Seasonal and spatial variability of Sea Surface

Seasonal and spatial variability of Sea Surface Temperature (SST) and Chlorophyll-a concentration using MODIS data

in East Kalimantan waters, Indonesia

IDHAM BIN KHALIL March, 2007

Seasonal and spatial variability of Sea Surface Temperature (SST) and Chlorophyll-a concentration using MODIS data

in East Kalimantan waters, Indonesia

by

Idham bin Khalil Thesis submitted to the International Institute for Geo-information Science and Earth Observation in partial fulfilment of the requirements for the degree of MSc Geo-information Science and Earth Observation for Environmental Modelling and Management, Specialisation: Environmental Modelling and Management. Thesis Assessment Board Chairman : Dr. Ir. Kees de Bie (ITC, The Netherlands) External Examiner : Prof. Peter Atkinson (University of Southampton, UK) First Supervisor : Dr. Ir. Chris Mannaerts (ITC, The Netherlands) Second Supervisor : Mr. Valentijn Venus (ITC, The Netherlands)

International Institute for Geo-Information Science and Earth Observation, Enschede, The Netherlands

Disclaimer This document describes work undertaken as part of a programme of study at the International Institute for Geo-information Science and Earth Observation. All views and opinions expressed therein remain the sole responsibility of the author, and do not necessarily represent those of the institute.

i

Abstract

Regular monitoring of near shore and open water parameters for marine management in East Kalimantan waters, Indonesia is still limited. The objective of this research is to determine and interpret the seasonal and spatial variability of sea surface temperature (SST) and chlorophyll-a concentration (Chl-a) in East Kalimantan waters. A standard MODIS SST split window algorithm and empirical chlorophyll-a OC-3M algorithm were used to generate the Level 2 MODIS SST and Chl-a images. From March 2005 to August 2006, the SST and Chl-a were retrieved from the sensor data in East Kalimantan coastal and open sea waters. In situ measurements from near shore waters were used to validate the MODIS Level 2 data. A comparison of MODIS with in situ values for SST and Chl-a shows promising result (RMSE=1.21 0C, Bias=-3.42, n=121 and RMSE=1.01mg.m-3, Bias=+2.45, n=75), although some anomalies were observed in the retrievals in both datasets. The analysis of seasonal variations indicates that there was low SST variability between wet and dry season. There was also low variability between SST values in near shore and open sea waters. However, for both seasons, open sea SST was paradoxically found to be warmer than the near shore waters. The Chl-a maps revealed low Chl-a variability between wet and dry season. Different value ranges in Chl-a were found between near shore waters (1.00-56.00 mg.m-3) and open waters (1.00-4.00 mg.m-3). The Chl-a values retrieved from MODIS for both seasons were higher in near shore water. The SST and Chl-a in near shore waters have a low positive interrelationship in wet season. During dry season, the relationship between these two variables varies from positive to negative. This study demonstrated that MODIS Level 2 data from Malaysia Ground Receiving Station (MGRS) can successfully be used to obtain SST and Chl-a in Southeast Asian coastal and open waters.

ii

Acknowledgements

I would like to express my deep and sincere gratitude to a number of people and organisations who became involved with this thesis, one way or another. GEM Erasmus Mundus consortium particularly Prof. Andrew Skidmore (ITC, The Netherlands), Prof. Petter Pilesjo (Lund University, Sweden), Prof. Katarzyna Dabrowska (University of Warsaw, Poland) and Prof. Peter Atkinson (University of Southampton, United Kingdom) who have always been very supportive throughout these 18 months. My first supervisor, Dr. Chris Mannaerts from Department of Water Resource, ITC for his understanding, encouraging and personal guidance has been a great value for me. Mr. Valentijn Venus, my second supervisor from Department of Natural Resources, ITC who has supported me throughout my thesis with his valuable comments and suggestions. I would have been lost without both of you. Malaysian Government particularly Malaysian Centre for Remote Sensing (MACRES) for providing the satellite imageries and giving me the opportunity to study in Europe. It was truly an eye-opening and rewarding experience of my life. I also wish to thank GEM students 2005, the most reliable friends one could have! Not forgotten, Malaysians in Enschede; Edna, Zaki, Fidah, Fauzi and Prof. Kamaruzaman Jusoff whose presence helped make the completion of my study possible. Ayahanda Khalil bin Yusof, Bonda Siti Rohani bt. Sh. Abdullah serta keluarga di Kuantan dan Pontian; your love, support, and patience have been a blessing throughout my study. I owe my loving thanks to my wife Nor Haslinda, my kids Muhammad Amnan, Muhrizah Humaira’ and Muhrizah Huda. Pengorbanan mereka tidak dapat digambarkan. Thesis ini didedikasikan istimewa buat mereka. Alhamdulillah. Idham Khalil Enschede

iii

Table of contents

1. Introduction ..................................................................................................... 1 1.1. Research background .............................................................................. 1 1.2. Research problem.................................................................................... 2 1.3. Research objectives ................................................................................. 4 1.4. Research questions .................................................................................. 4 1.5. Thesis Outline.......................................................................................... 5

2. Literature Review ............................................................................................ 7 2.1. Sea Surface Temperature (SST) .............................................................. 7

2.1.1. SST from space................................................................................... 8 2.1.2. In Situ SST Measurement ................................................................. 10

2.2. Chlorophyll-a concentration (Chl-a)..................................................... 11 2.2.1. Chlorophyll-a from space ................................................................. 11 2.2.2. In situ Chl-a measurement................................................................ 13

3. Study Area...................................................................................................... 15 3.1. Introduction ........................................................................................... 15 3.2. Climate .................................................................................................. 16 3.3. Oceanography........................................................................................ 16

3.3.1. Salinity.............................................................................................. 17 3.3.2. Dissolved Oxygen (DO) ................................................................... 17 3.3.3. Nitrate ............................................................................................... 17 3.3.4. Phospate............................................................................................ 17

4. Materials and Method ................................................................................... 19 4.1. MODIS Datasets.................................................................................... 19

4.1.1. Atmospheric correction .................................................................... 20 4.1.2. Cloud mask ....................................................................................... 20 4.1.3. Geometric correction ........................................................................ 20 4.1.4. Sea Surface Temperature.................................................................. 21 4.1.5. Chlorophyll-a concentration............................................................. 21 4.1.6. Correlation analysis between SST and Chl-a ................................... 22

4.2. Accuracy Assesment ............................................................................. 22 4.3. In situ measurement............................................................................... 23 4.4. Research approach................................................................................. 24

5. Results............................................................................................................. 25 5.1. Sea surface temperature......................................................................... 25 5.2. Chlorophyll-a concentration.................................................................. 29 5.3. Correlation between SST and Chl-a...................................................... 33

iv

6. Discussion .......................................................................................................35 6.1. Cloud cover in the study area ................................................................35 6.2. Sea Surface Temperature.......................................................................37

6.2.1. In situ SST versus MODIS Level 2 SST .......................................... 37 6.2.2. SST in the near shore waters............................................................. 38 6.2.3. SST in the open sea...........................................................................39

6.3. Chlorophyll-a concentrations ................................................................40 6.3.1. In situ Chl-a vs MODIS Level 2 Chl-a.............................................40 6.3.2. Chl-a in the near shore waters ..........................................................42 6.3.3. Chl-a in the open sea ........................................................................43 6.3.4. Striping lines in Chl-a images .......................................................... 43

6.4. Correlation between SST and Chl-a ......................................................44 7. Conclusions and Recommendations............................................................. 47

7.1. Conclussions..........................................................................................47 7.2. Recommendations .................................................................................47 References ............................................................................................................49 Appendices...........................................................................................................57 Appendix-1 : In situ stations, SST and Chl-a concentrations ..............................57 Appendix-1 : Con’t ..............................................................................................58 Appendix-1 : Con’t ..............................................................................................59 Appendix-1 : Con’t ..............................................................................................60 Appendix-2 : Bivarite Fit of IDWA SST by in situ SST......................................61 Appendix-3 : Bivarite Fit of Kriging SST by in situ SST ....................................62

v

List of figures

Figure 1-1: Berau Marine Protected Area (WWF, 2006)........................................... 3 Figure 2-1: NOAA AVHRR SST Global Map 6 – 8 May 2006 (NOAA, 2006) ....... 7 Figure 2-2: Plot of radiance from a blackbody against wavelength, with temperature as a variable (Lillesand, T. M., and R. W. Kiefer, 2004). .......................................... 8 Figure 2-3: Absorption coefficient and penetration depths of infrared waves (Wieliczka et al., 1989). ............................................................................................. 9 Figure 2-4: Idealized temperature profiles of the near-surface layer (10-m depth).. 10 Figure 3-1: Study area, East Kalimantan waters, Indonesia ..................................... 15 Figure 3-2: Map of Indonesian Throughflow (ITF) (Susanto et. al, 2006) .............. 16 Figure 4-1: SeaBass data points 2006 (SeaBass, 2006)............................................ 22 Figure 4-2: In situ sample stations............................................................................ 23 Figure 4-3: Simplified schema of the research approach ......................................... 24 Figure 5-1: Daily SST derived from MODIS; Wet Season...................................... 26 Figure 5-2: Daily SST derived from MODIS; Dry Season April – August 2005 .... 27 Figure 5-3: Daily SST derived from MODIS; Dry Season August 05–August 06 .. 28 Figure 5-4: Daily Chlorophyll-a concentration derived from MODIS; Wet Season 30 Figure 5-5: Daily Chl-a derived from MODIS; Dry Season, April – August 2005 . 31 Figure 5-6: Daily Chl-a derived from MODIS; Dry Season August 05–August 06 32 Figure 6-1: Percentage of SST pixels compared to cloud and land pixels ............... 35 Figure 6-2: Percentage of Chl-a pixels compared to cloud and land pixels ............ 36 Figure 6-3: Atmospheric Transmission Window (CRISP, 2006)............................. 36 Figure 6-4: In situ SST vs MODIS IDWA vs MODIS Kriging ............................... 37 Figure 6-5: Bivarate Fit of IDWA and Kriging SST................................................ 37 Figure 6-6: Minimum and maximum SST in near shore area .................................. 38 Figure 6-7: Minimum and maximum SST in open waters ....................................... 39 Figure 6-8: In situ Chl-a vs. MODIS IDWA vs. MODIS Kriging........................... 40 Figure 6-9: Bivarate Fit of IDWA and Kriging Chl-a.............................................. 41 Figure 6-10: Minimum and maximum Chl-a in near shore waters .......................... 42 Figure 6-11: Minimum and maximum Chl-a in open waters ................................... 43 Figure 6-12: SST and Chl-a correlation in East Kalimantan near shore waters...... 44 Figure 6-13: MODIS cloud mask for SST and Chl-a Level 2.................................. 45

