estimation of gross primary production using … · nilai gpp dari citra alos/avnir-2 lebih rendah...
Post on 07-Mar-2019
218 Views
Preview:
TRANSCRIPT
THESIS
ESTIMATION OF GROSS PRIMARY PRODUCTION USING SATELLITE DATA AND GIS IN URBAN AREA,
DENPASAR
ABD. RAHMAN AS-SYAKUR
POSTGRADUATE PROGRAM UDAYANA UNIVERSITY
DENPASAR 2009
THESIS
ESTIMATION OF GROSS PRIMARY PRODUCTION USING SATELLITE DATA AND GIS IN URBAN AREA,
DENPASAR
ABD. RAHMAN AS-SYAKUR NIM. 0791261011
MASTER DEGREE PROGRAM
STUDY PROGRAM OF ENVIRONMENTAL SCIENCE POSTGRADUATE PROGRAM
UDAYANA UNIVERSITY DENPASAR
2009
1
THESIS
ESTIMATION OF GROSS PRIMARY PRODUCTION USING SATELLITE DATA AND GIS IN URBAN AREA,
DENPASAR
Thesis to get Master Degree At Master Program on Environmental Science
Postgraduate Program Udayana University
ABD. RAHMAN AS-SYAKUR NIM 0791261011
MASTER DEGREE PROGRAM
STUDY PROGRAM OF ENVIRONMENTAL SCIENCE POSTGRADUATE PROGRAM
UDAYANA UNIVERSITY DENPASAR
2009
2
Agreement Sheet
THIS THESIS HAS BEEN AGREED
ON MARCH, 17 2009
First Supervisor
Dr. Takahiro Osawa
Second Supervisor
Prof. Dr. Ir. I Wayan Sandi Adnyana, MS NIP. 131572566
Knowing,
Chief of Master Program
of Environmental Science
Udayana University
Dr. Ir. I Wayan Arthana, MS.
NIP : 131644718
Director of
Post Graduate Program
Udayana University
Prof. Dr. Ir. Dewa Ngurah Suprapta, M.Sc
NIP : 131475047
iv
3
This Thesis Has Been Evaluated and Examined and Assessed
On March 16, 2009
Based on The Decree Letter of Director of Postgraduate Program Udayana University Number: 250/H.14.4/HK/2009 on March 04th 2009 Examiner Committee of Thesis Research Examiner as follow Head of Examiner : Dr. Takahiro Osawa Members : 1. Prof. Dr. Ir. I Wayan Sandi Adnyana, MS
2. Dr. Ir. I Wayan Arthana, MS 3. Dr. Ir. Ida Ayu Astarini, M.Sc. 4. Dr. Made Pharmawati, M.Sc.
v
4
Hidup bukanlah sekedar jalan untuk mencapai takdir yang telah ditentukan,
tapi hidup merupakan suatu perjalanan untuk menggapai nasib yang lebih baik demi suatu keberhasilan
saat takdir itu datang. Ya Allah, Engkau telah memberikan
suatu kenikmatan tersendiri atas kesempatan itu.
Untuk anugerah itu, saya persembahan tesis ini
bagi umat-umatMu yang ingin berusaha, yang ingin menggapai nasib yang lebih baik,
kepada para pendidik yang memanusiakan manusia, dan kepada orang-orang yang sabar menunggu
demi hari esok yang lebih baik. Saya juga persembahkan tesis ini kepada Aba, Mama dan Iien,
seerta kepada seseorang yang ku cintai
5
ACKNOWLEDGEMENT
Alhamdulillaahirabbil’alamiin. Firstly, the author would like to express sincere gratitude to the Almighty God, Allah S.W.T., for the Great full, Kindness, and Blessing him in finishing thesis research. Secondly, the author would like to thank Ministry of Education Republic of Indonesia for the scholarship support and also Udayana University for accepting the author as a student in Magister Program of Environmental Science with Oceanography and Remote Sensing concentration. The author choosed the topic about Gross Primary Production or total carbon assimilation by vegetation which important for global warming issues. In this opportunity, the author would like to acknowledge: 1. Dr. Takahiro Osawa, as a first supervisor and The Head of CReSOS, who
helping, supporting, and giving many information and literatures for author until this thesis can be finished.
2. Prof. Dr. Ir. I Wayan Sandi Adnyana, MS., as a second supervisor who helping, supporting, and giving many information and literatures for the author until this thesis can be finished.
3. Dr. Ir. I Wayan Arthana, M.S. as a chief of Master Program of Environmental Study Udayana University. Dr. Ir. I Wayan Suarna, MS and Prof. Dr. Ir. I Wayan Sandi Adnyana, MS. (again) as a chief and secretary of Environmental Research Center Udayana University to support and valuable suggestion.
4. Dr. Ir. I Wayan Arthana, M.S. (again), Dr. Ir. Ida Ayu Astarini, M.Sc., and Dr. Made Pharmawati, M.Sc. as examiners for spend time to criticism and give some improvement.
5. Prof. Dr. Ir. Dewa Ngurah Suprapta, M.Sc as Director of Postgraduate Program Udayana University
6. Very special thanks are extended to my parent Taufikurrahman Y. and Faizah for their sacrifices and understanding and also for my sister Iien for spending time and grow up together.
7. My family and all of my friends (especially Aji, Umi, Eka, Puji, Suri, Ende, Putri, Weda, Pande, Bu Ary, Made-PMIL, Nampa-PMIL, Mbo Putu-PMIL, Mbo Iluh-PPLH, and All PPLH researches) who praying, helping, supporting, and giving many information and literatures for the author until this thesis can be finished.
The author realizes that this thesis needs more improvement, so the author will appreciate all of the criticism and suggestion from the reader.
Denpasar, 17 March 2009
Abd. Rahman As-syakur
vi
6
ABSTRAK
Estimasi Gross Primary Production dengan Menggunakan Data Satelit dan SIG di Daerah Perkotaan, Denpasar
Data penginderaan jauh yang memiliki resolusi spasial yang tinggi sangat baik dalam menyediakan informasi sebaran Gross Primary Production (GPP), khususnya di daerah perkotaan. Sebagian besar model untuk menghitung pertukaran karbon di ekosistem yang memanfaatkan data penginderaan jauh menggunakan model light use efficiency (LUE), begitu juga dengan penelitian ini. Adapun tujuan penelitian ini adalah untuk menghitung GPP dengan menggunakan data satellite yang berbeda resolusi spasialnya (ALOS/AVNIR-2 dan Aster). Nilai GPP yang diperoleh dengan memanfaatkan citra ALOS/AVNIR2 adalah berkisar antara 0.130 gC m-2 yr-1 sampai 2586.181 gC m-2 yr-1, sedangkan dengan menggunakan Aster nilai GPPnya berkisar antara 0.144 gC m-2 yr-1 sampai 2595.264 gC m-2 yr-1. Nilai GPP dari citra ALOS/AVNIR-2 lebih rendah dari citra Aster karena ALOS/AVNIR-2 mempunyai resolusi spasial yang lebih kecil dan resolusi spektral yang lebih besar. Perbedaan penggunaan lahan mempengaruhi nilai GPP, karena setiap jenis penggunaan lahan memiliki jenis vegetasi, sebaran vegetasi dan proses fotosintesis yang berbeda-beda. Resolusi spasial yang tinggi sangat penting dalam membedakan jenis tutupan lahan di daerah perkotaan. Dengan tutupan lahan yang heterogen tersebut menghasilkan jumlah GPP dari citra Aster lebih tinggi di bandingkan dengan citra ALOS/AVNIR-2, akan tetapi rata-rata nilai GPP lebih tinggi pada citra ALOS/AVNIR-2 dibandingkan dengan citra Aster. Hasil perbandingan dengan penelitian-penelitian lainnya menunjukkan bahwa nilai GPP dari kedua citra satelit tersebut tidak berbeda jauh dengan nilai GPP yang diperoleh dari MODIS GPP product (MOD17) di daerah Denpasar dan di daerah tropis lainnya (Kalimantan-Indonesia dan hutan Amazon)
Kata Kunci: ALOS/AVNIR-2, Aster, gross primary production, resolusi spasial, SIG
viii
7
ABSTRACT
Estimation of Gross Primary Production using Satellite Data and GIS in Urban Area, Denpasar
Remote sensing data with high spatial resolution is very useful to provide information about GPP (Gross Primary Production) especially over spatial coverage in the urban area. Most models of ecosystem carbon exchange based on remote sensing use some form of the light use efficiency (LUE) model. The aim of this research is to analyze the distribution of annual GPP urban area of Denpasar. Further analysis was carried out using two types of satellite data (ALOS/AVNIR-2 and Aster), in order to know the different location of spatial resolution can affects to detect various ecosystem processes in Denpasar. Annual value of GPP using ALOS/AVNIR-2 varies from 0.130 gC m-2 yr-1 to 2586.181 gC m-2 yr-1. Meanwhile, using Aster the value varies from 0.144 gC m-2 yr-1 to 2595.264 gC m-2 yr-1. The annual value of GPP ALOS is lower than the value of Aster, because ALOS have high spatial resolution and smaller interval of spectral resolution compared to Aster. Different land use can effect the value of GPP, because the different land use has different vegetation type, distribution, and different photosynthetic pathway type. The high spatial resolution of the remote sensing data is crucial discriminating different land cover types in urban land cover. With the surface heterogeneous of land cover, maximum value of GPP using ALOS is smaller than the value GPP Aster, but the annual mean of GPP value by ALOS/AVNIR-2 is higher than the annual mean of GPP by Aster. Comparing with another scientific research, the maximum value of GPP using ALOS/AVNIR-2 and Aster satellite data showed more accurate result in Denpasar area derived from MODIS GPP product (MOD17) and the tropic area (Kalimantan-Indonesia and Amazon forest).
