagoon;
• to validate Sentinel-2 and Sentinel-3 (in case it will be fully operational) data with collected in situ and APEX data;
• to improve the knowledge of instrumentation currently available for monitoring activities;
Proposed work. The hyperspectral APEX data will be collected over entire Lithuanian part of the Curonian Lagoon (approximately 400 km2) and the entrance to the Klaipeda strait (approximately 25 km2), where turbid and productive lagoon waters form the plume in the coastal waters. Additionally, 2 or 3 overpasses will be planned over the selected area in the Curonian Lagoon but at different time in order to assess the dynamics of cyanobacteria and other ecological parameters during the day. Simultaneously with the acquisition of hyperspectral APEX data a series of radiometric field campaigns will be organized aiming at the collection of data for calibration and validation of the airborne radiometric signal. Field measurements will be carried out at different stations covering optically different water masses due to presence of cyanobacteria, high amount of CDOM or suspended particulate matter. Moreover, we will acquire ancillary data (like aerosol optical thickness), which are necessary to parameterize atmospheric correction models. In the littoral part of the Curonian Lagoon macrophyte functional groups (emergent, submersed and free-floating) will be identified and their distribution will be assessed. In order to improve the characterization of the optical properties, collected data will be integrated with those acquired in recent years (2009-2015) and during monthly 2016 field campaigns organized in the frame of the INFORM Project.
Anticipated output. The final product will be the maps of the relevant optically active parameters chlorophyll-a, coloured dissolved organic matter, suspended particulate matter, different phytoplankton groups/functional types, distribution of macrophytes. The obtained results will be integrated with existing historical trends of investigated parameters in order to assess temporal patterns of the ecological conditions, algal blooms, presence of cyanobacteria, the effect of discharge by the Nemunas River on CDOM, spatial distribution of macrophytes in the Curonian Lagoon.

The ideal conditions for the acquisition are: clear-sky with low wind speed (< 10 m/s), no rain.

Time between 1st June and 30th September 2016 would be acceptable, although July and August are preferred since the occurrence of algal blooms in the study area is mostly probable. The dates would need to be agreed well in advance so that the ground validation teams can make the necessary arrangements for access to research vessels and laboratory facilities. If possible, we will synchronize the APEX flights with an overpass of Sentinel-2, Landsat 8 or Sentinel-3 (if it will be fully operational).

The Curonian Lagoon (Fig. 1, A) is the largest lagoon in the Europe (total area 1584 km2). It is also a highly complex system in terms of its hydrodynamic and bio-geo-optical properties. The northern part of the Lagoon is connected to the SE Baltic Sea by a narrow strait. Under specific conditions (wind, water level difference), brackish less turbid and less productive waters may enter into the Curonian Lagoon and change the water transparency and optical properties. Although the Curonian Lagoon is shallow (average 3.6 m depth) waterbody the Klaipeda Strait is artificially deepened up to 16 m in the harbor area. During summer period under specific hydrometeorological conditions the two layer water masses occur, due to the inflow of brackish Baltic Sea waters. The temporal stratification can be formed, where upper layer is fresh and more turbid, whereas lower water layer is brackish less turbid and colder. At the eastern part of the Curonian Lagoon itself has a main freshwater inlet - the Nemunas River. The river drains a large watershed, hosting wetlands and forests, and discharges annually nearly 23 km3 of CDOM-rich freshwater. CDOM absorption at 440 nm ranges from 1.3 m-1 up to 4.0 m-1. Nowadays the Curonian Lagoon is considered as eutrophic or hyper-eutrophic with recurring spring diatom blooms (mainly Stephanodiscus spp., Aulacoseira spp.) followed by summer cyanobacteria blooms, mainly involving Aphanizomenon flos-aquae, Planktothrix agardhii and Microcystis spp. Mean chlorophyll-a concentrations is approximately 75 mg m-3, however in some cases it can be extremely high (up to about 500 mg m-3). Shallowness, wind regimes, sediment resuspension and frequent blooms make this system permanently turbid, with Secchi depth ranging from 0.6 up to 2.0 m. Despite low water transparency the submersed macrophytes form meadows (especially charophytes: Chara aspera and Chara contraria), which are mainly restricted to the eastern part of the lagoon. Compared to historical data, there is increase of vegetated areas by charophytes and their maximum depth in the estuarine part of the lagoon. Other submersed macrophytes (Cladophora, Ulva, Potamogeton, Zannichellia and Myriophyllum species) also generally distributed in the same part of the Curonian Lagoon. Emergent macrophytes (dominant by Phragmites australis) are found on both sides of the lagoon. The width of emergent and submersed macrophytes’ belts vary from 15 to 1500 m wide, being widest in the eastern part of the lagoon, where the bottom slope is relatively gentle compared to the western littoral.
In the recent years, the Curonian Lagoon has been object of several field campaigns aiming at the acquisition of radiometric data (2009, 2011-2012 and 2014-2015) and limnological data (monthly or seasonally during period 2009-2015). However, there are no existing hyperspectral airborne data available to support algorithm development and validation studies.