vi

List of tables

Table 2-1: MODIS specifications (NASA, 2006) ....................................................11 Table 2-2: Detailed MODIS band specification (NASA, 2006)...............................12 Table 4-1: MODIS Level 2 data structure ................................................................19 Table 4-2: WGS_1984_UTM_Zone_50N parameters .............................................20 Table 4-3: Coefficients for the MODIS Band 31 and 32 SST retrieval ...................21 Table 5-1: SST of East Kalimantan Water from March 2005 to August 2006.........25 Table 5-2: Chl-a of East Kalimantan Water from March 2005 to August 2006 ......29 Table 5-3: SST and Chl-a relationship ....................................................................33 Table 6-1: Cold pixels percentage in near shore waters ...........................................39 Table 6-2: Cold pixels percentage in open waters....................................................40

vii

List of Abbreviations

ASEAN The Association of Southeast Asian Nations ATBD Algorithm Theoretical Basis Documents Chl-a Chlorophyll-a concentration CO2 Carbon dioxide CH4 Methane GeoTIFF Geographic Tagged Image File Format HDF Hierarchical Data Format ICMM Integrated coastal and marine management IDWA Inverse distance weighted average ITF Indonesian Throughflow IUCN International Union for the Conservation of Nature & Natural Resources K Kelvin LWIR long-wave Infrared mg.m-3 milligram per cubic meter mg/l milligram per liter MODIS Moderate Resolution Imaging Spectroradiometer MPA Marine protected area NO2 Nitrogen dioxide ppt part per thousand RMSE Root Mean Square Error SeaBass SeaWiFS Bio Optical Archive and Storage System SeaWIFS Sea-viewing Wide Field-of-view Sensor SST Sea Surface Temperature SWIR short-wave Infrared UNESCO United Nations Educational, Scientific and Cultural Organization UTM Universal Transverse Mercator System WGS84 World Geodetic System 1984 WWF World Wildlife Federation 0C degree Celcius µm micrometer

viii

SEASONAL AND SPATIAL VARIABILITY OF SEA SURFACE TEMPERATURE (SST) AND CHLOROPHYLL-a CONCENTRATION USING MODIS DATA

IN EAST KALIMANTAN WATERS, INDONESIA

1

1. Introduction

1.1. Research background

The Association of Southeast Asian Nations (ASEAN) consist of 10 countries, Brunei Darussalam, Cambodia, Indonesia, Lao PDR, Malaysia, Myanmar, the Philippines, Singapore, Thailand and Vietnam. For the past few decades, ASEAN has been successfully been a leader in economic development among developing countries. However in a development and environmental study, McDowell (1989) described that apart from the most promising economic growth in this region, it is suffering with the depletion of natural resources. Deforestation, erosion and over fishing mainly because of the rapid population growth have been exhibited as an indicator to an unbalanced development in ASEAN. As there is a close relationship between people and the environment, economic development and environmental issues should not be approached separately in this region. McDowell (1989) further described that in order for ASEAN to maintain its position as a new growing economic power, management of natural resource in this region is very crucial. In reaction to the above suggestion and parallel to The 1992 UN Rio de Janeiro – the ‘Earth Summit’, the ASEAN Summit 1997 formed a sustainable development framework for ASEAN countries. Known as ASEAN Vision 2020, the main aim was described as : “… a clean and green ASEAN with fully established mechanisms for sustainable development to ensure the protection of the region’s environment, the sustainability of its natural resources and the high quality of life of its peoples” (UP-MSI et al., 2002). In terms of the marine environment, ASEAN contributed to about 14% of the world’s marine fish production and has 30% of the world’s coral reefs: the most species diversity in the world (UP-MSI et al., 2002). In order to achieve sustainable fisheries, The ASEAN Vision 2002 has translated into 2001 Plan of Action on Sustainable Fisheries for Food Security for the ASEAN Region. One major step in this plan is to establish marine protected areas (MPAs). The International Union for the Conservation of Nature and Natural Resources (IUCN) defines marine protected areas (MPA) as “an area of sea especially dedicated to the protection and maintenance of biological diversity, and of natural and associated cultural resources,



2

and managed through legal or other effective means” (IUCN, 1994). In lay man terms, marine protected areas could be described as an alternative way to protect marine fauna and flora. As more marine protected areas have been established in the ASEAN region, the challenge to continuously provide decision makers with sufficient and reliable data is consider as a major constraint in MPAs management. With limited budget, a cost estimated at USD 1,584.00 per km2 per year for collecting data and maintaining MPAs (Balmford et al. 2004) is consider very high for developing countries. The rapid development in remote sensing technology and its capability to provide two-dimensional synoptic view offers an alternative to help in these issues. While there are always arguments on the ‘surface’ monitoring of ocean, there are positive tendencies in ecosystem–based precautionary management to rely on surrogate’s data (Vanderklift and Ward, 2000) e.g geophysical features (Ray, 1997) because of the complexity of ecosystems and lack of data. The reliability of the ocean surface ‘skin’ information has also been proven by Macintyre (1977) who stated that the first millimetres below the surface represents the top half of the ocean. In this study, the capability of geophysical product from Moderate Resolution Imaging Spectroradiometer (MODIS) which is Level 2 sea surface temperature (SST) and chlorophyll-a concentration (Chl-a) were assessed in providing ocean parameters for Indonesian waters from May 2005 – August 2006 for a better marine protected areas management.

1.2. Research problem

This study focuses on marine resources depletion in Indonesian marine environment. The study area, East Kalimantan waters which consist of Berau Marine Protected Area as shown in Figure 1-1 is well known as one of the most beautiful area for coral reef in the world. With 507 species of hard coral, Berau coral biodiversity is considered as the second highest in Indonesia and the third in the world. Among 120 potential outstanding universal values, Derawan islands which are located in Berau waters have been classified as one of the top seven (UNESCO, 2002). Several studies had shown that the total fish catch from Berau waters have been declining for the past ten years (Ismuranty, 2003) due to the depletion of the coral reefs. Less coral reef areas means less fish stock as coral reefs area are breeding

Chapter 1 - Introduction

3

grounds for fish. While the caused of coral depletion is well known coming from inland activities e.g deforestration and sea temperature changes e.g El-Nino 1997-1998, continuous monitoring and studies for better marine management is very limited (Burke et al. 2002) .

Figure 1-1: Berau Marine Protected Area (WWF, 2006). In order to maintain the sustainable marine resources in the area, the Indonesian government under Berau Regent Regulation No. 31, 2005 has declared Berau District’s coastal and marine area as a marine conservation area (WWF, 2006). Integrated coastal and marine management (ICMM) of this area requires mapping of marine resources and monitor basic changes (e.g SST, chlorophyll-a concentration etc.) as revealed by Dahuri (2000); Erdmann and Mossa (1990). Considering the huge area of 1.2 million hectares of the Berau Marine Protected Areas (MPA), this study used satellite remote sensing imageries as to provide such oceanographic datasets. Previous research has proved that continuous monitoring of SST from space can act as early warning to detect any SST anomalies that could harm the coral reef such as coral bleaching events (Liu et al., 2003). In day to day operations, spatial information on SST and Chl-a distribution has been used widely, e.g. in coral reef management (Udy et al., 2005), fishery forecast (Solanki et al., 2003; Solanki et al., 2005) or as a fundamental data source for coastal management planning (Mumby et al., 1999).



4

1.3. Research objectives

The main objective of this research is to determine and interpret the seasonal and spatial variability of SST and Chl-a in the study area. The specific objectives of this study are:

i. To map the spatial distribution of SST in East Kalimantan waters using MODIS data during wet (November – March) and dry season (April – October).

ii. To map the spatial distribution of Chl-a in East Kalimantan waters using MODIS data during wet and dry season.

iii. To determine the relationship between SST and Chl-a during wet and dry season in open and near shore waters.

1.4. Research questions

The research will try to answer the following questions:

i. Are there any major differences in SST distribution between wet and dry season?

ii. Are there any major differences in SST distribution in open sea and near shore?

iii. Are there any major differences in Chl-a distribution between wet and dry season?

iv. Are there any major differences in Chl-a distribution in open sea and near shore?

v. What kind of relationship exists between SST and Chl-a distribution in Kalimantan during wet and dry season in open waters, and

vi. What kind of relationship exists between SST and Chl-a distribution in Kalimantan during wet and dry season in near shore waters?

Chapter 1 - Introduction

5

1.5. Thesis Outline

This thesis contains seven chapters. The first chapter contains the general overview of the research. Research problems were clearly defined in section two. Research objectives and research questions will then explained in section three and four. Chapter two describes the overall description of the study area. Chapter three reports the previous studies in the research field. Major concepts and definition has been reported in 2 sections. The first section covers the description of detection of sea surface temperature followed by the literature of chlorophyll-a in second section. Chapter 4 presents a comprehensive methodology used in this study including the relevant algorithms and coefficients. Chapter 5 presents results for the SST and Chl-a distribution in East Kalimantan waters during dry and wet season. The result also included the variation between the near shore water and open sea. Chapter 6 presents the discussion for the SST and Chl-a distribution in East Kalimantan waters during dry and wet season. The variation of minimum and maximum values between near shore and open sea also will be discussed. Furthermore, the correlation between SST and Chl-a in this region was also explained. Finally, Chapter 7 presents the conclusions derived from the analysis and suggestions for future works.



6



7

2. Literature Review

2.1. Sea Surface Temperature (SST)

Historically, measurements of SST have been made in situ, from ships and buoys. MetOffice (2006) described that such in situ observations has two disadvantages. Firstly, in situ measurement offer very limited spatial coverage. Secondly, it enlarges the margin of error between one measurement to the next. These two issues could be resolved by satellite observation throughout the globe. McClain et al. (1985; 1989) stated that the retrieval of SST from satellite measurements offers the advantage of global coverage (Figure 2-1) and has been performed routinely since 1981 to a point accuracy of ~ 0.5 K.

Figure 2-1: NOAA AVHRR SST Global Map 6 – 8 May 2006 (NOAA, 2006) Marine researchers found that the measurement of SST is essential to understand how our ocean behaves; understand its relationship with ocean flora and fauna and not least to understand the global climate. Operationally, SST information has been used as a boundary layer input and an assimilation data set for atmospheric circulation/forecast models (Emery et al., 2001). Alexander and Scott (2002) stated that measurements of SST are extremely useful for applications such as monitoring



8

large-scale anomalies e.g El Nino and trends possibly associated with global climate change (e.g. Lawrence et al., 2004).

