Monitoring Baltic Sea eutrophication from space-

from research to applications

Susanne Kratzer, Associate ProfessorCoordinator of NordAquaRemS

Department of Systems Ecology, SUSuse@ecology.su.se

Marine remote sensing and management

• Covers large areas • Cost-effective• Synoptic view• Visual format easy to

understand• New window into

marine ecosystems

Cyanobacteria Bloom in the Baltic Sea, MERIS FR image Date20050713_Time094518_Orbits17611, ESA

Elements for Ecological Status in the EC Water Framework Directive (WFD):

Biological elements:Phytoplankton, aquatic flora, benthic invertebrate fauna, fish fauna

Hydro-morphological elements (supporting the biological elements): Morphological conditions, Hydrological and Tidal regime

Chemical and physicochemical elements (supporting the biological elements): General: dissolved oxygen, nutrients, transparency, temperature; Specific: synthetic and non-synthetic pollutants

Chl-a and Secchi depth can be used as indicators for eutrophication

SSTSecchi depth

Productivity

Baltic Sea remote sensing and managementBaltic Sea remote sensing and management

Light quality& quantity

Eutrophication

Large scale dynamics

Uppwelling

Currents

Algal blooms

Humic Substances

Sedimenttransport

SST in the NW Baltic Sea Derived from AVHRR

7-16 July 27 July - 5 August 6-15 August

15-26 August 26 August - 4 Sept

Himmerfjärden

°C

BY31 BY31

BY31 BY31

BY31

Sea Surface Temperature (SST) derived from NOAA/AVHRR data during7 July-4 September 2001, binned images (10 day composites)

Attenuation of light in the sea

90% of TOA signal

10 % of signalfrom the sea

Attenuation of light

Classification: Morel and Prieur, 1977

water

Case-2

Case-1

Kurt Holacher, 2002

Oceanic – Optical Case-1 waters Coastal – Optical Case-2 waters

Light penetration in the open sea and in coastal waters

Jerlov’s water mass classification

Jerlov's optical classification into the oceanic water types I-III and the coastal water types 1-9 (Jerlov, 1976).

Baltic Sea

Clear tropical

Baltic Sea remote sensing and management

Decision makingEnvironmental policies and

managementSocio-economics

(e.g. tourism and fisheries)Public health

(e.g. toxic blooms)

Satellite remotesensing

Sea-truthing

Atmosphericcorrections

Bio-opticalmodelling

Coastal Zone monitoring:•Ecological and physiologicalvariables

•Physics and chemistry•Optical biogeochemical variables:

CDOMCDOM, ChlorophyllChlorophyll, , SPMSPM

Platforms:Research and monitoring vessels, moorings, ships-of-opportunity

Large scalesynopticinformation

Small and meso scaleinformation

from Kratzer et al, 2003

Södertälje

BY31

BIIIBII

BI

B1

H2

H3

H4

H5

Trosa

Askö

STP

Landsat image of Himmerfjärden5 May 1986

Main area of investigation:

Himmerfjärden and Open Baltic Sea

Landsort

Deep

STP outlet

In-water algorithm: SD = (0.55 * Kd (490) – 0.04) -1

SeaWiFS Kd(490) imageSecchi Depth map of the Baltic Sea derived from SeaWiFS(end of July/beginning of August 1999)

m

Kratzer et al. , 2003

Cyanobacteria Bloom in the Baltic Sea, MERIS FR image Date20050713_Time094518_Orbits17611, ESA

FR

RR

RR

FR

H5

H4

H3

H2

B1

H5

H4

H3

H2

B1

Sewage Treatment Plant

FR (300 m resolution)

RR (1.2 km resolution)

08/04/2010 15

Sea-truthing onboard Limanda

Kratzer et al, 2008

MERIS Kd490 image (300 m resolution) algorithm base on MERIS channels 3 and 6

Testing Kd 490 algorithm derived from MERIS against sea-truthing data

Kd490 (MERIS FR 19/08/2002, ch3_ch6_TACCS)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10 20 30 40 50 60

Distance to sewage outlet, km

mod

elle

d K

d490

, m-1

Measured Kd(490)

MERIS FR transect

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 5 10 15 20 25 30 35 40 45 50 55 60

Horizontal distance from sewage outlet, km

Kd(4

90) -

Kw(

490)

, m-1

inorganic SPM CDOM Chl-a

Himmerfjärden Open Baltic Sea

Attenuation model of the coastal zone

Contribution of each optical component to Kd (490) modelled from empirical data; polynomial decline of optical components from source (land) to sink (BY31) from Kratzer and Tett, 2009.