Keyword: ALOS/AVNIR-2, Aster, gross primary production, spatial resolution
ix
8
SUMMARY
Abd. Rahman As-syakur. Estimation of Gross Primary Production using Satellite Data and GIS in Urban Area, Denpasar. First supervisor: Dr. Takahiro Osawa and second supervisor: Prof. Dr. Ir. I Wayan Sandi Adnyana, MS.
Remote sensing could be used to estimate surface–atmosphere CO2 exchange. Most models of ecosystem carbon exchange based on remote sensing use some form of the light use efficiency (LUE) model. The LUE model states that carbon exchange is a function of the amount of light energy absorbed by vegetation and the efficiency with which that light energy is used to fix carbon (Monteith, 1972 in Sims, et al., 2006). Denpasar represent one of urban city in Bali Island. Remote sensing is a tool for mapping and monitoring urban area. Imagery with moderate until high spatial resolution is needed for application remote sensing in urban area. The satellite imagery with a high spatial resolution has been effectively used to classify homogeneous landscapes. Remote sensing often requires other kinds of ancillary data to achieve both its greatest value and the highest level of accuracy as a data and information production technology. Geographic Information Systems (GIS) can provide this capability (Star and Estes, 1990). Despite various shortcoming, results for the analysis indicated that the used of GIS and remote sensing data is useful to provide information about GPP especially over spatial coverage in the urban area. In this study, analysis was focused on the distribution of annual GPP, thus provide basically background information. The aims of the research are (1) to know how GIS application can estimate GPP using satellite data, (2) to evaluate the value of GPP in urban area, Denpasar that estimated using satellite data, (3) to evaluate the different spatial resolution from satellite data which can effect the value of GPP, and (4) to evaluate the different land use that can effect the value of GPP. Research framework has been designed base on the research background, which used satellite data and GIS to estimate GPP in urban area, Denpasar. Vegetation index will be obtained from ALOS/AVNIR-2 and ASTER, Land use map from Bapedalda Badung, and Climatology data from Indonesian Meteorology and Geophysics Agency (BMG). Satellite is employed to obtain vegetation index value, when index vegetation correlate with fAPAR, and can use for estimation GPP. GIS is used to analyst spatial data and to get map of GPP. The research location is surrounding in Denpasar city. Geographically, Denpasar located in 08o36’56”S – 08o42’01”S and 115o10’23”E – 115o16’27”E, and it is located between two neighborhood regency namely Badung and Gianyar Regencies. With material use is ALOS/AVNIR-2 digital image, Aster digital image, Quick Bird Digital image, land use map, topography map an solar radiation data. The data processing is devided into 3 steps, there are (1) Pre processing phase, (2) Digital Spatial Analysis, and (3) Data presentation.
x
9
The GPP annual value is different with differences satellite data. Annual value of GPP in ALOS/AVNIR-2 varies from 0.130 gC m-2 yr-1 to 2586.181 gC m-2 yr-1, mean value of annual GPP is 836.230 gC m-2 yr-1. In Aster satellite, minimum value of GPP is 0.144 gC m-2 yr-1 and maximum value is 2595.264 gC m-2 yr-1 with mean value of GPP is 776.830 gC m-2 yr-1. Totally GPP per year in Denpasar from ALOS/AVNIR-2 is 52421.462 tC yr-1 with the area of 6267.560 ha and analysis with Aster satellite data, total GPP in Denpasar per year is 59355.493 tC yr-1 with the area of 7647.840 ha. The GPP pixel value distribution from ALOS/AVNIR-2 satellite data is dominated by low pixel value (< 250 gC m-2 yr-
1) with the area of 1236.620 ha, which is decreasing until the GPP pixel value is high (2250 – 2587 gC m-2 yr-1) with the area of 17.170 ha. In the other case, the GPP pixel value distribution from Aster satellite data is dominated by low pixel value (< 250 gC m-2 yr-1) with the area of 1694.565 ha, which is decreasing until the GPP pixel value is in high range (2250 – 2595 gC m-2 yr-1) with the area of 6.593 ha. The maximum value of GPP from two satellite data is smaller than the 2707.8 gC m-2 yr-1 is the maximum GPP value derived from MODIS GPP product (MOD17) in Denpasar area, and smaller than the 2859 gC m-2 yr-1 measured over a tropical peat swamp forest in central Kalimantan-Indonesia (Hirano et al., 2005) and smaller than to the 3040 gC m-2 yr-1 measured over a tropical forest in central Amazonia, Brazil (Malhi et al., 1998) The different land use will effect the different of annual GPP value. In ALOS/AVNIR-2 satellite data, the maximum value of annual GPP is come from forest (Mangrove) land use which the value is 2586.181 gC m-2 yr-1 and the minimum value of annual GPP is 0.130 gC m-2 yr-1 from all land use. In Aster satellite data, the maximum value of annual GPP is come from forest (mangrove) land use which the value is 2595.264 gC m-2 yr-1 and the minimum value of annual GPP is from all land use, which value is 0.144 gC m-2 yr-1. In ALOS/AVNIR-2 satellite data, the maximum value of annual GPP in South Denpasar district which the value is 2586.181 gC m-2 yr-1, in West Denpasar district the maximum value of GPP is 2511.426 gC m-2 yr-1, in North Denpasar district the maximum value of GPP is 2462.299 gC m-2 yr-1, and in East Denpasar district the maximum value of GPP is 2449.191 gC m-2 yr-1. In Aster satellite data, the maximum value of annual GPP in South Denpasar district which the value is 2595.264 gC m-2 yr-1, in West Denpasar district the maximum value of GPP is 2289.775 gC m-2 yr-1, in North Denpasar district the maximum value of GPP is 2304.257 gC m-2 yr-1, and in East Denpasar district the maximum value of GPP is 2322.598 gC m-2 yr-1. The value of GPP was closely related to the value of total solar radiation, concentration of carbon dioxide and the light use efficiency. The highest value of GPP is normally closely related with areas of highest vegetation cover. This is mainly due to the highest reflectance in near infrared region. In Denpasar area, the GPP value from ALOS/AVNIR-2 and Aster satellite data is smaller than the GPP value from MODIS product (MOD17). MODIS and Aster image has a short spectral resolution. Smaller interval of spectral resolution will give higher capability of sensor in detecting object in surface.
xi
10
The different effect land use that caused the different value of GPP is quite a lot, because the different land use has a different vegetation type, percentage vegetation cover and dissemination. The value of vegetation index related to percentage vegetation cover (Horning, 2004; Inoue et al., 2008). Settlement land use has a high heterogeneous landscape that influencing the object reflectance value on earth surface. This problem could be solved by use the image satellite with higher resolution. With the higher image satellite resolution, the object detection is get more specific which can make higher accurate of data. This matter caused the annual mean GPP from ALOS/AVNIR-2 in settlement land use higher than in Aster. Higher total of annual GPP from Aster in settlement land use is caused by the different of spatial resolution and the different of spectral resolution between Aster and ALOS/AVNIR-2. Although identify as a low pixel, most places in Denpasar which covered by little vegetation in settlement is identified as a high vegetation by Aster, this problem caused the annual GPP by Aster satellite is higher than the total annual GPP by ALOS/AVNIR-2 in settlement land use. South Denpasar has two types land use with a high total GPP influencing to total GPP value in South Denpasar. Those two types of land use are ricefield and forest (mangrove). West Denpasar has a low amount of GPP value because in West Denpasar that covered by a small area of ricefield, especially without contribution from forest (mangrove) The conclusions of this research are (1) GIS application is used to estimate total annual GPP in urban area, such as in Denpasar, (2) value of GPP from ALOS is smaller than the value from Aster, because ALOS have a high spatial resolution and smaller interval of spectral resolution, (3) different land use can effect to different value of GPP, because the different land use has a different vegetation type, dissemination, and different photosynthetic pathway type, (4) the high spatial resolution of the remote sensing data is crucial discriminating different land cover types in urban land cover compared to Aster, (5) with the surface heterogeneous of land cover, maximum value of GPP from ALOS is smaller than the value GPP that get from Aster, but the annual mean of GPP value by ALOS/AVNIR-2 is higher than the annual mean of GPP by Aster, because ALOS/AVNIR-2 has a high spatial resolution and has more significant detection quality and condition of vegetation than Aster, (6) the maximum value of GPP from two satellite data is smaller than the 2707.8 gC m-2 yr-1 is the maximum GPP value derived from MODIS GPP product (MOD17) in Denpasar area, and smaller than the 2859 gC m-2 yr-1 measured over a tropical peat swamp forest in central Kalimantan-Indonesia (Hirano et al., 2005) and smaller than to the 3040 gC m-2 yr-1 measured over a tropical forest in central Amazonia, Brazil (Malhi et al., 1998). The suggestion of the research is the differences of spatial and spectral resolutions influence accuration of object detection. The object detection for heterogeneous area such as settlement land use is recommended to use satellite with high spatial resolution, meanwhile for homogeneous area such as forest (mangrove) and ricefield is recommended to use satellite with high spectral resolution.