2.1.1. SST from space

According to the Planck Law, each warm object on Earth emitted certain amount of thermal radiation at a particular wavelength depending on its temperature (Rybicki et al., 1979). For a much clearer picture, the relationship between Earth’s surface, its brightness temperature and the spectral radiance can be understood by referring to Figure 2-2. From the graph, it was clearly shown that for Earth, the spectral radiance peaks at a wavelength around 10 µm.

Figure 2-2: Plot of radiance from a blackbody against wavelength, with temperature

as a variable (Lillesand, T. M., and R. W. Kiefer, 2004).

This is the basic concept why most of the infrared satellite sensors used band 10 µm to detect infrared radiation for SST detection. As there will be never 100% or ‘clear sky’ condition (without atmospheric disturbance), one channel measurement will never be a success. Water vapor, CO2, CH4, NO2 and aerosols are the major constituents that determine the atmospheric disturbance (Minnett, 1990). McMillin

Chapter 2 – Literature Review

9

(1975) and Barton (1995) described that most satellite SST algorithms are derived from the measurement using two channels. Two commonly use wavelengths range are the long-wave Infrared (LWIR; 10.00-12.5 µm) and the short-wave IR (SWIR; 3.7-4.2 µm) (Brown et al., 2005). The absorption for each wavelength is different. These differences were then used to correct the atmospheric disturbance before the sea surface temperature can be obtained. Apart from thermal infrared remote sensing, SST can also be detected using microwave sensors. However due to lower signal strength of the Earth's Planck radiation curve in the microwave region (approximately in the range of 30 cm to 1 mm), accuracy and resolution is poorer for SST derived from passive microwave measurements. Bear in mind that, although it is possible to obtain SST reading from thermal infrared sensors, the maximum sea depth that infrared wavelength can penetrate is around 500 µm (Wieliczka et al., 1989). In other words, if our interest is in getting the sea surface temperature more than that depth, we might end up with such a big uncertainty. In his research, Wieliczka et al., (1989) has plot a detailed relationship between infrared optical constants and clear water as shown in Figure 2-3.

Figure 2-3: Absorption coefficient and penetration depths of infrared waves

(Wieliczka et al., 1989).



10

2.1.2. In Situ SST Measurement

While the understanding of SST measurement will become a dominant knowledge, we could not admit the importance of the in situ SST measurement. There are several important terms that have been used to describe in situ SST. Donlon et al. (2002) categorized the vertical structure of SST into four major categories /definitions as shown in Figure 2-4. The first category is interface SST, SSTint. Interface SST is the temperature at the very top of the ocean. To be more specific, it is the temperature at the air-sea thin layers. The second category is the skin SST, SSTskin. By definition, the SSTskin is the temperature that can be measure up to 500 µm. Just below SSTskin, we have the layer of ocean where heat transfer process is actively happened. SST from this layer is called subskin SST, SSTsubskin. The final in situ SST categories is the subsurface SST, SSTdepth. In layman term, SSTdepth is also called ‘bulk’ SST.

Figure 2-4: Idealized temperature profiles of the near-surface layer (10-m depth) of the ocean (Donlon e. al., 2002).


11

2.2. Chlorophyll-a concentration (Chl-a)

Similar to SST, the capability of satellite remote sensing in providing global coverage snapshot has been an advantage to the measurement of Chl-a in the marine environment. In marine remote sensing, Chl-a has been used a proxy to the existence of phytoplankton. Butler et al. (1972) reported that the Chl-a above 0.2 mg/l indicate the presence of sufficient fish food to sustain a viable commercial fishery. In general term, chlorophyll-a can be described as a vital pigment for photosynthesis in phytoplankton. Phytoplankton are marine microscopic plants which well known as the base of marine food chain. A detailed definition has been described by Water Resource Disciplines USGS (2005) as “Phytoplankton is the plant part of the plankton. They usually are microscopic, and their movement is subject to the water currents. Phytoplankton growth is dependent upon solar radiation and nutrient substances. Because they are able to incorporate as well as release materials to the surrounding water, the phytoplankton have a profound effect upon the quality of the water. They are the primary food producers in the aquatic environment and commonly are known as algae”.

2.2.1. Chlorophyll-a from space

The earliest sensor that has been used for mapping and measuring chlorophyll-a concentration was Coastal Zone Color Scanner (CZCS) (Mitchell, 1994), Feldman et al. (1989). Realising the importance of synoptic global observation of marine parameters, several ocean color sensors than were launched,; Sea-viewing Wide Field-of-view Sensor (SeaWIFS) – July 1997; Moderate Resolution Imaging Spectroradiometer AM (MODIS AM) also known as MODIS Terra in December 1999 and MODIS PM or MODIS Aqua - May 2002 (NASA, 2006) . The detail specifications of the MODIS sensors are shown in Tables 2-1 and 2-2.

Table 2-1: MODIS specifications (NASA, 2006)

Parameters Details

1. Orbit 705 km, 10:30 a.m. descending node (Terra) or 1:30 p.m ascending node (Aqua), sun-synchronous, near-polar, circular

2. Swath Dimensions 2330 km (cross track) by 10 km (along track at nadir)

3. Spatial Resolution 250 m (bands 1-2) 500 m (bands 3-7) 1000 m (bands 8-36)



12

Table 2-2: Detailed MODIS band specification (NASA, 2006)

Primary Use Band Bandwidth Primary Use Band Bandwidth 1 620 - 670 20 3.660 -

3.840 Land/Cloud/Aerosols Boundaries 2 841 - 876 21 3.929 -

3.989

3 459 - 479 22 3.929 - 3.989

4 545 - 565

Surface/ Cloud Temperature

23 4.020 - 4.080

5 1230 - 1250 24 4.433 -

4.498

6 1628 - 1652

Atmospheric Temperature

25 4.482 - 4.549

Land/Cloud/Aerosols Properties

7 2105 - 2155 26 1.360 -

1.390

8 405 - 420 27 6.535 - 6.895

9 438 - 448

Cirrus Clouds Water Vapor

28 7.175 - 7.475

10 483 - 493 Cloud Properties 29 8.400 -

8.700

11 526 - 536 Ozone 30 9.580 - 9.880

12 546 - 556 31 10.780 - 11.280

13 662 - 672

Surface/ Cloud Temperature 32 11.770 -

12.270

14 673 - 683 33 13.185 - 13.485

15 743 - 753 34 13.485 - 13.785

Ocean Color/ Phytoplankton/ Biogeochemistry

16 862 - 877 35 13.785 - 14.085

17 890 - 920 36 14.085 - 14.385

18 931 - 941

Atmospheric Water Vapor

19 915 - 965

Cloud Top Altitude

Notes : Bands 1 to 19 are in nm; Bands 20 to 36 are in µm


13

The main reason Chl-a being used as an indicator for phytoplankton existence is because it is the most common pigment in most of marine phytoplankton (Li et al., 2002). The spectral characteristic of this pigment which located between blue–green wavelengths of light spectrum has made it detectable from optical remote sensing. Other pigments that can be also found in plants are Carotene - an orange pigment, Xanthophyll- a yellow pigment, Chlorophyll b - a yellow-green pigment and Phaeophytin - a gray pigment. In other literature, chlorophyll pigment in phytoplankton can also be categorized as chlorophyll a, b, c and carotenoids (Doerffer et al., 1999). To date, most Chl-a detection methods from satellite sensors used band-ratio algorithms (blue-green band reflectance ratio). In general, phytoplankton absorbs blue light more strongly than green light. Thus, the estimation of the relative amounts of phytoplankton can be achieved by measuring the amounts of light leaving the ocean waters at those two bands (blue and green).

2.2.2. In situ Chl-a measurement

The normal practice for in situ Chl-a measurement is by direct water sampling (NRMA, 2006). Filter papers with the size of 0.45 μ normally were used during the filtering process. The concentration of chlorophyll-a measured in µg/l then determined using spectrophotometer. With the current technology, it is now possible to directly measure the chlorophyll from water column using fluorescence.



14



15

3. Study Area

3.1. Introduction

The study area located within the rectangular region 00 45’ 2.93”N to 40 39’ 43.86”N and 1170 11’ 22.38”E to 1200 34’ 24.21”E encompass an area of about 228,254 km2 (Figure 3-1).

Figure 6 : Study area : East Kalimantan waters

Figure 3-1: Study area, East Kalimantan waters, Indonesia



16

In order achieve the objective of understanding the variability of SST and chlorophyll-a between near shore and open waters, Kalimantan waters has been subdivided into Open sea area: 17,771 km2 and near shore area: 13,046.6 km2 which is actually the Berau Marine Protected Area. The open sea was defined as the area more than 24 nautical miles (44 km) from the shoreline and near shore area is the area within the 24 nautical miles.

3.2. Climate

Like other parts of Indonesian waters, Kalimantan waters is affected by the Asia-Australia (AA) monsoon; southeast monsoon (April-October) and northwest monsoon (November- Mac), (Susanto et. al, 2006) . The southeast monsoon brings the warm and dry air from Australia into Kalimantan waters whereas the northwest monsoon is associated with warm and moist air. Nurlidasari (2004) stated that there are not many differences in SST values between coral reef area and the area close to the Berau river. SST in the coral reef area is reported to have values ranges from 29.50C to 300C while 29.50C to 30.50C for the second one. The highest wind speeds in this area were recorded during July and August while the lowest wind speed is in October and November.

3.3. Oceanography

Kalimantan waters play an important role in the transfer of heat and fresh water in the region because it is located in the Indonesian Throughflow (ITF) as shown in Figure 3-2.

Figure 3-2: Map of Indonesian Throughflow (ITF) (Susanto et. al, 2006)

Chapter 3 – Study Area

17

Notes : i. The solid black arrows - North Pacific thermocline water; ii. The black dashed arrows - South Pacific lower thermocline water. iii. The dashed arrows (green and blue) - seasonally reversed surface water flow due to the Asia-Australia monsoon. Wiryaman et al. (2004) reported a detailed summary of physical oceanographic in East Kalimantan waters as follows:

3.3.1. Salinity

The average surface salinity in East Kalimantan for open sea is 33.5 ppt. However, as approaching the river mouth, the salinity decreases (32.5 – 33.0 ppt). For salinity at 100 meter depth, the value range from 34.0-34.5 ppt for open sea and 33.5 ppt for river mouth area.

3.3.2. Dissolved Oxygen (DO)

Dissolved oxygen (DO) could be explained as the oxygen gas that is dissolved in water. Most of DO exist in marine environment produced by phytoplankton during photosynthesis process. In East Kalimantan waters, the average value for coastal and open waters is 2.5 to 4.5 mg/l.

3.3.3. Nitrate

The open sea and coastal water surface nitrate not vary much (0.4 to 1.8 mg/l). The 100 meter depth nitrate however show slightly different pattern. For near shore the value is 0 to 1.2 mg/l while > 1.2 mg/l for open sea.