The preliminary flight plan covers 2 days:

Day 1. to cover all Lithuanian part of the Curonian Lagoon (approximately 400 km2) and the plume of lagoon water (25 km2) where temporal water mass stratification is mostly probable (Fig. 1, B);
Day 2. to have 2 or 3 overpasses over a smaller area (approximately 200 km2) but at 2-3 different times (i.e. morning, noon and/or afternoon) during the day. The selected area represents the confined region of the Curonian Lagoon, where cyanobacteria accumulation is most probable and dynamics of CDOM-rich riverine waters can be visible (Fig. 1, C).

Spatial resolution of 3 m (swath width of 3 km, i.e. 1000 pixels) was discussed with the APEX operator. The flight lines should be directed into the sun, or away from it to minimize BRDF effects. This preliminary flight plan resulted in approximately 16 flight lines on Day 1 and approximately 22 flight lines on Day 2.

In total 10 APEX flight hours (transit + acquisition) are requested to cover the Lithuanian part of the Curonian Lagoon on day 1 and 2-3 times smaller area on Day 2.

Hyperspectral at-sensor radiances (L1b) for computation of water-leaving reflectances following atmospheric correction.

APEX

No additional instruments

0

In situ data will be collected simultaneously to the planed APEX overpasses at NADIR. Two research vessels will be deployed to obtain in situ measurements of remote sensing reflectance, inherent optical properties (absorption, scattering and backscattering). During field campaigns water samples for laboratory analysis of chlorophyll-a, phycocyanin, total suspended matter (plus inorganic/organic fractions), coloured dissolved organic matter, absorption by particles (plus pigments and non-algal particles), dissolved organic carbon, particulate organic carbon, phytoplankton species composition and abundance, water turbidity will be collected. Additionally primary productivity will be measured using O2-evolution and 14C techniques. Simultaneously, measures of macrophytes species composition and abundance will be performed. Biomass estimation will be derived from measurements of individual plants and plant density. We expect to sample between 10 and 15 stations per day over the [2-day] campaign. The date of the APEX flight need to be agreed with the PI well in advance in order to arrange access to vessels and lab facility and to mobilize the ground team.

In situ measurements and water sampling
In situ data will be collected at 10-15 stations simultaneously to the planed APEX overpasses. The planed measurements are following:
I. Inherent optical properties:
Those data will be collected together with subsurface and above surface reflectances and with optically active in-water constituents.
• Vertical profiles of total backscattering coefficient (bbtot) will be collected in the following channels: 442, 488, 510, 550, 620, and 676 nm with HOBILabs HydroScat-6.
• Absorption by pigments and non-algal particles will be determined using filter pad technique. The absorption with be measured with UV/Vis spectrophotometry in all range (350-750 nm) with 1 nm of spectral resolution.