xii
11
LIST of CONTENTS
APPROVAL PAGE OF THESIS AGREEMENT ...................................... iv EXAMINED OF THESIS ............................................................................. v ACKNOWLEDGEMENT ............................................................................. vi ABSTRAK ...................................................................................................... viii ABSTRACT .................................................................................................... ix SUMMARY .................................................................................................... x LIST of CONTENTS ..................................................................................... xiii LIST of TABLES ........................................................................................... xv LIST of FIGURES ......................................................................................... xvii LIST of APPENDIX....................................................................................... xx CHAPTER I INTRODUCTION.................................................................. 1 1.1 Background ............................................ ........................ 1 1.2 Problems Formula .................................................. ........ 4 1.3 Aim of Research.............................................................. 4 1.4 Research Benefits............................................................ 4 CHAPTER II LITERATURE REVIEW .................................................... 6 2.1. Remote Sensing............................................................... 6 2.2. Spectral Characteristics of Vegetation ............................ 11 2.3. Vegetation Index ............................................................. 14 2.4. Estimation CO2 Assimilations by Vegetation Using Satellite Data ................................................................... 18 2.5. Characteristic of ALOS/AVNIR-2 and Aster ................. 23 2.5.1. ALOS/AVNIR-2 ................................................... 23 2.5.2. Aster ...................................................................... 26 2.6. Geographic Information Systems (GIS) ......................... 28 CHAPTER III FRAMEWORK OF RESEARCH ..................................... 35 CHAPTER IV RESEARCH METHODS ................................................... 36 4.1. Research Location..................................................... ..... 36 4.2. Research Material........................................................... 37 4.3. Research Instrument....................................................... 39 4.4. Data Source .................................................................... 39 4.5. Data Processing.............................................................. 40 4.5.1. Pre processing phase ........................................... 40 4.5.2. Digital spatial analysis ........................................ 42 4.5.3. Data presentation................................................. 45 CHAPTER V RESULTS .............................................................................. 47 5.1. Total GPP..................................................... .................. 47 5.2. The Annual GPP by Land Use in Denpasar................... 50 5.3. The Annual GPP by Districts in Denpasar..................... 67
xiii
12
CHAPTER VI DISCUSSIONS .................................................................... 79 CHAPTER VII CONCLUSSIONS AND SUGGESTIONS ...................... 86 6.1. Conclusions..................................................... ............... 86 6.2. Suggestions................... ................................................. 87 REFERENCES ............................................................................................... 88 APPENDIX ..................................................................................................... 96
xiv
13
LIST of TABLES
Table 2.1. ALOS characteristics .................................................................... 25 Table 2.2. Characteristics of the 3 ASTER Sensor Systems.......................... 28 Table 3.1. Shows the unit conversion coefficient for each band.................... 42 Table 5.1. Annual and total value of GPP with differences satellite data...... 48 Table 5.2. Total pixels and hectarage of annual GPP value with differences satellite data ................................................................ 48 Table 5.3. Annual value of GPP with differences land use in
ALOS/AVNIR-2 satellite data ...................................................... 51 Table 5.4. Annual value of GPP with differences land use in Aster satellite data................................................................................... 52 Table 5.5. The totally of annual value of GPP with differences land use in
ALOS and Aster satellite data....................................................... 53 Table 5.6. Total pixels and hectarage of annual GPP value with differences satellite data in settlement land use ............................ 54 Table 5.7. Total pixels and hectarage of annual GPP value with differences satellite data in ricefield land use ............................... 56 Table 5.8. Total pixels and hectarage of annual GPP value with differences satellite data in forest (mangrove) land use................ 58 Table 5.9. Total pixels and hectarage of annual GPP value with differences satellite data in shrub land use.................................... 60 Table 5.10. Total pixels and hectarage of annual GPP value with differences satellite data in perennial plant land use..................... 62 Table 5.11. Total pixels and hectarage of annual GPP value with differences satellite data in dryland land use ................................ 64 Table 5.12. Total pixels and hectarage of annual GPP value with differences satellite data in bareland land use............................... 48 Table 5.13. The totally of annual value of GPP by district in ALOS/AVNIR-2 satellite data ...................................................... 68 Table 5.14. The totally of annual value of GPP by district in Aster satellite data................................................................................... 69
xv
14
Table 5.15. The totally of annual value of GPP by district in ALOS and Aster satellite data ......................................................................... 70 Table 5.16. Total pixels and hectarage of annual GPP value with differences satellite data in South Denpasar district ..................... 71 Table 5.17. Total pixels and hectarage of annual GPP value with differences satellite data in West Denpasar district ...................... 73 Table 5.18. Total pixels and hectarage of annual GPP value with differences satellite data in North Denpasar district ..................... 75 Table 5.19. Total pixels and hectarage of annual GPP value with differences satellite data in East Denpasar district........................ 77 Table 6.1. Land use area by district ............................................................... 84 Table 6.2. Cloud Hectarage by district........................................................... 85
xvi
15
LIST of FIGURES
Figure 2.1. Scheme of remote sensing ........................................................ 7 Figure 2.2. Spectra of Sample Cover Types and Landsat TM/ETM Bands (Angel et al., 2005)................................................................... 10 Figure 2.3. Reflectance and transmittance spectra of a typical fresh, green
leaf (Rautiainien, 2005) ............................................................ 12 Figure 2.4. Simplified cross-sectional view of behavior of energy interacting
with a canopy of living vegetation (Short, 2006) ..................... 13 Figure. 2.5. The relationship between fAPAR and top of the canopy (TOC)
NDVI (Myneni and Williams, 1994)........................................ 18 Figure. 2.6. The seasonal dynamics of predicted of gross primary production
(GPP) with MODIS (a) and VGT (b) satellite data and observed GPP at Harvard Forest in 2001 (Xiao, et al., 2004) ................. 21
Figure 2.7. ALOS satellite .......................................................................... 23 Figure 2.8. The ASTER satellite................................................................. 27 Figure 2.9. Data integration is the linking of information in different forms
through a GIS (USGS, 2007).................................................... 32 Figure 2.10. The concept of layers (ESRI, 1996) ......................................... 33 Figure 3.1 Framework of Research............................................................ 35 Figure 4.1. Study site .................................................................................. 37 Figure 4.2. Digital image of ALOS/AVNIR-2 and Aster of
Denpasar area in 2006 .............................................................. 38 Figure 4.3. Land use map............................................................................ 38 Figure 4.4. Research Schema...................................................................... 45 Figure 4.5. GIS Process .............................................................................. 46 Figure 5.1. Annual distribution of GPP from ALOS/AVNIR-2................. 49 Figure 5.2. Annual distribution of GPP from Aster.................................... 49 Figure 5.3. Annual distribution of GPP from MODIS product (MOD17) . 50
xvii
16
Figure 5.4. Graphic of annual GPP value with different land use in
ALOS/AVNIR-2 satellite data ................................................. 51 Figure 5.5. Graphic of annual GPP value with different land use in Aster satellite data .............................................................................. 52 Figure 5.6. Totally of annual value of GPP with differences land use in
ALOS and Aster satellite data .................................................. 53 Figure 5.7. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in settlement land use ........ 55 Figure 5.8. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in ricefield land use ........... 57 Figure 5.9. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in forest (mangrove) land use..................................................................................... 59 Figure 5.10. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in shrub land use................ 61 Figure 5.11. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in perennial plant land use. 63 Figure 5.12. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in dryland land use ............ 65 Figure 5.13. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in bareland land use........... 67
xviii
17
Figure 5.14. Totally of annual value of GPP by district in ALOS/AVNIR-2 satellite data .............................................................................. 68
Figure 5.15. Totally of annual value of GPP by district in Aster satellite data .............................................................................. 69 Figure 5.16. Totally of annual value of GPP by district in ALOS and Aster
satellite data .............................................................................. 70 Figure 5.17. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in South Denpasar District 72 Figure 5.18. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in West Denpasar District.. 74 Figure 5.19. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in North Denpasar District 76 Figure 5.20. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in settlement land use ETM+ (d) in East Denpasar District... 78 Figure 6.1. GPP distribution in ALOS/AVNIR-2 (a) and GPP distribution in Aster (b), compare with vegetation distribution from Quickbird satellite data (c) and compare to with ground check picture............................................................................. 83
xix
18
LIST of APPENDIX APPENDIX 1. GPP Map Derived from Satellite Data and GIS echnique Aplication ............................................................. 96
xx
CHAPTER I
INTRODUCTION
1.1. Background
Considered globally, the most important interaction between the biosphere
and atmosphere are the transfer of energy, water, and carbon. Carbon is
assimilated by the biosphere through photosynthesis and released through
autotrophic and heterotrophic respiration (Malhi et al., 1998). Emissions and re-
absorption of these gases from natural ecosystem have been in equilibrium for
million of years. However, this balance has been disturbed by human activities.