3.3.4. Phospate

The average 100 meter depth phosphate in East Kalimantan for open sea is 1.2 to 2.4 mg/l and 0 to 1.2 mg/l for river mouth area.



18



19

4. Materials and Method

4.1. MODIS Datasets

The primary remotely sensed data used in this study were obtained from Malaysia Ground Receiving Station (MGRS) which has been acquiring MODIS data since 2nd February 2005. TeraScan software was used to process MODIS raw data into Level 2 MODIS with a spatial resolution of 1 km. The initial MODIS Hierarchical Data Format (HDF) was exported into Geographic Tagged Image File Format (GeoTIFF) for further analysis in ArcGis. For the purpose of this study, only Channel 1 of SST MODIS Level 2 was used while Channels 5 and Channel 10 for Chl-a analysis. Channel 10 was used for the cloud masking. The detail specification of the 32 byte GeoTIFF files are shown in Table 4-1 below:

Table 4-1: MODIS Level 2 data structure

SST MODIS Level 2 GeoTIFF Channels Parameters Channel 1 SST Channel 2 SST4

Chl-a MODIS Level 2 GeoTIFF Channels Parameters Channel 1 CZCS_pigment Channel 2 : K_490 Channel 3: L2_flags Channel 4 : calcite_conc Channel 5 chlor_MODIS mgm-3 Channel 6 chlor_MODIS_log mgm-3 Channel 7 chlor_fluor_base Channel 8 chlor_fluor_effic Channel 9 chlor_fluor_ht Channel 10 cldmsk_flags Channel 11 cocco_conc_detach Channel 12 cocco_pigmnt_conc Channel 13 common_flags Channel 14 phycoeryth_conc Channel 15 phycou_conc Channel 16 pigment_c1_total Channel 17 quality Channel 18 susp_solids_conc



20

4.1.1. Atmospheric correction

The atmospheric corrections applied to MODIS raw data can be categorised into two components. The first component uses the atmospheric correction algorithm described by Gordon and Voss (2004) in Algorithm Theoretical Basis Documents (ATBD) 18 version 5. The output of the algorithm is the normalized water-leaving reflectance which then be used for determinations of Chl-a as shown below:

[ρω(λ)]N= ρω(λ)/t*(θ0, λ) where; ρω = water leaving reflectance t* = diffuse transmittance θ0 = solar zenith angle λ = MODIS band wavelength The second component of atmospheric correction which is for SST data has been taken into account in the SST split window algorithm (Brown and Minnet, 1999). See SST algorithm in section 4.1.4 for further details. All atmospheric correction procedures were readily built into the TeraScan software processing module.

4.1.2. Cloud mask

A cloud mask with a spatial resolution of 1 km used in this study was generated by MODIS algorithm. A total of 17 to 19 spectral bands ranging from 0.55-13.93 µm were considered in the process of developing the mask as described by Baum and Platnick (2006) and Kathleen (2006).

4.1.3. Geometric correction

Shoreline vector and MODIS dataset were geo-referenced to the same World Geodetic System 1984 (WGS84) Universal Transverse Mercator (UTM) Projection System in ArcGIS software. The detail parameters are listed in Table 4-2 below.

Table 4-2: WGS_1984_UTM_Zone_50N parameters

Projection : Transverse_Mercator (WGS84_UTM_Zone_50N)

False_Easting: 500000.000000 False_Northing: 10000000.000000 Central_Meridian: 117.000000 Scale_Factor: 0.999600 Latitude_Of_Origin: 0.000000

Chapter 4 – Materials and Method

21

4.1.4. Sea Surface Temperature

A total of 21 scenes of SST were processed using split window technique based on ATBD-MOD25 (Brown and Minnet, 1999). The equation is given as follows:

SST = C1 + C2 * T31 + C3*T3132 +C4(sec(θ)-1)* T3132

where; T31 is the band 31 brightness temperature (BT) T3132 is (Band32 - Band31) BT difference θ is the satellite zenith angle Ci coefficients listed in Table 4-3

Table 4-3: Coefficients for the MODIS Band 31 and 32 SST retrieval algorithm ( Brown and Minnet, 1999)

By using the above coefficients, Brown and Minnet found that the SST result has a predicted RMS error of 0.337K with a bias near 0.

4.1.5. Chlorophyll-a concentration

Chlorophyll-a data were processed based on ATBD MOD19, Carder et al. (2003) using empirical chlorophyll-a OC3M algorithm as follows:

Ca = 100.283 – 2.753R + 1.457 R 2 + 0.659R 4 ;

where R = log10

Ca : Chlorophyll-a concentration (milligram per cubic metre, mg.m-3) Rrs : remote sensing reflectance

R : blue-green band ration [dimensionless]

Coefficients

T30 – T31 <= 0.7 T30 – T31 > 0.7 C1 1.228552 1.692521 C2 0.9576555 0.9558419 C3 0.1182196 0.0873754 C4 1.774631 1.199584

⎥⎦

⎤⎢⎣

⎡ >

rs551

rs488 rs443

RRR



22

4.1.6. Correlation analysis between SST and Chl-a

The SST and chlorophyll-a images which has the same spatial resolution (1 km) and projection system were brought into ArcGis. Due to ArcGis limitation in performing raster to raster correlation analysis, the raster layers were converted into point features. Then, an extraction of x-y coordinates were performed to the point features in order to get the x-y coordinates. This step was repeated for every SST and chlorophyll-a layers. Having the SST values, chlorophyll-a values together with their x-y coordinates for every pixel, a correlation analysis were carried out for each pixel.

4.2. Accuracy Assesment

This study does not call for the creation of a new local empirical MODIS algorithm. Research approach was conducted using the existing standard MODIS algorithm and compared with limited in situ data. The current accuracy of standard SST MODIS algorithm based on ATBD 25 is 0.337K (Brown and Minnet, 1999). In order to maintain the accuracy of their Chl-a algorithm, MODIS has a continuous validation programme using the in situ data from SeaWiFS Bio Optical Archive and Storage System (SeaBass) as shown in Figure 4-1. The MODIS Terra chlorophyll-a product gives root mean square error (RMSE) of 0.554 mg.m-3.

Figure 4-1: SeaBass data points 2006 (SeaBass, 2006)

Chapter 4 – Materials and Method

23

Near shore waters

Berau District

4.3. In situ measurement

In situ measurements were conducted from 11 to 29th August 2006. Measurements of SST were made using the Horiba U-10 water quality checker. The accuracy of the instrument is ± 0.3 0C as stated by Cole-Parmer (2006). SST was measured between 0-20 cm from the ocean surface (bulk temperature). Chlorophyll-a concentrations measurement were made using Aquafluor Handheld Fluorometer and Turbidimeter with a detection limit of 0.30 mg.m-3 (Turnerdesigns, 2006). The station locations for the sea truth are shown in Figure 4-2. Detailed location of in situ stations appear in Appendix-1

Figure 4-2: In situ sample stations



24

4.4. Research approach

A simplified flow chart of research approach is indicated in the following diagram.

Figure 4-3: Simplified schema of the research approach

IMAGE PROCESSING (MODIS)

DATA PREPARATION

SATELLITE IMAGES

LITERATURE REVIEW

Atmospheric Correction

Geometric Correction

Cloud covers % statistic determination

SST dry season

SST wet season

Chl-a dry season

Chl-a wet season

Correlation between SST & Chl-a in near shore

Correlation between SST & Chl-a in open sea

ACCURACY ASSESSMENT

REPORT WRITING

Land and cloud masking



25

5. Results

5.1. Sea surface temperature

A total of 21 scenes of SST were processed and the detailed results are shown in Table 5-1. The SST maps are shown in Figure 5-1, 5-2 and 5-3.

Table 5-1: SST of East Kalimantan Water from March 2005 to August 2006

Date Near Shore SST (0C) Open Sea SST (0C)

Min Max Min Max 3rd March 2005 18.02 27.66 22.99 27.66 4th March 2005 16.49 27.48 24.43 27.66 5th March 2005 19.64 27.66 21.93 27.66 30th April 2005a 8.23 27.66 22.85 27.65 30th April 2005b 9.57 27.66 14.17 27.67 7th June 2005 10.34 27.66 12.90 27.51 27th July 2005 23.02 27.77 27.55 27.55 30th July 2005 14.55 27.66 24.34 27.66 6th August 2005 20.81 27.66 26.06 27.66 11th August 2005 10.67 27.54 - - 12th August 2005 21.62 27.66 26.29 27.66 20th August 2005 19.34 27.66 - - 29th August 2005 14.11 27.66 - - 4th September 2005 11.31 27.66 - - 9th October 2005 20.84 27.63 18.42 27.64 5th May 2006 10.41 27.62 20.29 27.65 9th May 2006 20.16 27.64 21.96 27.66 18th May 2006 22.42 27.66 - - 10th July 2006 8.31 27.65 23.77 27.61 21th July 2006 18.61 27.67 25.13 27.65 7th August 2006 19.67 27.67 - -



26

Figure 5-1: Daily SST derived from MODIS; Wet Season

3rd March 2005 4th March 2005

5th March 2005

0

28

SST 0C25

Cloud and land mask

Chapter 5 - Results

27

Figure 5-2: Daily SST derived from MODIS; Dry Season April – August 2005

30th Apr 05a 30th Apr 05b 7th June 05

27th July 05 30th July 05 6th Aug 05

11th Aug 05 12th Aug 05 20th Aug 05

0

28

SST 0C 25

Cloud and land mask



28

Figure 5-3: Daily SST derived from MODIS; Dry Season August 05–August 06

9th Oct 05 4th Sept 0529th Aug 05

5th May 06 9th May 06 18th May 06

10th July 06 21st July 06 7th Aug 06

0

28

SST 0C25

Cloud and land mask

Chapter 5 - Results

29

5.2. Chlorophyll-a concentration

A total of 21 scenes of chlorophyll-a were processed and the detailed result appears in Table 5-2 below. The Chl-a maps are shown in Figure 5-4, 5-6 and 5-7.