Seasonality of optical constituents- BY31 spring 1999

0.300

0.320

0.340

0.360

0.380

0.400

0.420

0.440

0.460

0.480

16/03/1999 24/03/1999 01/04/1999 09/04/1999 17/04/1999 25/04/1999

Spring 1999

G44

0, m

-1

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

16/03/1999 24/03/1999 01/04/1999 09/04/1999 17/04/1999 25/04/1999

Spring 1999

SPM

, g m

-3

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

16/03/1999 24/03/1999 01/04/1999 09/04/1999 17/04/1999 25/04/1999

Spring 1999

Pigm

ents

, µg

l-1

Chlorophyll aTotal carotenoids

Kratzer, unpublished

CDOMSPM

Inverted Secchi depth derived from MERIS data 19 August 2002 using a new algorithm (Kratzer et al., 2008)

Inverse Secchi depth (m-1) map as derived from the MERIS scene from 22 August2002, using a Secchi depth algorithm derived from sea-truthing data. Station 7a-7din Himmerfjärden are the same as H5-H2 in the national monitoring program.

NASA’s AERONET-OC stationGustaf Dalén (JRC & Strömbeck Consulting).

Aeronet-OC Gustav Dalén – algal bloom 2005

0

5

10

15

20

25

30

1-Jun-05 31-Jul-05 29-Sep-05

Chl

orop

hyll

(ug/

l)

B1BY31Model

Regular monitoring data(point sampling, monthly means)

Experimental monitoring(reflectance measurementson Gustav Dalén)

Courtesy Niklas Strömbeck

Match-ups NW Baltic Sea July 2008

Sampling stations and transects during Askö field campaign 2008 (see PINS on each RR scene). The conditions were very good for sea-truthing. The MERIS RR RGB composites show how patchy the waters become during good conditions in summer.

MERIS FR image 31 July 2008

Sea-truthing @ STN CI- CII-CIII

(30 km long transect) Gustav Dalén light house

close to CII

Testing ICOL processor

MERIS RR data, transect on 31 July 2008

near Gustaf Dahlén light house

both ICOL and non-ICOL processed and compared

to the sea-truthing data.

Processing: Christian Vinterhav; MEGS processing by Gerald Moore

Improvement of MERIS processors 2004-2009

Chlorophyll a - % error

-100

-50

0

50

100

150

200

250

300

2004

-IPF-

open

sea

2008

-FU

B-co

asta

l20

08-F

UB-

open

sea

2009

-MEG

S-op

en s

ea20

09-F

UB-

open

sea

2009

-MEG

S-co

asta

l20

09-M

EGS-

open

sea

2009

-FU

B IC

OL-

coas

tal

2009

-FU

B IC

OL-

open

sea

2009

-BPA

C IC

OL-

ope

n se

a

Optical variables as indicators of ecosystem state

• Humic substances (CDOM): terrestrial inputs of freshwater,

• suspended particulate inorganic matter (SPIM): land drainage and wind-stirring in shallow waters,

• and phytoplankton pigments: productive status of the pelagic ecosystem, influenced by anthropogenic nutrients from land

(Kratzer and Tett, 2009)

From research to applications -Development and quality assurance of an operational monitoring system

User consultation meetings• Vattenfall Power Consultant’s user group• EU FP6 SPICOSA stakeholder group• Himmerfjärden nitrogen study reference group

Direct contact to user, feed-back on what can be improved; need to develop user friendly productsProblem: many users are not familiar with the method- many perceive the topic as ‘too technical’

Temporal resolution in Hf area 2008

0

2

4

6

8

April May June July August September

Num

ber o

f mea

surm

ents

Monitoring

Remote.sensing

Regular cruises by monitoring vessels: 17Remote sensing by MERIS: 20Total set of chlorophyll a measurements: 37

Measurements between April and SeptemberMERIS passes by every 2-3 day

How representative is the satellite-derived chlorophyll-a product?

• In the inner fjord chlorophyll a concentrations are overestimated by 91 %.

• In open Sea and the outer basin of Himmerfjärden the Chlorophyll a concentrations is overestimated by 25 %.

(Kratzer and Vinterhav, 2010)• Correct for this in the chl-a estimates

Paired t-test Monitoring Df p-value

Remote sensing at B1 n.s 4 0.63

Remote sensing at B1 corrected n.s 4 0.59

Monthly average station B1

0

2

4

6

8

10

12

April May June July September

Chl

a c

onc

µg m

-1

MonitoringRemote sensingRemote sensing corrected

Therese Arredal Harvey, PhD student

Summary• MERIS provides us with a new tool to assess coastal systems from

space• Indicators for eutrophication, e.g. chl-a and Secchi depth, can be

derived from space reliably (if data is validated)• Chl-a concentrations from remote sensing not significantly different

from conventional monitoring data• Remote sensing data provides improve spatial and temporal

resolution• Together with Vattenfall Power Consultant we have developed a

user-friendly operational system• Continuation of operational system by joining GMES MarCoast

downstream services for 2010-2012• Aeronet-OC stations both for validation of satellite data and for

providing local continuous time series• Continuation of MERIS mission through Sentinel-3 (ocean colour

sensor OLCI and SST / (A)ATSR) through 2023

Possible future application – using MERIS data to test biogeochemical model

Sediment

Eukaryotic andpico-phytoplankton

N2-fixingcyanobacteria

DINDIP

Annualgrazing cycle

Surface water

Bottom water

Temp

(Permanent burial)