Consequently, the atmospheric concentrations of CO2 have been increasing
rapidly and it is widely believed that higher concentration of these gases is
responsible for global warming (Hazarika and Yasuoka, 2002). Understanding the
control on spatial and temporal pattern of surface–atmosphere CO2 exchange is
therefore needed so that improved prediction of future level of atmospheric CO2
could be made (Jenkins, et al., 2007). This highlights the need to monitor plant
cover and corresponding surface CO2 uptake on a large scale. Such data will aid
to obtain accurate estimation of regional and global carbon budget and, ultimately,
more accurate prediction of carbon source-sink relationships and atmospheric CO2
concentration (Hunt et al., 2002).
Remote sensing could be used to estimate surface–atmosphere CO2
exchange. Remotely sensed optical signatures have proved useful for estimating
ecological variables such as leaf area index (LAI) and the absorbtivity of
1
2
photosynthetically active radiation (fAPAR) (Asrar et al., 1984 in Inoue, 2007;
Turner et al., 2002). Fraction of absorbed photosynthetically active radiation
(fAPAR) by the vegetation cover is related to the normalized difference
vegetation index (NDVI). The strong relationship between NDVI and fAPAR has
been examined in detail with theoretical and experimental analyses (Myneni and
Williams, 1994; Kumar and Monteith, 1981 in Hooda and Dye, 1996; Inoue et al.,
2008). The NDVI has become a popular tool for assessing different aspects of
plant processes, while simultaneously determining spatial variation in vegetation
cover (La Puma et al., 2007).
Most models of ecosystem carbon exchange based on remote sensing use
some form of the light use efficiency (LUE) model. The LUE model states that
carbon exchange is a function of the amount of light energy absorbed by
vegetation and the efficiency with which that light energy is used to fix carbon
(Monteith, 1972 in Sims, et al., 2006). Monteith (1972) in Bradford (2005)
developed method for estimating plant productivity from observation of absorbed
photosynthetically active radiation (APAR) and estimates of LUE.
Denpasar represent one of urban city in Bali Island. Remote sensing is a tool
for mapping and monitoring urban area. For application of remote sensing to
urban area, we need imagery with moderate until high spatial resolution. The
satellite imagery with a high spatial resolution has been effectively used to
classify homogeneous landscapes. A higher spatial resolution is greatly desirable
for land application (e.g., ecosystem and hydrology) (Liang et al., 2007) and very
useful to acquire vegetative information (Yüksel et al., 2008) in urban areas.
3
Operational potential of urban remote sensing will depend on the capacity of
remote sensing to capture objects in urban areas (Tang, 2007). The urban
landscapes the flux measurement gets more complicated due to the surface
heterogeneity and the NDVI loses its importance for scaling the CO2 exchange
(Soegaard and Møller-Jensen, 2003). Landsat imagery with a moderate spatial
resolution of 30 m has been effectively used to classify homogeneous landscapes.
However, their accuracy may diminish in regions with highly heterogeneous
landscapes (Yüksel et al., 2008). Therefore we need imagery higher than Landsat
spatial resolution to classify and identify heterogeneous landscapes.
Remote sensing often requires other kinds of ancillary data to achieve both
its greatest value and the highest level of accuracy as a data and information
production technology. Geographic Information Systems (GIS) can provide this
capability (Star and Estes, 1990). GIS can make order to develop the required
capability of natural resources mapping and periodical monitoring (Muzein,
2006).
Several previous scientist such as in Kalimantan tropical forest showed thad
value of GPP is from 2859 until 3227 gC m-2 yr-1 (Hirano et al., 2005) and the
research in Amazon tropical forest got the value of GPP that is 3040 gC m-2 yr-1
(Malhi et al., 1998). According to Xiao et al. (2004), the seasonal dynamics of
GPP prediction from satellite data were similar to those of GPP from observation.
Seasonally integrated GPP observation over eight month period accounted for
98% of annual GPP prediction. In tropical evergreen forest, Amazon-Brazil,
prediction of GPP from MODIS satellite data is consistent with GPP estimation
4
from the eddy flux tower (Xiao et al., 2005; Seleska et al., 2003), GPP value
prediction from MODIS satellite data is about 2977 gC m-2 year-1 (Xiao et al.,
2005). The research in Labanan Concession Area, East Kalimantan-Indonesia got
the annual value of GPP ranges from 1710 to 2635 gC m-2 year-1 estimation from
MODIS satellite data (Nugroho, 2006). Gitelson et al. (2008) found linearly value
of GPP from Landsat related with daytime maize GPP from observation with root
mean squared error less than 1.58 gC m-2 d-1 in a GPP range of 1.88 to 23.1 gC m-
2 d-1.
Despite various shortcoming, results of the analysis indicated that the used
of GIS and remote sensing data can be very useful to provide information about
GPP especially over spatial coverage in the urban area. In this study, analysis was
focused on the distribution of annual GPP, thus provide basically background
information.
1.2 Problems Formula
1. How can GIS application estimate GPP using satellite data?
2. How much value of GPP in urban area, Denpasar can be estimated using
satellite data?
3. How does different spatial resolution from satellite data affect to GPP value?
4. How does different land use affect value of GPP?
5
1.3 Aim of Research
1. To evaluate how GIS application can estimate GPP using satellite data.
2. To evaluate how much value of GPP in urban area, Denpasar can be
estimated using satellite data.
3. To evaluate how different spatial resolution from satellite data influence
value of GPP.
4. To evaluate how different land use can affect value of GPP.
1.4 Research Benefits
1. To provide more alternative for image processing using GIS analysis.
2. To provide information of how much value of GPP in urban area, Denpasar
estimated using satellite data.
3. It can be used as a reference for further analysis and research on amount of
carbon assimilation by vegetation in urban area.
6
CHAPTER II
LITERATURE REVIEW
2.1. Remote Sensing
Remote sensing is defined as the collection of information about an object
without being in physical contact with the object. Aircraft and satellites are the
common platforms from which remote sensing observations are made (Sabins,
1977). According to Lillesand and Kiefer (1994), remote sensing is the science
and the art of obtaining information about an object, area, or phenomenon through
the analysis of data acquired by a device that is not in contact with the object,
area, or phenomenon under investigation. The characteristics measured by a
sensor are the electromagnetic energy reflected or emitted by the Earth’s surface.
This energy relates to some specific parts of the electromagnetic spectrum:
usually visible light, but it may also be infrared light or radio waves (Kerle et al.,
2004).
In much of remote sensing, the process involves an interaction between
incident radiation and the target of interest. This is exemplified by the use of
imaging system where the following seven elements are involved; 1) Energy
Source or Illumination (A), 2) Radiation and the Atmosphere (B), 3) Interaction
with the Target (C), 4)Recording of Energy by the Sensor (D), 5)Transmission,
Reception, and Processing (E), 6) Interpretation and Analysis (F), and 7)
Application (G) (Fig. 2.1) (CCRS, 2007).
6
7
Fig. 2.1. Scheme of remote sensing (CCRS, 2007)
Remote Sensing in the most generally accepted meaning refers to
instrument-based techniques employed in the acquisition and measurement of
spatially organized (most commonly, geographically distributed) data/information
on some properties (spectral; spatial; physical) of an array of target points (pixels)
within the sensed scene that correspond to features, objects, and materials, doing
this by applying one or more recording devices not in physical, intimate contact
with the items under surveillance.