Table 5-2: Chl-a of East Kalimantan Water from March 2005 to August 2006

Date

Near Shore Chl-a mg.m-3

Open Sea Chl-a mg.m-3

Min Max Min Max

3rd March 2005 1.00 56.00 1.00 2.00

4th March 2005 1.00 4.00 1.00 1.00

5th March 2005 1.00 20.00 1.00 2.00

30th April 2005a 1.00 56.00 1.00 2.00

30th April 2005b 1.00 52.00 1.00 1.00

7th June 2005 1.00 52.00 1.00 2.00

27th July 2005 1.00 35.00 1.00 3.00

30th July 2005 1.00 38.00 1.00 4.00

6th August 2005 1.00 57.00 1.00 2.00

11th August 2005 1.00 16.00 - -

12th August 2005 1.00 22.00 1.00 2.00

20th August 2005 1.00 55.00 1.00 2.00

29th August 2005 1.00 54.00 1.00 1.00

4th September 2005 1.00 29.00 1.00 1.00

9th October 2005 1.00 54.00 1.00 2.00

5th May 2006 1.00 41.00 1.00 3.00

9th May 2006 1.00 38.00 1.00 2.00

18th May 2006 1.00 38.00 1.00 3.00

10th July 2006 1.00 60.00 1.00 13.00

21st July 2006 1.00 59.00 1.00 2.00

7th August 2006 1.00 18.00 1.00 3.00



30

3rd March 05 4th March 05

5th March

Figure 5-4: Daily Chlorophyll-a concentration derived from MODIS; Wet Season

0 - 1

10 - 60

3 - 4 Chl-a mg.m-3

Cloud and land mask

Chapter 5 - Results

31

Figure 5-5: Daily Chl-a derived from MODIS; Dry Season, April – August 2005

30th Apr 05a 30th Apr 05b 7th June 05


11th Aug 05 12th Aug 05 20th Aug 05

0 - 1

10 - 60

3 - 4 Chl-a mg.m-3

Cloud and land mask



32

Figure 5-6: Daily Chl-a derived from MODIS; Dry Season August 05–August 06

9th Oct 05 4th Sept 0529th Aug 05

5th May 06 9th May 06 18th May 06


0 - 1

10 - 60

3 - 4 Chl-a mg.m-3

Cloud and land mask

Chapter 5 - Results

33

5.3. Correlation between SST and Chl-a

A correlation analysis between SST and Chl-a was performed for near shore and open sea. The detail result for near shore appears in Table 5-3. The correlation for open sea has not been reported due to no match up points between SST and Chl-a.

Table 5-3: SST and Chl-a relationship * Rs : Spearman correlation coefficients

Date Rs P value

n

3rd March 2005 0.1569 0.0130 68

4th March 2005 0.1403 0.0390 217

5th March 2005 0.1192 0.0155 412

30th April 2005a -0.0848 0.6558 30

30th April 2005b - - -

7th June 2005 -0.0026 0.9938 11

27th July 2005 - - -

30th July 2005 0.0876 0.7135 20

6th August 2005 -0.2189 0.0416 87

11th August 2005 - - -

12th August 2005 -0.5161 0.0042 29

20th August 2005 0.1472 0.054 172

29th August 2005 0.222 0.0176 114

4th September 2005 -0.0063 0.9808 17

9th October 2005 - - -

5th May 2006 -0.7364 0.0591 7

9th May 2006 - - -

18th May 2006 - - -

10th July 2006 - - -

21st July 2006 0.0671 0.1149 554

7th August 2006 -0.5536 0.0001 45



34



35

6. Discussion

This study demonstrated that MODIS Level 2 data from Malaysia Ground Receiving Station (MGRS) can successfully be used to obtain SST and Chl-a in East Kalimantan coastal waters and the Sulawesi open sea waters. This is the first attempt to validate in situ SST and Chl-a data with MODIS Level 2 data in East Kalimantan waters exclusively in Berau Marine Protected Area since it was gazetted in December 2005. This chapter will start with a section that considered as an important step for satellite SST and Chl-a measurement which is cloud cover percentage.

6.1. Cloud cover in the study area

Efforts to obtain SST and Chl-a information in Kalimantan waters during wet and dry season encountered major obstacle of cloud cover especially during wet season. Queries from MACRES Data Quicklook System for wet season result in only three scenes that were cloud free and suitable for further analysis. These three scenes obtained in Mac 2005 were used to represent the whole wet season period. Query result for dry season however ended up with better results with 18 suitable scenes. Analysis in ArcGis was then conducted to determine the detail percentage of cloud free pixels. As this study focuses in the ocean region, the land area has been added into the cloud statistic. The percentage of cloud covers and land area for SST and Chl-a MODIS Level 2 data are shown in Fig 6-1 and Fig 6-2.

SST pixels and cloud mask pixels

80

90

100

3rd Mac 05

5th Mac 05

30 April 05b

27th July 05

6th Aug 05

12th Aug 05

29th Aug 05

9th Oct 05

9th May 06

10th July 06

7th Aug 06

Dates

%

% SST pixels

% Cloud & Land pixels Figure 6-1: Percentage of SST pixels compared to cloud and land pixels



36

Chlorophyll-a pixels and cloud mask pixels

80

90

100

3rd Mac 05

5th Mac 05

30 April 05b

27th July 05

6th Aug 05

12th Aug 05

29th Aug 05

9th Oct 05

9th May 06

10th July 06

7th Aug 06

Dates

%

% Chl pixels

% Cloud & land pixels

Figure 6-2: Percentage of Chl-a pixels compared to cloud and land pixels Figure 6-1 and 6-2 clearly indicated that although it is possible to map the SST and Chl-a in the region using MODIS datasets, the chances of getting 100% ‘clear sky’ throughout the year is very slim. In general, Chl-a data had higher cloud free pixels compare to SST datasets. The average value for cloud free pixels in Chl-a datasets was 6.09 % and 2.17 % for SST datasets. One possible explanation is that the visible blue wavelengths (0.438 – 0.493 µm) used for generating Chl-a map is located in the bigger region of atmospheric transmission window (Figure 6-3) compared to the thermal infared wavelengths (10.780 - 12.270 µm) to generate SST values. CRISP (2006) defined the atmospheric transmission window as the wavelength region where there are less absorption and scattering by the atmospheric gases.

Figure 6-3: Atmospheric Transmission Window (CRISP, 2006)

Chapter 6 - Discussion

37

6.2. Sea Surface Temperature

6.2.1. In situ SST versus MODIS Level 2 SST

The August 2006 in situ SST values from 121 stations had been compared to the 7th August 2006 MODIS data as shown in Figure 6-4 below. We acknowledge there is a time difference between the satellite SST retrieval and the ground measurements, and are fully aware that validation should be carried out at time of satellite overpass, but as an indication, we performed the comparison and validation exercise.

SST : In situ vs MODIS IDWA vs MODIS Kriging

20.00

22.00

24.00

26.00

28.00

30.00

32.00

34.00

D1

TA2

TA9

TB4

TB12

TC6

TE1

TE8

TE15

TE22

TF5

TF12

TG7

TH2

TH9

TI7

TI14

TI21

In situ Stations

SST 0 C '

InsituIDWAKriging

Figure 6-4: In situ SST vs MODIS IDWA vs MODIS Kriging

Figure 6-4 clearly indicated that MODIS data tends to underestimate the SST values in East Kalimantan waters. Further analysis then found that root mean square error (RMSE) for in situ SST vs MODIS inverse distance weighted average (IDWA) was 1.35 0C and for in situ SST vs MODIS Kriging was 1.21 0C. The details are shown in Figure 6-5 and Appendix 2.

Kriging SST = 8.9002256 + 0.5842152 Insitu SST (n=121, r =0.48) IDWA SST = 10.163081 + 0.5374372 Insitu SST (n=121, r = 0.42)

Figure 6-5: Bivarate Fit of IDWA and Kriging SST



38

At this stage, one assumption that best explained why MODIS underestimated the SST in this region is because of cold cloud pixels, which affect neighbouring SST pixels. Further analysis than were conducted to the remaining 20 SST datasets. There categories of SST values; SST Maximum=Tmax, SST Minimum=Tmin and SST Average=Tavg were discussed. The discussion also focused on the differences between the SST values for open sea and near shore; during wet and dry season.

6.2.2. SST in the near shore waters

Result shows that Tmax for near shore waters during wet and dry season does not vary much. Tmax for the near shore during wet season is 27.66 0C and 27.77 0C during dry season. The difference is only 0.11 0C. Throughout the year, Tmax for this region can be concluded as relatively constant as shown in Figure 6-6 below.

Min and Max SST values:Near Shore (3rd March 2005 - 7th August 2006)

0.00

5.00

10.00

15.00

20.00

25.00

30.00

3rd Mac 05

4th Mac 05

5th Mac 05

30th Apr 05a

30th April 05b

7th June 05

27th July 05

30th July 05

6th Aug 05

11th Aug 05

12th Aug 05

20th Aug 05

29th Aug 05

4th Sept 05

9th Oct 05

5th May 06

9th May 06

18th May 06

10th July 06

21th July 06

7th Aug 06

Dates

SST 0C

SST MinSST Max

Figure 6-6: Minimum and maximum SST in near shore area

Tmin however shows a fluctuating pattern all year round. The minimum SST value recorded during wet season is 16.49 0C and 8.23 0C during dry season. The low SST values (pixels < 15 0C) could be unrealistic for SST in tropical sea SST such as East Kalimantan waters. Detail inspections to all datasets then revealed that the percentage of the low SST values pixels is very low. It ranged from 0.34% to 5.07% of the total SST pixels in the respective datasets. This confirmed that the low SST values are cold cloud pixels or edge cloud pixels and not the true SST. The detail counts of the cold pixels represented in Table 6-1.


39

Table 6-1: Cold pixels percentage in near shore waters

Date Number of pixels

< 15 0C Total

% 30th Apr 05a 7 138 5.07 30th April 05b 5 216 2.31 7th June 05 1 197 0.51 30th July 05 1 158 0.63 11th August 05 7 161 4.35 29th August 05 4 567 0.71 4th September 05 2 588 0.34 5th May 06 7 176 3.98 10th July 06 1 26 3.85

By setting the cut off point of valid SST to 15 0C, the average temperature, Tavg were recalculated for all the SST imageries. The new Tavg in near shore for wet season were 26.23 0C and 26.47 0C during dry season.

6.2.3. SST in the open sea

Similar to the near shore waters, Tmax for open sea during wet and dry season does not vary much. Tmax for the open sea during wet season is 27.66 0C and 27.67 0C during dry season. The average Tmax throughout the year is 27.64 0C. Tmin in the open sea however shows a unique pattern compared to the near shore waters. Figure 6-7 clearly indicated that Tmin for open sea is much higher than near shore waters.

Min and Max SST values: Open Sea (3rd March 2005 - 7th August 2006)

0.00

5.00

10.00

15.00

20.00

25.00

30.00

3rd Mac 05

4th Mac 05

5th Mac 2005

30th Apr 05a

30th April 05b

7th June 05

27th July 05

30th July 05

6th Aug 05

12th Aug 05

9th Oct 05

5th May 06

9th May 06

10th July 06

21th July 06

Dates

SST 0C

SST MinSST Max

Figure 6-7: Minimum and maximum SST in open waters



40

Analyses were also carried out to calculate the cold cloud pixels in the open sea area. Similar to near shore areas, result shows that the low SST values in open waters does not depict the overall pictures. The low SST values only exist in two SST images and the percentage is low as shown in Table 6-2 below.