II. Subsurface and above surface reflectances:
In situ water reflectance data will be collected with the spectroradiometer FieldSpec (Analytical Spectral Device Inc.) that will be operated both above the water surface and into the water column. Above surface in situ water reflectance will be measured with WISP-3 (Water Insight) and TriOS RAMSES radiometers.
This measure could be generate the water leaving reflectance above the water surface Rr(0+), subsurface water leaving reflectance above the water surface Rr(0-).

III. Optically active in-water constituents. Water samples will be taken from the surface using a bucket or Ruttner water sampler (where vertical profiles are performed). Samples (three replicates per station) will be collected and transferred in opaque HDPE bottles and transported with ice packs to the laboratory within 2-4 hours, for subsequent treatment and analysis.
• Samples for chl-a and phycocyanin measurement will be filtered through glass fiber GF/F filters with a nominal pore size 0.7 μm (in triplicate) and frozen in liquid-N prior to analysis by spectrophotometry.
• The concentration of total suspended matter (TSM) will be determined gravimetrically by passing a sample aliquot in triplicate through pre-weighed and pre-combusted 47 mm Whatman GF/F filters. The inorganic and organic fraction will be determined by burning the organic material on the filter papers at 500 ºC for 4 h.
• The spectral absorption coefficients of coloured dissolved organic matter (CDOM) will be determined using UV/Vis spectrophotometry. Prior the analysis water samples will be filtered through 47 mm Nucleopore 0.2 µm filters.
• The concentration of dissolved organic carbon and particulate organic carbon will be determined using a Shimadzu TOC analyser.
• Phytoplankton samples will be preserved with acetic Lugol’s solution. Phytoplankton species composition, abundance and bio-volume will be determined using the inverted microscope technique (Utermöhl, 1958), according HELCOM (1988) recommendations.
• Phytoplankton primary production will be measured at selected stations using the O2-evolution and 14C methods.
• Turbidity will be measured with portable Eutech turbiditimeter TN-100.
• Nutrient concentration (if applicable) will be determined using a 4-channel continuous flow analyzer (San++, Skalar) using standard colorimetric methods (Grasshoff et al., 1983).
• Supplementary CTD, PAR (with LiCOR instrument) profiles and vertical distribution of main phytoplankton groups (cyanobacteria, green algae, diatoms+dinoflagellates, cryptophytes) and yellow matter will be determined by portable fluorometer FluoroProbe II. At the deeper sites the radiometric profiles for spectral Kd will be measured.

IV. Macrophytes.
Macrophyte distribution will be surveyed along transects extending from the coastline or reed stands down to the lower macrophyte depth limit. Emergent plant density will be measured using a 1.0 m2 frame, which will be randomly placed in replicates over macrophytes. Samples of submersed plants will be collected from the boat using a S. Bernatowicz type grab. Total coverage (%) of each macrophyte species will be assessed every 0.25 m depth interval (at least 3 replicates in each depth zone). Additionally the distribution and abundance of submerged macrophytes will be assessed by rakes and echo-sounder (Humminbird 898c SI Combo). Macrophytes will be collected and brought back to the laboratory for species identification. The plant dry weight will be determined (70 °C). Total macrophyte biomass will be calculated multiplying their abundance by mean biomass. These measures will be used to calibrate data derived from space-born images.
At the same time will be acquired radiometric measurements of different sub-samples of leaves of macrophytes, both with contact probe and as reflectance considering the air interface vs vegetation.

V. Aerosol optical thickness. Sun photometer measurements (Microtops II) will be made to assist the atmospheric correction of the APEX data.