Techniques involve amassing knowledge pertinent to the sensed scene
(target) by utilizing electromagnetic radiation, force fields, or acoustic energy
sensed by recording cameras, radiometers and scanners, lasers, radio frequency
receivers, radar systems, sonar, thermal devices, sound detectors, seismographs,
magnetometers, gravimeters, scintillometers, and other instruments (Short, 2006).
Possibly the most significant characteristic of the image data in a remote
sensing system is the wavelength, or range of wavelengths, used in the image
acquisition process. The energy emitted by the earth itself can also be resolved
8
into different wavelengths that help us understand the properties of the earth
surface region being imaged (Richards and Jia, 2006). According to Short (2006),
for any given material, the amount of solar radiation that it reflects, absorbs,
transmits, or emits varies with wavelength. When that amount coming from the
material is plotted over a range of wavelengths, the connected points produce a
curve called the material's spectral signature (spectral response curve). The special
information in form of emitted electromagnetic radiation or transmitted from
surface of earth, because sensor attached far from object which is sensed,
therefore energy which radiated or emitted by object is needed. It is also possible
to collect information about an object or geographic area from a distant vantage
point using specialized instrument (sensors) (Lillesand and Kiefer, 1994; Black,
2006).
Sensor can obtain very specific information about object or the geographic
extent of a phenomenon. The electromagnetic energy emitted or reflected from an
object or geographic area is used as a surrogate for the actual property under
investigation. The electromagnetic energy measurements must be turned into
information using visual and/or digital processing techniques (Moriyama, 2005).
According by Lillesand and Kiefer (1994), interaction happened between energy
of an object. Every object has different characteristic or attitude in its interaction
of energy, for example water absorbs much light and only reflected a little light,
on the other hand calcify rock or snow absorb a few light and reflected much
light.
9
Remotely sensed data can be used from local to global scales in
characterizing various ecological variables that are applicable in monitoring, for
example, changes in land and vegetation cover, land use, vegetation structure,
phenological cycles, natural disasters or biodiversity of habitats (Rautiainen,
2005). As states by Current (2004), remotely sensed data can be used to provide
information on land cover (e.g., different vegetation types or classes of vegetation
amount) and thereby habitat.
According to Angel et al., (2005), Remote sensing images have four
different types of resolutions: spectral, spatial, radiometric, and temporal. Spectral
resolution characterizes the range of sensitivity of sensors to different
wavelengths of electromagnetic radiation, as well as the width and placement of
those bands. Spatial resolution characterizes with the fineness of detail afforded
by the sensor optics and platform altitude. Radiometric resolution refers to the
number of unique quantization (brightness) levels in the data. And temporal
resolution characterizes the frequency of revisits by a remote sensing platform.
The generic nature of remote sensing techniques and the wide range of
spatial and temporal resolutions of the data sets make it possible to apply remote
sensing in studying the processes and structure of a multitude of terrestrial
ecosystems such as forests, agricultural fields, wetlands and urban vegetation. It is
also important to acknowledge the interactions between different parts of the
biosphere, and thus obtaining simultaneous time series data from vegetation,
oceans and atmosphere helps us assess many global environmental phenomena
10
(Rautiainen, 2005). Therefore, each application itself has specific demands, for
spectral resolution, spatial resolution, and temporal resolution (CCRS, 2007).
In addition to spectral reflectance (i.e., color or tone), a human analyst will
employ other criteria in the visual-cognitive process of interpreting remote
sensing imagery: texture, pattern, size, shape, shadow, and context, among other
visual cues. In contrast, however, most methods for computer-assisted
classification of digital remote sensing data that do not involve a human observer
utilize a “per-pixel, spectral data-alone” approach (Angel et al., 2005). Figure 2.2
presents the spectral reflectance properties of several land cover types.
µm
Fig. 2.2. Spectra of Sample Cover Types and Landsat TM/ETM Bands (Angel et
al., 2005)
11
2.2. Spectral Characteristics of Vegetation
Remote sensing technologies employing satellites provide the ability to
repeatedly image and monitor vegetation dynamics over large areas, including the
entire globe (Horning, 2004). The spectral (i.e. wavelength dependent) variability
of reflectance is probably the most utilized information source in the remote
sensing of land surfaces (Heiskanen, 2007).
The vegetation shows typically a low reflectance in the visible range of the
spectrum, particularly in the blue and red wavelengths, a steep increase in
reflectance around 700 nm (red edge) and high reflectance in the near infrared
(NIR) (Heiskanen, 2007). The original incident radiation on a leaf is divided into
the spectral hemispherical reflectance, transmittance and absorption of a leaf.
Typically, only approximately 2 to 3 % of the radiation which initially is incident
on the leaf surface is immediately (without entering the leaf) reflected from the
leaf surface (Tucker and Garratt, 1977 in Rautiainien, 2005)
According to Horning (2004), chlorophyll absorbs light in specific
wavelength bands. Vegetation appear green to the human eye because
preferentially more green light is reflected (and hence less absorbed) from the
leaf's surface and internal structure than the rest of the visible portion of the
spectrum. Moreover, vegetation reflect infrared light even more than green, so if
human eyes were sensitive to infrared light, the leaves would appear very bright
and reflective to us (Fig. 2.3).
12
Fig. 2.3. Reflectance and transmittance spectra of a typical fresh, green leaf
(Rautiainien, 2005)
Absorption centered at about 0.65 µm (visible red) is controlled by
chlorophyll pigment in green-leaf chloroplasts that reside in the outer or palisade
leaf. Absorption occurs to a similar extent in the blue. With these colors thus
removed from white light, the predominant but diminished reflectance of visible
wavelengths is concentrated in the green. Thus, most vegetation has a green-leafy
color. There is also strong reflectance between 0.7 and 1.0 µm (near IR) in the
spongy mesophyll cells located in the interior or back of a leaf, within which light
reflects mainly at cell wall/air space interfaces, much of which emerges as strong
reflection rays. The intensity of this reflectance is commonly greater (higher
percentage) than from most inorganic materials, so vegetation appears bright in
the near-IR wavelengths (Fig. 2.4) (Short, 2006)
13
Fig. 2.4. Simplified cross-sectional view of behavior of energy interacting with a canopy of living vegetation (Short, 2006)
A health green leaf intercepts incident radian flux directly from the sun or
from diffuse skylight scattered onto the leaf. This incident electromagnetic energy
interacts with the pigments, water, and intercellular air spaces within plant leaf.
The amount of radiant flux reflected from the leaf, the amount of radian flux
absorbed by the leaf, and the amount radian flux transmitted through the leaf can
be carefully measured as we apply the energy balance equation and attempt to
keep track of what happens to all the incident energy. The energy reflected from
the plant leaf surface is equal to the incident energy minus the energy absorbed
directly by the plant for photosynthetic or other purposes and the amount of
14
energy transmitted directly throught the leaf onto other leaves or the ground
beneath the canopy (Jensen, 2000).
2.3. Vegetation Index
The reflectance and transmittance of leaves is a function of both the
concentration of light absorbing compounds (chlorophyll, water, dry plant matter
etc) and the surface/internal scattering of light that is not efficiently absorbed
(Ustin et al., 2005). This is basic theory to know vegetation index from remotely
sensed data. Vegetation indices are among the oldest tools in remote sensing
studies. Although many variations exist, most of them ratio the reflection of light
in the red and NIR sections of the spectrum to separate the landscape into water,
soil, and vegetation (Glenn et al., 2008).
A vegetation index is generated by combining data from multiple spectral
bands into a single value. Usually simple algebraic formulations, Vegetation
index’s are designed to enhance the vegetation signal in remotely sensed data and
provide an approximate measure of live, green vegetation amount (Horning,
2004). When vegetation density is low, background reflectance significantly
influences canopy reflectance and when the vegetation density is high, leaves are
the primary scattering elements and the background contributes little to overall
canopy reflectance (Daughtry, 2006).
The vegetation index are built on the observation that chlorophylls a and b in
green leaves strongly absorb light in the Red, with maximum absorption at about
690 nm, while the cell walls strongly scatter (reflect and transmit) light in the NIR
15
region (about 850 nm). This results in a strong absorption contrast across a narrow
wavelength band of 650 - 850 nm, captured by the vegetation indices (Glenn et
al., 2008). When solar radiation interacts with matter, it may be reflected,
transmitted, or absorbed. The spectral reflectance of crop canopies is determined
by 1) leaf spectral properties, 2) leaf area and canopy geometry, 3) background
(soil or residue) reflectance, 4) illumination and view angles, and 5) atmospheric
transmittance (Bauer, 1985 in Daughtry, 2006).
Spectral vegetation indices have been found to be related to a number of
biophysical parameters (variables) of interest to many researchers, including Leaf
Area Index (LAI), percent vegetation cover, green leaf biomass, fraction of
absorbed photosynthetically active radiation (fAPAR), photosynthetic capacity,
and carbon dioxide fluxes (Horning, 2004). According to Glenn et al. (2008),
vegetation indices are now indispensable tools in land cover classification,
climate- and land-use-change detection, drought monitoring, and habitat loss, to
name just a few applications.