Table 6-2: Cold pixels percentage in open waters

Number of pixels Date < 15 0C Total %

30th April 05b 7 424 1.65 7th June 05 2 20 10.00

By setting the cut off point of valid SST to 15 0C, the average temperature were recalculated from the SST imageries and it was found that Tavg for open sea during wet season were 26.69 0C and 26.49 0C for dry season.

6.3. Chlorophyll-a concentrations

6.3.1. In situ Chl-a vs MODIS Level 2 Chl-a

The August 2006 in situ Chl-a values from 121 stations were compared to the 7th August 2006 MODIS data as shown in Figure 6-8 below. As stated earlier, we agree that the time of satellite Chl-a retrieval does not coincide fully with the ground measurements (± 2 week difference). As an indicative exercise, we however conducted a comparison.

Log 10Chl-a : In situ vs MODIS IDWA vs MODIS Kriging

-2.5

-0.5

1.5

D1

TA2

TA9

TB4

TB12

TC6

TE1

TE8

TE15

TE22

TF5

TF12

TG7

TH2

TH9

TI7

TI14

TI21

In situ Stations

Log 1

0 Chl

-a m

g.m

-3

Log_InsituLog_IdwaLog_Kriging

Figure 6-8: In situ Chl-a vs. MODIS IDWA vs. MODIS Kriging


41

Figure 6-8 clearly indicated that Level 2 MODIS chlorophyll-a product has overestimated the Chl-a values in East Kalimantan waters. Similar to SST, further analysis has been carried out to find the RMSE values for this result. It was found that the RMSE for in situ Chl-a vs MODIS IDWA was 1.01 mg.m-3 and for in situ Chl-a vs MODIS Kriging was 1.22 mg.m-3. The cut off point for in situ Chl-a value is 0.1 mg.m-3 due to the accuracy of the instrument used. From 121 values, 46 Chl-a values were excluded. The details are shown in Figure 6-9 and Appendix 3.

IDWA CHL = 2.2111408 + 0.1802974 Insitu (n = 75, r = 0.05) Kriging CHL = 2.5006469 + 0.1483307 Insitu (n=75, r = 0.04)

Figure 6-9: Bivarate Fit of IDWA and Kriging Chl-a

The overestimation of low Chl-a values by MODIS also reported by Barbini et al. (2004). Barbini et al. (2004) further describe MODIS sensors tend to underestimate high Chl-a. This statement also supported by Blondeau-Patissier et al. (2004). In his finding, Blondeau-Patissier et al. (2004) stated that in order to gain more accurate result in detecting Chl-a using MODIS, effort should be focused on improving the atmospheric correction rather than more complicated Chl-a algorithms. This might explained very well why there were reasonable differences between MODIS Chl-a and in situ values in East Kalimantan waters. With none of the validation stations of SeaWiFS Bio Optical Archive and Storage System (SeaBass) located in the study area, this result was expected. However, the RMSE of 1.01 mg.m-3 gained can be a good starting point in understanding the variation of actual in situ values and MODIS sensors for East Kalimantan waters. A detailed analysis



42

was then carried out in order to understand the variability of Chl-a in near shore and open sea area during the wet and dry season.

6.3.2. Chl-a in the near shore waters

Result shows that there is a large variability of Chl-a in the near shore waters. No unique pattern exists between wet and dry season. The average Chl-a in near shore calculated from the imageries were 2.12 mg.m-3 during wet season and 3.46 mg.m-3 during dry season. The Chl-a range for wet season is between 1.00 to 56.00 mg.m-3 while it varies from 1.00 to 59.00 mg.m-3 during dry season. Our result shows huge differences when compared to previous Chl-a mapping in this area. Abu Daya (2004) in his study found that the range of Chl-a in near shore East Kalimantan waters ranges from 0 – 9 mg.m-3. Although it is true to say that it might be due to the different sensor used (SeaWiFS) and different acquisition dates but further research need to be carried out to compare these two sensors. Blondeau-Patissier et al. (2004) stated that SeaWiFs was more accurate (RMS = 0.24; n = 26) compared to MODIS Level 2 (RMS = 0.40; n = 26). However, as his study was in Northern European waters, the result might be different in tropical waters. One important advantage of MODIS Chl-a dataset due to its complex turbid water algorithm, is the ability to measure Chl-a in near shore waters. Barbini et al. (2004) stated that, SeaWifs atmospheric correction failed in some near shore areas. This study proved the capability of MODIS in measuring near shore Chl-a. The minimum and maximum Chl-a in near shore waters shown in Figure 6-10.

Min and Max Chl-a Values : Near Shore (3rd March 2005 - 7th August 2006)

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

3rd Mac 05

5th Mac 05

30th April 05b

27th July 05

6th Aug 05

12th Aug 05

29th Aug 05

9th Oct 05

9th May 06

10th July 06

7th Aug 06

Dates

mgm-3 '

Chl Min

Chl Max

Figure 6-10: Minimum and maximum Chl-a in near shore waters


43

6.3.3. Chl-a in the open sea

Result in Figure 6-11 shows that the value of Chl-a ranges from 1.00 to 13 mg.m-3 in the East Kalimantan open waters. Further inspection found that the highest concentration value (13.00 mg.m-3) which happened on 10th July 2005 was only from one pixel. Assuming this as noise, the new Chl-a range for open waters is 1.00 to 4.00 mg.m-3 which is considerably low. The average Chl-a calculated from the imageries were 1.00 mg.m-3 during wet season and 1.02 mg.m-3 during dry season.

Min and Max Chl-a Values : Open Sea (3rd March 2005 - 7th August 2006)

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

3rd Mac 05

5th Mac 05

30th April 05b

27th July 05

6th Aug 05

20th Aug 05

4th Sept 05

5th May 06

18th May 06

21th July 06

Dates

mgm-3 '

Chl MinChl Max

Figure 6-11: Minimum and maximum Chl-a in open waters The possible explanation for this small range might be best described by Arnone and Parsons (2005) which stated that in open seas, chlorophyll maximum layer normally located below the surface, out from satellite sensors view.

6.3.4. Striping lines in Chl-a images

In few Chl-a images as shown in Figure 5-5, 5-6 and 5-7, striping lines were very obvious. These lines were generated by the residual differences between MODIS Terra detectors during solar diffuser (SD) calibrations. Bryan (2007) reported that, NASA Ocean Biology Production Group (OBPG) is still trying to reduce the residual detector striping lines. Bryan (2007) further described that the responsivity of MODIS/Terra sensors were degraded up to 40% since launched (compared to 15% for Aqua). This issue needs to be taken into account seriously if MODIS Level 2 Chl-a data will be use for future study in this area.



44

6.4. Correlation between SST and Chl-a

Having the opportunity to obtain the SST and Chl-a at the same time and date is always a dream of marine researchers as it is valuable information for further application research e.g fish forecasting and modelling. Mansor et al. (2001) found that SST and Chl-a relationship are the most important variables to model the fish distribution in tropical regions. Solanki et al. (1998) and Solanki et al. (2003) revealed that there was negative/inverse correlation between SST and Chl-a. Solanki et al. (2003) then used the correlation of SST and Chl-a information to model the fish distribution in Arabian Sea. In contrast to his finding, this research however found that that there was no specific pattern for correlation between SST and Chl-a in East Kalimantan waters except for wet season. Figure 6-12 shows that, the SST and Chl-a correlation in wet season has positive relationship. Despite the relatively low Spearman correlation cooefficient (~ Rs = 0.10), the relationship was significant (p<0.05). In dry season the relationship varies from negative to positive throughout the dry months.

SST-Chl-a concentration correlation : Mac 2005 - August 2006 East Kalimantan waters

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

3rd Mac 05

4th Mac 05

5th Mac 05

30th Apr 05a

7th June 05

30th July 05

6th Aug 05

12th Aug 05

20th Aug 05

29th Aug 05

5th May 06

21th July 06

7th Aug 06

Dates

Rs

Spearman Rs

Figure 6-12: SST and Chl-a correlation in East Kalimantan near shore waters The Spearman non-parametric correlation test has been used due to the non normal distribution of the data. The analysis reported above was only for the near shore waters as lack of match up points in open waters. Although MODIS can provide SST and Chl-a at the same date and time, these dataset had one drawback; the cloud mask for SST and Chl-a is different as shown in Figure 6-13.


45

Figure 6-13: MODIS cloud mask for SST and Chl-a Level 2

Bryan Franz (2006) reported that for MODIS SST Level 2 product, the determination of invalid pixel values is based on how different the pixels value is from the reference SST (Reynolds). If the difference is huge then the pixels will be labelled as SSTWARN. If the SST algorithm could not calculate any values, the pixels will then be labelled as SSTFAIL. As the final SST pixels also took into account the night and day value, sun glint did not effect SST pixels. The Chl-a determination however is very sensitive to sun glint. This is the main reason why both datasets cannot have similar ‘cloud mask’ and analysis need to be carefully done to select only overlapping pixels for the correlation analysis. The limited overlapping pixels available in open sea of East Kalimantan need to be taken as an important consideration if fish forecasting model need to be run using MODIS Level 2 data in this waters.

SST 30th Apr 05 Chl-a 30th Apr 05



46



47

7. Conclusions and Recommendations

7.1. Conclussions

The seasonal and spatial variability of SST and Chl-a in East Kalimantan near shore and open waters, Indonesia were determined. The minimum, maximum and average values calculated from temporal datasets were examined and reported. In situ data were used to validate the accuracy of MODIS Level 2 SST and MODIS Level 2 Chl-a data. The major findings of the research revealed: SST maps of East Kalimantan waters from March 2005 to August 2006 show that there was low SST variability between wet and dry season. There was also low variability between SST values in near shore and open sea waters. However, for both seasons, open sea SST was found to be paradoxically slightly warmer than the near shore waters. Chl-a maps of East Kalimantan waters from March 2005 to August 2006 waters revealed that there was low Chl-a variability between wet and dry season. There was also low variability between Chl-a values in near shore and open sea waters. The Chl-a values for both seasons is higher in near shore waters. SST and Chl-a in near shore waters has low positive relationship in wet season. During dry season, the relationship between these two variables varies from positive to negative.

7.2. Recommendations

The work presented in this study has left some important questions still unanswered. Effort should be focused in several issues as discussed below if further research need to be conducted.