APEX data
Standard, the APEX data will be radiometrically corrected, geocoded and atmospherically corrected to reflectance by VITO.
For studying water quality the APEX data will be atmospherically corrected to water-leaving reflectance by VITO.
APEX processing up to Level-2 (geocoded reflectance) is performed using the semi-automated VITO Central Data Processing Center (CDPC).
For further APEX processing and analysis the participants will use e.g. ENVI-IDL image processing software. For estimating water quality properties e.g. BOMBER, an ENVI/IDL add-on tool, will be used (Giardino et al., 2012) and semi-empirical algorithms.
The atmospherically corrected APEX data will be used to test at Curonian Lagoon a range of water quality algorithms (phytoplankton functional types, yellow matter and macrophytes) developed within the FP7 INFORM project.

The field work and the analysis of the field data, airborne data and satellite data for water quality and macrophytes will be mainly supported by the FP7 INFORM (GA 606865) project. The work will also be partly supported by national project MAURAKUMA (MIP-14210) funded by the National Research Council. The work will be led by Dr. Diana Vaičiūtė and Dr. Artūras Razinkovas-Baziukas and supported by other researchers (Dr. Martynas Bučas, Dr. Jolita Petkuvienė), PhD students and several technicians. The work will be done in collaboration with experienced researchers Dr. Claudia Giardino, Dr. Mariano Bresciani, Dr. Paolo Villa from the Institute for Electromagnetic Survey of the Environment of the National Research Council (CNR), Italy, Dr. Peter Hunter, Professor Andrew Tyler, Dr Evangelos Spyrakos, Dr Claire Neil, from the Earth observation group of University of Stirling (USTIR), UK and Dr. Marco Bartoli, Dr. Gianmarco Giordani and Dr. Rossano Bolpagni from Parma University, Italy. The USTIR, CNR and Parma University teams are also very well equipped for in situ optics and radiometry with a range of state-of-the-art instrumentation. Further expertise at VITO, Belgium (Sindy Sterckx, Els Knaeps and Liesbeth De Keukelaere) can be called upon through the FP7 INFORM project, particularly in relation to Sentinel-2 atmospheric correction for water. Klaipeda University (KU) has three laboratories that are equipped by basic instruments that can be fully exploited for this project and boats that will be used during the field campaigns in the Curonian Lagoon.

The full list of participants:
Klaipeda University (Lithuania): Dr. Artūras Razinkovas-Baziukas, Dr. Diana Vaičiūtė, Dr. Martynas Bučas, Dr. Jolita Petkuvienė;
USTIR (UK): Dr. Peter Hunter, Professor Andrew Tyler, Dr. Evangelos Spyrakos, Dr. Claire Neil, Pierre Mercatoris, Maria Maestro.
CNR (Italy): Dr. Claudia Giardino, Dr. Mariano Bresciani; Dr. Paolo Villa;
Parma University (Italy): Dr. Marco Bartoli, Dr. Gianmarco Giordani, Dr. Rossano Bolpagni.

1st June to 30th September 2016. 10:00 to 15:00 CEST.

No

The results of the project will be published in the international scientific journals. Results will be delivered for the Department of Marine Research, Environmental Protection Agency, in order to provide useful information for the management and monitoring of the Curonian Lagoon. PhD, masters-level and undergraduate students will be fully involved within the project, assisting in the field campaigns, laboratory analysis, data processing and reporting.

Damien Bouffard, damien.bouffard@epfl.ch
Tiit Kutser, Estonia, Tiit.Kutser@sea.ee
Peter Gege, DLR, peter.gege@dlr.de
Steef Peters, The Netherlands, peters@waterinsight.nl

The airborne data acquired within the proposal will enable us to extend ongoing activities of FP7 INFORM project on the development of inland water quality algorithms focused on Sentinel-2, Sentinel-3 and future hyperspectral satellites and test these algorithms using COOLAPEX airborne hyperspectral data simultaneously acquired during an FP7 INFORM and MAURAKUMA (MIP-14210) funded field campaign carried out in the Curonian Lagoon.

Yes

1

7

EUFAR website