The simple ratio vegetation index (termed SR or RVI) is calculated using the
following formula (Jordan, 1969 in Huete et al., 1999):
Band RedBand Red InfraNear RVI= (1)
Where the value of band can be digital counts, at satellite radiances, top of
the atmosphere apparent reflectances, land leaving surface radiances, surface
reflectances, or hemispherical spectral albedos. However, for densely vegetated
16
areas, the amount of red light reflected approaches very small values and this
ratio, consequently, increases without bounds (Huete et al., 1999).
There are many vegetation indices such as normalized difference vegetation
index (NDVI), infrared percentage vegetation index (IPVI), soil adjusted
vegetation index (SAVI), modified soil adjusted vegetation index (MSAVI), and
green vegetation index (GVI). But the most popular is the NDVI, is one of the
ratio vegetation indices by rationing the difference between the NIR and red bands
by their sum (Zhang, 2006), this ratio from -1 to +1 (Huete, et al., 1999; Glenn et
al., 2008) dense vegetation has a high NDVI, while soil values are low but
positive, and water is negative due to its strong absorption of NIR (Glenn et al.,
2008). NDVI varies between 0 and 1 for vegetated surfaces, with desert values
near zero and those for tropical forests near one (Tu, 2000). The following
formula is:
BandRedBandRed InfraNear Band Red - Band Red InfraNear NDVI
+= (2)
The NDVI has become a popular tool for assessing different aspects of plant
processes, while simultaneously determining spatial variation in vegetation cover
(La Puma et al., 2007). NDVI is also the most commonly used vegetation index
which has been widely used to evaluate cover, above-ground biomass, chlorophyll
content, leaf area, penology (Zhang, 2006; Horning, 2004), fraction absorbed
photosynthetically active radiation (fAPAR/fPAR) (Myneni and Williams, 1994;
Kumar and Monteith, 1981 in Hooda and Dye, 1996; Inoue et al., 2008), Gross
Primary Productivity (GPP), Net Primary Productivity (NPP) (Running et al.,
17
1999; Inoue et al., 2008), roughness lengths for turbulent transfer, albedo,
emissivity and other biophysical properties of the landscape (Glenn et al., 2008).
However, NDVI is influenced by many environmental factors such as topography,
bare soil (soil fraction, soil type, and soil moisture), atmospheric condition,
vegetation association, rainfall, and non-photosynthetic materials (Zhang, 2006)
The strong relationship between the NDVI and fAPAR has been examined
in detail with theoretical and experimental analyses (Myneni and Williams, 1994;
Kumar and Monteith, 1981 in Hooda and Dye, 1996; Inoue et al., 2008) because
of its positive linear relationship (Fig. 2.5) ((Myneni and Williams, 1994). Other
studies have shown that the relation between FAPAR and NDVI is similar for
one-dimensional and three-dimensional canopies (Myneni et al.,. 1992 in Lotsch,
1996). The relationship between NDVI and fAPAR can be used to determine total
CO2 assimilation by vegetation or gross primary production (GPP) using models
of light use efficiency (LUE) (Running et al., 1999). Both NDVI and fAPAR
integrate the effects of the leaf quantity (LAI) and leaf quality (chlorophyll) (Tu,
2000)
18
r2 = 0.919 N = 252 fAPAR = 1.1638 (NDVI) - 0.1426
Fig. 2.5. The relationship between fAPAR and top of the canopy (TOC) NDVI
(Myneni and Williams, 1994)
2.4. Estimation CO2 Assimilations by Vegetation Using Satellite Data
Plants use solar energy in a chemical reaction which converts carbon dioxide
and water into carbohydrates (Horning, 2004). This crucial process is known as
photosynthesis. Photosynthesis is the process whereby plants synthesize organic
compounds using inorganic raw materials in the presence of light energy (Black,
2006). Photosynthetic Carbon Assimilation (PCA) has often been summarized by
the equation:
19
This portrays a reaction sequence, driven by light energy, in which carbon
dioxide (CO2) and water (H2O) are consumed, oxygen (O2) is liberated and
glucose (C6H12O6) is the end product. However, although sugars such as glucose
and fructose eventually appear in plant leaves as a result of photosynthesis, they
are not formed directly from carbon dioxide and are not particularly important
compounds in leaf metabolism (Edwarsd and Walker, 2001)
CO2 is also emitted by forests through plant respiration and through the
processes of death and decay. The net balance of CO2 uptake and release will
determine whether an ecosystem is acting as a sink or source of carbon (ECCM,
2002). This process is referred to as terrestrial carbon sequestration, as the carbon
is removed from the atmosphere and assimilated into the vegetation and soil
(Black, 2006) and stored in wood, other biomass and soil organic matter, is
highest in young forests and will tend to reduce as forests reach maturity (ECCM,
2002).
Two types of primary production, Gross Primary Productivity (GPP) and
Net Primary Productivity (NPP), are used to describe the rate at which ecosystems
produce organic matter through photosynthesis. The total of the converted energy
is called Gross Primary Productivity (GPP) and Net Primary Productivity (NPP) is
the difference between GPP and energy lost during plant respiration (Campbell
1990 in Heinsc et al., 2003). GPP indicates carbon uptake by plants from the
atmosphere, and NPP describes the net flux of carbon between plants and the
atmosphere. The net carbon gained through NPP is used to increase plant biomass
or to supply herbivores and decomposers. Plant biomass is defined as the sum of
20
dry mass of all plant tissues contained in a defined area and can be reported in its
carbon equivalent (i.e., kg C ha-1). It can be partitioned into two parts,
aboveground biomass (leaf, stem and pole) and belowground biomass (roots)
(Zhao, 2007).
Primary production can be estimated by combining remote sensing with
carbon cycle processing (Heinsc et al., 2003). GPP and NPP have been estimated
based on biophysical parameters derived from vegetation indices (such as the
normalized difference vegetation index, NDVI), land-cover data, light-use-
efficiency parameters, and carbon allometric equations (Running et al., 1999).
A CO2 assimilation model is needed to estimate the exchange of CO2
between an ecosystem and its environment through photosynthetic and respiration
processes (Grose, 2004). Remote sensing of ecosystem-atmosphere CO2 exchange
requires the use of models that relate these variables to rates of photosynthesis.
The challenge remains to develop models that can provide estimates of land
surface CO2 exchange comparable to those obtained by ground-based eddy
covariance (Tu, 2000). Xiao et al. (2004) shown seasonal dynamics of GPPpred
estimation with MODIS satellite data (Fig. 2.6a) and VGT satellite data (Fig.
2.6b) were similar to those of GPPobs at Harvard Forest in 2001.
21
Fig. 2.6. The seasonal dynamics of predicted of gross primary production (GPP)
with MODIS (a) and VGT (b) satellite data and observed GPP at Harvard Forest
in 2001 (Xiao, et al., 2004).
Most models of ecosystem carbon exchange based on remote sensing use
some form of the light use efficiency (LUE) model. The LUE model states that
carbon exchange is a function of the amount of light energy absorbed by
vegetation and the efficiency with which that light energy is used to fix carbon
(Monteith, 1972 in Sims, et al., 2006; Zhao, 2007). Some models included in this
type are canopy photosynthesis models or CPMs, production efficiency models or
PEMs (Gitelson et al., 2008) and Vegetation Photosynthesis models or VPMs
(Xiao et al., 2004). These satellite-based studies have used the LUE approach to
estimate GPP (Running et al., 1999). Monteith (1972) in Bradford (2005)
developed methods for estimating plant productivity from observations of
absorbed photosynthetically active radiation (APAR) and estimates of LUE, The
following formula is:
GPP = ε × fAPAR × PAR = ε × APAR (3)
22
Where GPP is gross primary productivity (gC m-2 time-1), PAR is
Photosynthetically Active Radiation (MJ m-2 time-1), APAR is absorbed
photosynthetically active radiation (MJ m-2 time-1) and ε is light-use efficiency
(gC MJ-1).
The LUE term represents a conversion efficiency, or the ratio of carbon
biomass produced for each unit of absorbed light. In natural ecosystems, LUE is
determined by many biological and biophysical factors, principally, maximum
photosynthetic rates under light saturated conditions, fraction of photosynthesis
consumed by autotrophic respiration, quantum yield of photosynthesis,
photosynthetic pathway (C3 vs. C4), and climate (Still et al., 2004). According to
Jongschaap (2006), net RUE varies among crops and crop varieties, but is more
stable with location and/or crop management. Crop location determines available
growth resources, such as incoming radiation, temperature and precipitation. Soil
characteristics modified by management co-determine the availability of these
resources for crop growth
Both projects exploited remotely sensed data and biophysical or ecosystem
models to estimate GPP and biomass. Altering the resolution of spatial data and
summary units may generate different results. The spatial resolution effects need
to be characterized to evaluate scale-related uncertainties associated with carbon
estimates (Zhao, 2007).