48

In section 6.2.1 and 6.3.1, this study had successfully compared the August 2006 MODIS Level 2 data with August 2006 in situ data although we admit the time difference between satellites overpass and ground sampling. There were 4 days difference between the MODIS acquisition date and the in situ campaign. In ideal case, it is necessary to validate the satellite imageries with in situ that coincide with the satellite acquisition time. The RMS error between MODIS and in situ value could be refined if this can be achieved. In section 6.2.2, this study had made the assumption that the cut off value for minimum SST value in East Kalimantan waters was 15 0C. This has raised two interesting questions. What is the best cut off SST value for East Kalimantan region? How would the changes in the cut off value affect the SST values for this region?. The unique positive relationship between SST and Chl-a in wet season was very interesting. This was somehow questionable because only three images were used to representing wet season due to the unavailability of satellite data (cloud cover). For future works, more datasets for the wet season should be obtained and analysed to proof the relationship.



49

References

Abu Daya, M.I. (2004), Coastal water quality monitoring with remote sensing in

East Kalimantan, Makassar strait, Indonesia. MSc. Thesis. International Institute for Geo-Information Science and Earth Observation (ITC).

Arnone R.A and Parsons A.R. (2005), “Real-time Use of Ocean Color Remote

Sensing for Coastal Monitoring,” chapter 14 in Remote Sensing of Aquatic Coastal Environments, ed. R.L. Miller, C.E. Del Castillo, and B.A. McKee, Springer, Netherlands

Alexander, M., Scott, J. (2002), The influence of ENSO on air–sea interaction in the

Atlantic. Geophysical Research Letters, 29(14). Balmford A., Pippa Gravestock, Neal Hockley, Colin J. McClean and Callum M.

Roberts, (2004), The worldwide costs of marine protected areas. Proceedings of the National Academy of Sciences of the United States of America, 101(26): 9694 - 9697.

Barbini R., F. Colao, L. De Dominicis, R. Fantoni, L. Fiorani, A. Palucciand and E.

S. Artamonov., (2004), Analysis Of Simultaneous Chlorophyll Measurements by Lidar, Fluorosensor, Modis And Seawifs. Int. J. Remote Sensing, 25(11):2095–2110

Barton, I. J., (1995), Satellite-derived sea surface temperatures: Current status. J.

Geophys. Res., 100: 8777–8790. Blondeau-Patissier, D., Tilstone, G. H., Marinez-Vicente, V., & Moore, G. F.

(2004). Comparison of bio-physical marine products from SeaWiFS, MODIS and a bio-optical model with in situ measurements from NorthernEuropean waters. Journal of Optics. A, Pure and Applied Optics, 6: 875–889.

Butler, M.J.A. et al., (1988) The application of remote sensing technology to marine

fisheries: an introductory manual. FAO Fish. Tech.Pap., (295):165 p.



50

Baum B.A. and Platnick S. (2006), Introduction to MODIS Cloud Products. Retrieved 12th October 2006 from http://climate.gsfc.nasa.gov/publications/ fulltext/Baum_and_Platnick_2006.pdf

Brown C.W, Connor L.N, Lillibridge J.L, Nalli N.R and Legeckis R.V (2005), “An

Introduction to satellite sensors, observations and techniques)” Chapter 2 in Remote Sensing of Aquatic Coastal Environments, ed. R.L. Miller, C.E. Del Castillo, and B.A. McKee, Springer, Netherlands

Brown, O.B. and Minnet P.J., (1999), MODIS Infrared Sea Surface Temperature

Algorithm. Algorithm Theoretical Basis Document (MOD25). Retrieved 21st November 2006 from http://modis.gsfc.nasa.gov/data/atbd/ atbd_mod17.pdf

Bryan Franz (2006), Implementation of SST Processing within the OBPG.

Retrieved 12th October 2006 from http://oceancolor.gsfc.nasa.gov/DOCS/ modis_sst/

Bryan Franz (2007), MODIS Terra Ocean Processing Status. Retrieved 12th

February 2007 from http://oceancolor.gsfc.nasa.gov/DOCS/modist_proces sing

Burke, L., Selig, E. and Spalding, M. (eds). (2002), Reefs at Risk in Southeast Asia.

World Resources Institute, Washington, D.C., 72 pp. Carder K.L., Chen F.R., Lee Z., Hawes H.K. and Cannizzaro J.P. (2003), MODIS

Ocean Science Team Algorithm Theoretical Basis Document (ATBD 19) Version 7 Retrieved 21st November 2006 from http://modis.gsfc.nasa.gov/ data/atbd/ atbd_mod19.pdf

CRISP (2006), Atmospheric effect. Retrieved 12th October 2006 from http

://www.crisp.nus.edu.sg/~research/tutorial/atmoseff.htm#windows Cole-Parmer (2006), Horiba U-10 Series Water Quality Meters–Technical

Appendix. Retrieved 12th October 2006 from http://www.Coleparmer .com/catalog/product_view.asp?sku=9959000#tech

References

51

Dahuri, R., and I. M. Dutton (2000), Integrated Coastal and Marine Management Enters a New Era in Indonesia. Coastal Zone Management Journal 1:11-16.

Doerffer, R.; Sfrensen, k. & Aiken, J. (1999), MERIS potential for coastal zones

applications. International Journal of Remote Sensing, 20(9): 1809-1818. Donlon, C. J., P. Minnett, C. Gentemann, T. J.Nightingale, I. J. Barton, B. Ward

and, J. Murray (2002), Towards Improved Validation of Satellite Sea Surface Skin Temperature Measurements for Climate Research, J. Climate, 15(4): 353-369

Emery, W. J., Castro, Sandra, Wick, G. A., Schluessel, Peter and Donlon, Craig.

(2001), Estimating Sea Surface Temperature from Infrared Satellite and In Situ Temperature Data. Bulletin of the American Meteorological Society 82: 2773-2785

Erdmann, M.V. and M.K. Moosa (1999), A new discovered home for “old

fourlegs”: The discovery of an Indonesian living coelacanth, Pesisir dan Lautan. Indonesian Journal of Coastal and Marine Resources. 2(1):12-20

Feldman, G., N. Kuring, C. Ng, W. Esaias, C.R. McClain, J. Elrod, N. Maynard, D.

Endres, R. Evans, J. Brown, S. Walsh, M. Carle, and G. Podesta. (1989), Ocean color, availability of the global data set. Eos, Trans., AGU 70:634-641.

Gordon H.R and Voss K.J (2004), MODIS Normalized Water-leaving Radiance

Algorithm Theoretical Basis Document, (MOD 18) Version 5 Retrieved 21st November 2006 from http://modis.gsfc.nasa.gov/data/atbd/ atbd_mod18.pdf

Ismuranty, C., (2003), Building the co-management for the conservation and

sustainable use of the Derawan Islands, East Kalimantan, Indonesia. Second International Tropical Marine Ecosystems Management Symposium (ITMEMS 2), Manilla, Philippines. March 24-27, 2003.

IUCN (1994), Guidelines for Protected Area Management Categories. CNPPA and

WCMC. Cambridge, UK, pp. 5-10.



52

Kathleen Strabala (2006), MODIS Cloud Mask User’s Guide. Retrieved 12th Oct

2006 from http://cimss.ssec.wisc.edu/modis1/pdf/Cmusersguide .pdf Lawrence, S. P., D. T. Llewellyn-Jones, and S. J. Smith (2004), The measurement

of climate change using data from the Advanced Very High Resolution and Along Track Scanning Radiometers, J. Geophys. Res., 109, C08017, doi:10.1029/2003JC002104.

Li, H.-P., G.-C. Gong and T.-M. Hsiung (2002), Phytoplankton pigment analysis by

HPLC and its application in algal community investigations. Bot. Bull. Acad. Sin 43: 283-290

Lillesand, T.M. and R.W.Kiefer. (2004), Remote Sensing and Image Interpretation,

Fifth edition. John Wiley & Sons. United States of America. pp. 8 Liu G., Alan E.Strong and Andwilliam Skirving (2003), Remote Sensing of Sea

Surface Temperatures During 2002 Barrier Reef Coral Bleaching. Eos, 84(15): 137-141

MacIntyre, F. (1977), The top millimeter of the ocean. In: Menard, H.W.(ed.),

Ocean Science, San Fransisco, W.H. Freeman, p.145-155) (first published in Scientific American, May, 1974)

McClain, E. P., W. G. Pichel, and C. C. Walton (1985), Comparative performance

of AVHRR-based multichannel sea surface temperatures, J. Geophys. Res., 90(11):578-601

McClain, E. P. (1989), Global sea surface temperatures and cloud clearing for

optical depth estimates, Int. J. Remote Sens., 10:763- 769 McDowell Mark A. (1989), Development and environment in ASEAN. Pacific

Affairs. 62(3): 307-329 McMillin, L. M., (1975), Estimation of sea surface temperatures from two infrared

window measurements with different absorption. J. Geophys. Res., 80: 5113–5117.

References

53

MetOffice (2006), Sea Surface Temperature Conversion. Retrieved 30th May 2006 from http://www.metoffice.com/research/nwp/satellite/infrared/sst/

Minnett, P.J. (1990), The regional optimisation of infrared measurements of sea

surface temperature from space. Journal of Geophysical Research 95: 13,497–13,510.

Mitchell, B.G. (1994), Coastal zone scanner retrospective. Journal of Geophysical

Research. 99: 7291-7292 Mumby, P.J., E.P. Green, A.J. Edwards anf C.D. Clark (1999), The cost-

effectiveness of remote sensing for tropical coastal resources assessment and management. Journal of Environmental Management. 55: 17-166

NASA (2006), MODIS Specification. Retrieved 12th January October 2007 from

http://aqua.nasa.gov/about/instrument_modis.php NOAA (2006), Global Sea Surface Temperature. Retrieved 12th January October

2007 from http://www.osdpd.noaa.gov/PSB/EPS/SST/data/FS_km5000.gif NRMA (2006), Chlorophyll-a. Retrieved 14th November 2006 from http://www.

nrm.gov.au/monitoring/indicators/estuarine/chlorophyll-a.html Nurlidiasari M. (2004), The application of Quickbird and multi-temporal Landsat

data for coral reef mapping. Case study : Derawan Island, East Kalimantan, Indonesia. MSc. Thesis. International Institute for Geo-Information Science and Earth Observation (ITC).

Ray G.C. (1997), Biogeography as a criterion for selection of marine and coastal

protected areas. In: M.P. Crosby, K. Geenen, D. Laffoley, C. Mondor & G. O’Sullivan (eds) Proceedings of the 2nd International Symposium and Workshop on Marine and Coastal Protected Areas: Integrating Science and Management, 1995. Silver Spring, MD, USA: NOAA, pp. 81–93.