DIPDIN

Sedimentation

Vertical mixing

Advection

System boundary

Uptake

Mineralization

(Denitrification)

Mineralization

AdvectionSedimentation

Himmerfjärden SSA conceptual model – basin levelExternalinputs Sunlight

Eukaryotic andpico-phytoplankton

Advection

Advection

Annual light cycle

Temperature controlmortality

Annual nutrient release cycle

Mineralization

Sediment

Eukaryotic andpico-phytoplankton

N2-fixingcyanobacteria

DINDIP

Annualgrazing cycle

Surface water

Bottom water

Temp

(Permanent burial)

DIPDIN

Sedimentation

Vertical mixing

Advection

System boundary

Uptake

Mineralization

(Denitrification)

Mineralization

AdvectionSedimentation

Himmerfjärden SSA conceptual model – basin levelExternalinputs Sunlight

Eukaryotic andpico-phytoplankton

Advection

Advection

Annual light cycle

Temperature controlmortality

Annual nutrient release cycle

Mineralization

SPICOSA Conceptual model 2 Courtesy: Jakob Walve

Model input: Kd490 as proxy for light forcing

Model output: chl-a as proxy for phytoplankton biomass

MVT intercalibration work-shop July 2008, Askö

COASTCOLOUR Project from ESA

• Stockholm University and SYKE are champion users of the COASTCOLOUR project

• Round Robin of different algorithms for Baltic Sea remote sensing products (Sept 2010- Feb 2011)

• Validation of coast colour algorithm in the Baltic Sea

COASTCOLOUR sites

Thanks for listening!

SST bio-sensor, Kratzer Ltd. at Landsort Deep (BY31).

• Swedish National Space Board (SNSB): ‘Algorithm development and validation of MERIS data over optically complex waters’, 2008- 2010. Focus: Fundamental research.

• ESA/ESRIN: Technical Assistance for the validation of MERIS products in lake Vänern and coastal waters of the north-western Baltic Sea (Sweden), mid 2008- mid 2011. Focus: Validation of MERIS data and intercalibration of radiometers.

• Swedish Environmental Protection Agency (SEPA) and SYVAB, 2008-2009. Focus: Applications.

• NordForsk: NORDic network for AQUAtic REMote Sensing (NordAquaRemS), Sept 2008-Sept 2010; coordinator: S.Kratzer. Focus: Networking and PhD training.

• Participation in SPICOSA, and EU FP6 project on integrated coastal zone management (PI: Ragnar Elmgren): Focus: Academic training & developing remote sensing as diagnostic tool for integrated coastal zone management.

Acknowledgements

Close collaborations

• Petra Philipsson, Vattenfall Power Consultant AB, www.vattenkvalitet.nu

• Niklas Strömbeck, Strömbeck Consultant, Aeronet-OC Stationshttp://aeronet.gsfc.nasa.gov/new_web/ocean_color.html, (in collaboration with Giuseppe Zebordi, JRC, Ispra, Italy and NASA and ESA)

• Anu Reinart, Tartu Observatory, Estonia• Gerald Moore, Bio-Optika, UK• Carsten Brockmann, Brockmann Consult, Germany• Roland Doerffer, GKSS, Germany• Paul Tett, Napier University, UK

NORDic network for AQUAtic REMote Sensing (NordAquaRemS)

Coordinator:Associated Professor Susanne Kratzer,

Department of Systems Ecology, Stockholm UniversitySuse@ecology.su.se

Organizing committee:Prof. Matti Leppäranta, Helsinki University, Finland

Dr. Anu Reinart, Tartu Observatory, EstoniaDr. Piotr Kowalczuk, IOPAS, Poland

Dr. Are Folkstad, NIVA, Norway

Main Communication:E-mail & Skype conferences

Aims and Objectives- NordAquaRemS

(1) Advanced education for PhD students and young researchers, increased mobility and international experience

(2) Organize PhD and research training courses and seminars /workshops for students & researchers.

(3) Share expertise and define the most relevant common problems when applying remote sensing methods in Nordic conditions.

(4) Stimulate the application of new algorithms for the atmospheric correction for satellite images of Nordic water bodies (Baltic Sea and Nordic lakes)

Aims and Objectives- NordAquaRemS

(5) Share experience, data and instruments according to agreed conditions and rules.

(6) Improve exchange between scientists, users and service providers, including small/medium enterprises.

(7) Span the research and education across the whole seasonal cycle in Nordic waters, including the winter season.

(8) Create a Virtual Nordic Institute in Aquatic Remote Sensing.