Running et al. (1999) suggested the variation of three factors deserve further
study in the context of using remote sensing to derive spatial estimates of GPP: 1)
spatial resolution 2) land cover and 3) ε estimates..
23
2.5. Characteristics of ALOS/AVNIR-2 and Aster
2.5.1. ALOS/AVNIR-2
The Advanced Land Observing Satellite (ALOS) is a Satellite following the
Japanese Earth Resources Satellite-1 (JERS-1) and Advanced Earth Observing
Satellite (ADEOS) which will utilize advanced land observing technology. The
ALOS satellite vehicle was launched on January 24, 2006. ALOS is equipped
with a mapping stereocamera (PRISM) allowing to obtain images with the
resolution up to 2.5 m, as well as with a multispectral camera (AVNIR-2)
allowing to obtain color images with the resolution of 10 m. The ALOS satellite is
also equipped with radar with the PALSAR synthetic aperture operating in L-
range and allowing obtaining the radar data with the resolution up to 8 m,
including in the interferometric and polirimetric survey modes (Fig. 2.7)
(SOVZOND, 2007).
Fig. 2.7. ALOS satellite
24
PRISM can be used for digital elevation mapping, AVNIR-2 is used for
precise land coverage observation, and PALSAR is used for day-and-night and
all-weather land observation. In order to utilize fully the data obtained by these
sensors, the ALOS was designed with two advanced technologies: the former is
the high speed and large capacity mission data handling technology and the latter
is the precision spacecraft position and attitude determination capability. They
will be essential to high-resolution remote sensing satellites in the next decade
(JAXA, 2007). The ALOS will provide “homogenous quality data for 1/25,000
scale global maps” including elevation, vegetation, land use and land cover data.
ALOS is one of the largest Earth observing satellite ever developed. Its
objectives are (EORC-JAXA, 2008):
1. Cartography: to provide maps for Japan and other countries including those in
the Asian-Pacific region,
2. Regional Observation: to perform regional observation for "sustainable
development" and harmonization between earth environment and
development,
3. Disaster Monitoring: to conduct disaster monitoring around the world,
4. Resource Surveying: to survey natural resources, and
5. Technology Development: to develop technology necessary for future Earth
observing satellite.
The AVNIR-2 is a successor to ADEOS/AVNIR, which was a four-band
optical sensor launched in 1996. The AVNIR-2 has almost same optics and
25
configuration as the AVNIR. The main modification is detectors and following
electronics. These changes are given to achieve spatial resolution compared to 16
m of the AVNIR. Another modification from the AVNIR is pointing capability,
which is ± 44 degree from nadir in cross-track direction. Its flexible pointing
ability realizes frequent observation, e.g. every 48 hours in higher latitude area
(Osawa, 2004). AVNIR-2 is a visible and near infrared radiometer for observing
land and coastal zones. It provides better spatial land-coverage maps and land-use
classification maps for monitoring regional environments with 10 meters spatial
resolution (JAXA, 2007) and can be used to PANSHARPEN higher resolution
PRISM data (PCI Geomatics, 2005). Table 2.1 presents the ALOS characteristics.
Table 2.1. ALOS characteristics
Number of bands 4
Wavelength
Band 1 : 0.42 to 0.50 micrometers Band 2 : 0.52 to 0.60 micrometers Band 3 : 0.61 to 0.69 micrometers Band 4 : 0.76 to 0.89 micrometers
Spatial Resolution 10m (at Nadir) Swath Width 70km (at Nadir)
S/N >200
MTF Band 1 through 3 : >0.25 Band 4 : >0.20
Number of Detectors 7000/band Pointing Angle - 44 to + 44 degree
Bit Length 8 bits Source: JAXA, 2007
On the report from Gonga-Saholiariliva et al. (2008), carrying out
comprehensive inventories and synoptic maps of water resources over large areas,
ALOS/AVNIR-2 images seem to represent a good compromise in terms of
resolution, computational tractability, data availability, geographic coverage, ease
26
of result interpretation and cost compared with other satellite image data
commonly employed for the same applications. According to PCI Geomatics
(2005), ALOS/AVNIR-2 can used large scale map creation, large scale city
planning, agriculture (crop identification etc.), forest management, coastal
management, gulf pollution control, vegetation monitoring, and large scale flood
monitoring
2.5.2. ASTER
The Advanced Spaceborne Thermal Emission and Reflection Radiometer
(ASTER) is an advanced multispectral imager that was launched on board
NASA’s Terra spacecraft in December, 1999. ASTER covers a wide spectral
region with 14 bands from the visible to the thermal infrared with high spatial,
spectral and radiometric resolution. An additional backward-looking near-infrared
band provides stereo coverage. The spatial resolution varies with wavelength: 15
m in the visible and near-infrared (VNIR), 30 m in the short wave infrared
(SWIR), and 90 m in the thermal infrared (TIR). Each ASTER scene covers an
area of 60 x 60 km (Abrams and Hook, 2002)
ASTER monitors cloud cover, glaciers, land temperature, land use, natural
disasters, sea ice, snow cover and vegetation patterns at a spatial resolution of 90
to 15 meters. The multispectral images obtained from this sensor have 14 different
colors, which allow scientists to interpret wavelengths that cannot be seen by the
human eye, such as near infrared, short wave infrared and thermal infrared.
27
ASTER data is expected to contribute to a wide array of global change-
related application areas, including vegetation and ecosystem dynamics, hazard
monitoring, geology and soils, land surface climatology, hydrology, land cover
change, and the generation of digital elevation models (DEMs) (Department of
Geography, University of Maryland, 2004). Figure 2.8 and Table 2.2 present the
ASTER satellite and the characteristics of the 3 ASTER Sensor Systems.
Fig. 2.8. The ASTER satellite
28
Table 2.2. Characteristic of the 3 ASTER Sensor Systems
Sub system
Band No.
Spectral Range (μm)
Spatial Resolution, m
Quantization Levels
VNIR
1 2
3N 3B
0.52-0.60 0.63-0.69 0.78-0.86 0.78-0.86
15 8 bits
SWIR
4 5 6 7 8 9
1.60-1.70 2.145-2.185 2.185-2.225 2.235-2.285 2.295-2.365 2.360-2.430
30 8 bits
TIR
10 11 12 13 14
8.125-8.475 8.475-8.825 8.925-9.275 10.25-10.95 10.95-11.65
90 12 bits
Source: Abrams and Hook, 2002
2.6. Geographic Information Systems (GIS)
Geographical Information System (GIS) is computer-based system that
enable user to collect, store, process, analyze and present spatial data (Prakash,
2001). According to Burrough (1986), GIS is disciplines are attempting the some
sort of operation to develop a powerful set of tools for collecting, storing,
retrieving at will, transforming, and displaying spatial data from the real world for
a particular set of purposes. As states by Star and Estes (1990), GIS is an
information system that is designed to work with data referenced by spatial or
geographic coordinates. In the other words, GIS is both a database system with
specific capabilities for spatially-referenced data, as well as a set of operations for
working with the data.
29
A GIS can be either manual (sometimes called analog) or automated (that is,
based on a digital computer). Manual GIS usually comprise several data elements
including maps, sheets of transparent material used a overlays, aerial and ground
photographs, statistical report, and field survey reports (Star and Estes, 1990).
Aronoff (1989) in Raju (2005), defined the automatically GIS as a computer-
based system that provides the following four sets of capabilities to handle geo-
referenced data: input, data management (data storage and retrieval), manipulation
and analysis, and output.
DeBy et al. (2004) distinguished three important stages of working with
geographic data, that is (1) data preparation and entry: the early stage which data
about study phenomenon is collected prepared to be entered to the system, (2)
data analysis: the middle stage in which collected data is carefully reviewed, and,
for instance, attempts are made to discover patterns, and (3) data presentation: the
final stage in which the result earlier analysis are presented in an appropriate way.
The basic data types in a GIS reflect traditional data found on a map.
Accordingly, GIS technology utilizes two basic types of data. These are (1)
Spatial data: spatial data describes the absolute and relative location of geographic
features and (2) Attribute data: attribute data describes characteristics of the
spatial features. These characteristics can be quantitative and/or qualitative in
nature. Attribute data is often referred to as tabular data (Prakash, 2001). The
coordinate location of a forestry stand would be spatial data, while the
characteristics of that forestry stand, e.g. cover group, dominant species, crown
closure, and height, would be attribute data.