Rybicki, G. and Lightman, A. P. (1979), "The Planck Spectrum." Radiative

Processes in Astrophysics. New York: Wiley-Interscience, pp. 3-4 and 20-23



54

SeaBass (2006), SeaBASS validation data summary Retrieved 30th October 2006 from http://seabass.gsfc.nasa.gov/cgi-bin/validation_cruises.cgi

Solanki, H. U., R. M. Dwivedi, S. R. Nayak V. S. Somvanshi, D. K. Gulati and S.

K. Pattnayak (2003), Fishery forecast using OCM chlorophyll concentration and AVHRR SST: validation results off Gujarat coast, India. Int. J. Remote Sensing, Vol. 24, No. 18, 3691–3699

Solanki, H.U., P.C. Mankodib, S.R. Nayaka and V.S. Somvanshi (2005), Evaluation

of remote-sensing-based potential fishing zones (PFZs) forecast methodology. Continental Shelf Research 25: 2163–2173

Solanki, H.U., Dwivedi, R. M., and Nayak, S. R., (1998), Relationship between IRS

MOS-B derived chlorophyll and NOAA AVHRR SST: a case study in the north west Arabian Sea. Proceedings of 2nd International Workshop on MOS-IRS and Ocean Colour, Institute of Space Sensor Technology, Berlin. pp 327-339

Susanto, R. D., T. S. Moore II, and J. Marra (2006), Ocean color variability in the

Indonesian Seas during the SeaWiFS era, Geochem. Geophys. Geosyst., 7, Q05021, doi:10.1029/2005GC001009.

Turnerdesigns (2006), Aquafluor Handheld Fluorometer and Turbidimeter –

Manual. Retrieved 12th January October 2007 from http://www.turner designs.com/t2/doc/brochures/aquafluor_ brochure.pdf

Udy, J., Gall, M., Longstaff, B., Moore, K., Roelfsema, C., Spooner, D.R. and

Albert, S. (2005), Water quality monitoring: a combined approach to investigate gradients of change in the Great Barrier Reef. Australian Marine Pollution Bulletin 51:224-238.

UNESCO (2002), The Hanoi Statement. [Online] http://whc.unesco.org/archive/

hanoi-statement02.pdf (accessed 10 July 2006) UP-MSI, ABC, ARCBC, DENR, ASEAN, (2002), Marine Protected Areas in

Southeast Asia. ASEAN Regional Centre for Biodiversity Conservation, Department of Environment and Natural Resources, Los Baños, Laguna, Philippines. WordShop Specialists Network, Inc. Philippines.

References

55

Vanderklift M. and Ward T. (2000), Using biological survey data when selecting

Marine Protected Areas: an operational framework and associated risks. Pacific Conservation Biology (6): 152–161.

Water Resources Data State, (2005), Definition of Terms. Retrieved 12th December

2006 from http://water.usgs.gov/ ADR_Defs_2005.pdf Wieliczka, D. M., S. Weng, and M. R. Querry (1989), Wedge shaped cell for highly

absorbent liquids: infrared optical constants of water, Applied Optic., 28: 1714-1719

Wiryawan, B., S.A. Stanley, I. Yulianto, H.A. Susanto. (2004), Derawan Islands

Profile, Berau Regency, East Kalimantan. The Nature Conservancy, and Pemerintah Kabupaten Berau. Tanjung Redeb - Kalimantan Timur, Indonesia. 71 pp.

WWF (2006), Berau Marine Protected Area. Retrieved 30th October 2006 from

http://assets.panda.org/ downloads/mpaberau_map.doc



57

Appendices

Appendix-1 : In situ stations, SST and Chl-a concentrations

Station East(X) North(Y) Chl-a

mg.m-3 SST 0C D1 638823.35 252255.86 0.00 29.80 D2 638968.63 252291.38 0.00 29.10 D3 639061.30 252326.95 0.00 28.30 D4 639102.19 252348.47 0.03 28.50 D5 639162.46 252360.31 0.05 28.40 D6 639230.26 252363.54 0.07 28.40

TA1 637376.10 250316.60 0.01 28.30 TA2 629185.50 238094.00 0.02 28.40 TA3 627205.50 237857.50 0.02 28.90 TA4 623541.20 236839.60 0.15 30.20 TA5 620012.70 236330.70 0.16 29.60 TA6 615568.00 235923.50 0.20 31.50 TA7 611496.60 235889.60 0.30 32.40 TA8 607696.60 235584.30 0.31 32.10 TA9 606373.40 237755.70 0.34 32.90

TA10 615362.70 243233.10 0.28 31.40 TA11 618854.50 246668.80 0.14 31.90 TA12 621780.10 251740.20 0.10 31.30 TB1 631849.90 250358.00 0.02 28.40 TB2 627712.70 248311.30 0.02 28.40 TB3 623240.70 245704.10 0.10 29.00 TB4 618986.00 243059.10 0.24 29.00 TB5 614885.20 241483.70 0.22 29.90 TB6 611226.20 240635.30 0.22 30.40 TB7 607428.40 240206.70 0.80 30.20 TB8 606150.20 238540.00 0.35 30.40 TB9 603819.40 239931.00 0.38 30.50

TB10 601510.50 243448.10 0.43 30.60 TB11 596694.80 242623.50 0.34 30.60 TB12 588686.80 242097.30 0.26 30.70 TB13 579982.20 241588.30 0.24 30.90 TB14 571842.40 243043.60 1.63 30.30 TB15 563327.60 241072.70 1.49 30.40 TB16 560910.10 239981.40 1.15 30.50



58

Appendix-1 : Con’t


mg.m-3 SST 0C TB17 554904.40 239804.80 0.70 30.40 TC1 612121.30 222504.30 0.38 29.80 TC2 607163.00 221009.40 0.70 29.70 TC3 598806.00 221463.00 0.26 30.30 TC4 592919.50 222632.00 0.25 30.20 TC5 593138.20 224263.60 0.39 30.30 TC6 584159.50 224157.50 0.20 30.00 TC7 577123.50 227848.50 0.35 30.20 TD1 597088.20 220135.40 0.47 31.30 TD2 599747.20 216820.00 0.84 31.70 TD3 605328.40 215349.30 0.40 30.80 TD4 614559.60 226007.20 0.45 30.40 TD5 620625.80 234342.90 0.10 29.20 TD6 626872.60 243145.00 0.13 29.40 TE1 627087.00 250059.00 0.26 28.70 TE2 623861.00 246572.00 0.19 29.20 TE3 620623.00 245640.00 0.14 29.50 TE4 619784.00 240967.00 0.41 29.40 TE5 617420.00 235121.00 0.66 29.20 TE6 617612.00 230508.00 0.19 30.20 TE7 617802.00 224962.00 0.21 29.80 TE8 616236.00 219918.00 0.13 29.80 TE9 614571.00 214864.00 0.11 30.40

TE10 614525.00 209828.00 1.71 30.40 TE11 613970.00 204964.00 0.28 30.60 TE12 613513.00 199918.00 0.50 31.40 TE13 606053.00 203492.00 0.66 31.90 TE14 602391.00 206929.00 1.21 32.90 TE15 611797.00 213737.00 0.22 30.70 TE16 614663.00 215273.00 0.20 30.60 TE17 618334.00 220065.00 0.15 30.40 TE18 622262.00 221997.00 0.10 29.70 TE19 625100.00 221948.00 0.07 29.70 TE20 629804.00 224277.00 0.00 29.70 TE21 629799.00 225358.00 0.14 29.90 TE22 629950.00 232579.00 0.26 29.50 TE23 630131.00 238780.00 0.34 30.00

Appendices

59



mg.m-3 SST 0C TE24 637705.00 250586.00 0.10 28.40 TF1 642570.00 248741.00 0.02 28.90 TF2 639713.00 254219.00 0.03 28.30 TF3 645350.00 243765.00 0.10 28.20 TF4 647907.00 238981.00 0.03 28.70 TF5 648174.00 237562.00 0.02 28.40 TF6 652909.00 234033.00 0.03 28.30 TF7 655420.00 231645.00 0.02 28.40 TF8 656759.00 230289.00 0.09 28.30 TF9 647105.00 235256.00 0.07 28.40

TF10 645214.00 239228.00 0.00 29.80 TF11 642383.00 241601.00 0.03 30.40 TF12 640832.00 245196.00 0.05 28.50 TG1 639034.00 250107.00 0.06 28.10 TG2 638470.00 255411.00 0.08 28.20 TG3 636394.00 257066.00 0.15 27.90 TG4 632608.00 261544.00 0.08 28.10 TG5 630119.00 265723.00 0.06 28.30 TG6 628065.00 270145.00 0.06 28.40 TG7 622600.00 267934.00 0.06 28.40 TG8 624227.00 259931.00 0.06 28.40 TG9 623565.00 252713.00 0.22 29.00

TG10 630115.00 254900.00 0.14 28.80 TG11 635056.00 254479.00 0.11 28.50 TG12 637714.00 250635.00 0.08 28.10 TH1 654168.20 250920.00 0.05 28.50 TH2 662242.90 252828.80 0.03 28.60 TH3 669831.20 252506.50 0.04 28.80 TH4 676812.50 253505.20 0.02 28.10 TH5 673593.30 248213.60 0.05 28.70 TH6 677990.90 241845.50 0.04 28.40 TH7 676917.90 239561.90 0.01 28.90 TH8 670192.90 242628.50 0.00 28.90 TH9 660174.10 246155.50 0.02 28.70 TI1 630062.60 238913.40 0.08 28.30 TI2 621975.50 236308.40 0.07 28.60 TI3 617327.70 234708.50 0.31 29.20



60



mg.m-3 SST 0C TI4 613058.20 233117.40 0.16 29.70 TI5 608054.80 231597.10 0.16 30.20 TI6 604478.60 230528.90 0.27 30.80 TI7 601225.80 229715.10 0.30 30.50 TI8 595821.20 230642.50 0.35 30.80 TI9 589856.50 236180.80 0.31 30.50

TI10 589271.20 240978.90 0.46 30.20 TI11 588178.30 241434.90 0.45 29.70 TI12 593791.70 242477.30 0.54 30.10 TI13 598158.20 240804.60 1.37 30.10 TI14 601883.40 237996.50 0.97 30.60 TI15 604153.10 238779.20 0.75 30.10 TI16 607247.20 239445.30 0.59 30.70 TI17 612398.20 240545.10 0.39 30.70 TI18 619713.40 241144.10 0.27 29.50 TI19 624299.80 246482.40 0.13 30.20 TI20 629612.60 248979.80 0.08 30.20 TI21 634072.80 250119.20 0.07 29.80 TI22 637670.00 250578.60 0.06 28.40

Appendices

61

Appendix-2 : Bivarite Fit of IDWA SST by in situ SST



62

Appendix-3 : Bivarite Fit of Kriging SST by in situ SST