30
GIS help in such study because they represent these phenomena digitally in
a computer. The digital representation can be subjected to analytic function
(computation) in the GIS (DeBy et al., 2004). To represent these phenomena we
need to do conversion of real world geographical variation into discrete objects
which is done through data models. It represents the linkage between the real
world domain of geographic data and computer representation of these features
for representing the spatial information (Raju, 2005). In this way, the content of a
spatial database is a model of the earth (Star and Ester, 1990).
GIS has two structure data to show the data model, Raster and Vector (Raju,
2005). In raster type of representation of the geographical data, a set of cells
located by coordinate is used; each cell is independently addressed with the value
of an attribute. Each cell contains a single value and every location corresponds to
a cell. One set of cell and associated value is a layer. Raster models are simple
with which spatial analysis is easier and faster. Raster data models require a huge
volume of data to be stored, fitness of data is limited by cell size and output is less
beautiful.
A vector based GIS is defined by the vectorial representation of its
geographic data. Base on the characteristics of this data model, geographic objects
are explicitly represented and, within the spatial characteristics, the thematic
aspects are associated. Vector data is comprised of lines or arcs, defined by
beginning and end points, which meet at nodes. The locations of these nodes and
the topological structure are usually stored explicitly. Features are defined by their
boundaries only and curved lines are represented as a series of connecting arcs
31
(Escobar, 2002). Vector data structures are based on elemental points whose
locations are known to arbitrary precision, in contrast to the raster or cellular data
structures we have described (Star and Ester, 1990). According to As-syakur
(2005) two model of structure data each owning excess and weakness. Analysis
using structure data raster can take a short cut time, however information
presented in attribute data not as complete as like analysis with vector data
structure.
A GIS makes it possible to link, or integrate, information that is difficult to
associate through any other means. Thus, a GIS can use combinations of mapped
variables to build and analyze new variables (Fig. 2.9) (USGS, 2007). For
example, using GIS technology, it is possible to combine agricultural records with
hydrography data to determine which streams will carry certain levels of fertilizer
runoff. Agricultural records can indicate how much pesticide has been applied to a
parcel of land. By locating these parcels and intersecting them with streams, the
GIS can be used to predict the amount of nutrient runoff in each stream. Then as
streams converge, the total loads can be calculated downstream where the stream
enters a lake.
32
Fig. 2.9. Data integration is the linking of information in different forms through a
GIS (USGS, 2007).
According to Prakash (2001), a GIS references these real-world spatial data
elements to a coordinate system. These features can be separated into different
layers. A GIS system stores each category of information in a separate "layer" for
ease of maintenance, analysis, and visualization. For example, layers can
represent terrain characteristics, census data, demographics information,
environmental and ecological data, roads, land use, river drainage and flood plains,
and rare wildlife habitats (Fig. 2.10). Different applications create and use
different layers. A GIS can also store attribute data, which is descriptive
information of the map features. This attribute information is placed in a database
separate from the graphics data but is linked to them. A GIS allows the
examination of both spatial and attribute data at the same time. Also, a GIS lets
users search the attribute data and relate it to the spatial data. Therefore, a GIS can
combine geographic and other types of data to generate maps and reports,
33
enabling users to collect, manage, and interpret location-based information in a
planned and systematic way.
Depending on the interest of a particular application, a GIS can be
considered to be a data store (application of a spatial database), a tool- (box), a
technology, an information source or a science (spatial information science) (Raju,
2005).
Fig. 2.10. The concept of layers (ESRI, 1996)
Data for GIS applications includes from digitized and scanned data,
databases, GPS field sampling of attributes, and from remote sensing and aerial
photography (Escobar et al., 2002). Remote sensing often requires other kinds of
ancillary data to achieve both its greatest value and the highest levels of accuracy
34
as a data and information production technology (Star and Ester, 1990). Remote
sensing image analysis systems and geographic information systems (GIS) show
great promise for the integration of a wide variety of spatial information as a
support to tasks such as urban and regional planning, natural resources
management, agricultural studies, topographic and thematic mapping, civil
engineering, hydrology studies, and geological exploration (Ehlers, 1992).
35
CHAPTER III
FRAMEWORK OF RESEARCH
Research framework has been designed based on the research background,
which used satellite data and GIS to estimate GPP in urban area, Denpasar.
Vegetation index is obtained from ALOS/AVNIR-2 and ASTER, Land use map
from Environmental Management Agency (Bappedalda) of Denpasar, and
Climatology data from Indonesian Meteorology and Geophysics Agency (BMG).
Satellite is employed to obtain vegetation index value, when index vegetation
correlate with fAPAR, and can use for estimation GPP. GIS is used to analyze
spatial data and to obtain map of GPP. The framework of research is shown in
Figure 3.1.
Satellite Data Land Use Map Climatology Data
ALOS/ AVNIR-2
ASTER
GIS Analysis
GPP Map in Urban Area, Denpasar - All Area
- By Land Use - By Sub Regency
Fig. 3.1. Framework of Research
35
36
CHAPTER IV
RESEARCH METHODS
4.1 Research Location
The research location is in Denpasar City. Geographically, Denpasar located
in 08o36’56”S – 08o42’01”S and 115o10’23”E – 115o16’27”E, and it is located
between two neighborhood regency namely Badung and Gianyar Regencies. It
has coasts at eastern and southern part and it also has an island which is separated
from Bali Island, called Serangan Island. Administratively, Denpasar consist of 4
districts by 43 country side, those are: South Denpasar District, East Denpasar
District, West Denpasar District, and North Denpasar District (Fig. 4.1). Denpasar
is surrounded by: North and West; Badung Regency, East; Gianyar Regency, and
South; Badung Strait. Denpasar has type of soil called latosol type. The condition
of oblique topography of Denpasar city is from north to south. The height level is
0-75 m above sea level, while the inclination slope morphology is from 0-5 %.
Denpasar has tropical climate with monthly mean temperature around 24-32 oC
and monthly mean precipitation around 13-358 mm. Land use cover of Denpasar
are forest (mangrove), perennial plant, settlement, rice field, shrub, dry land, and
bare land.
36
37
Denpasar
Serangan Island
Fig. 4.1. Study site
4.2 Research Materials
Materials used in this research are as follows:
1. Digital image of Denpasar area in 13 October 2006 from ALOS/AVNIR-2
(Fig. 4.2).
2. Digital image of Denpasar area in 5 September 2006 from ASTER (Fig.
4.2).
3. Land use map image of Denpasar area in 2006 from Quick Bird (Fig. 4.3)
4. Topography map 1 : 25.000 region of Denpasar from Bakosurtanal (2000)
5. Solar radiation data from Indonesian Meteorology and Geophysics
Agency (BMG)
6. Quick Bird image of Denpasar area in 2006
38
(a) (b)
Fig. 4.2. Digital image of ALOS/AVNIR-2 (a) and ASTER/VNIR (b) of Denpasar area in 2006
Fig. 4.3. Land use map
39
4.3 Research Instruments
Research instruments that are utilized for this research are as follows:
1. Computer Intel Pentium IV 3.0 GHz, RAM 1064 MB, Hard-disk 160 GB,
Keyboard 104 keys, and Mouse.
2. Remote Sensing software, the software used ENVI 4.4.
3. GIS software for analysis spatial data, the software used ArcView GIS 3.2
and the extensions.
4. Microsoft Office Excel 2003 for analysis attribute data
4.4 Data Source
First of all some references and materials were collected. The theoretical
base references and previous study are important to enhance the information. The
materials of this research are ALOS/AVNIR-2 and ASTER/VNIR digital image in
2006, land use map, topography map, and solar radiation data.
Ground check is used to determine specific ordinate to truth evidence of
field research. This process is conducted to observe the land cover vegetation.
40
4.5 Data Processing
4.5.1 Pre processing phase
In this phase, there are four processes which should be done as fallows; (1)
geometric correction, (2) radiometric correction, (3) cropping, and (4) conversion
image data to raster data.
1. Geometric correction
The remote sensing image has various geometric distortions. This problem
is inherent in remote sensing, as attempt to accurately convert the three-
dimensional surface of the Earth as a two-dimensional image. There are
two main processes which will be done in geometric corrections, those are
transformation and resampling process. Topography map whith the scale
1:25.000 is used in transformation process, with Universal Transverse
Mercator (UTM) projection as reference.
At least four pairs of field control points are needed to be analyze of two
order of polynomial transformation process with the root mean squares
(RMS) and it’s required by the National Map Accuracy Standard (NMAS)
it’s less of the same with the 1.7 pixels. In the geometric correction
process under ENVI 4.4.
2. Radiance correction
Pixel values in commercially available satellite imagery represent the
radiance of the surface in the form of Digital Numbers (DN) which is
calibrated to fit a certain range of values. Conversion of DN to absolute
41
radiance values is a necessary procedure for comparative analysis of
several images taken by different sensors. Since each sensor has its own
calibration parameters used in recording the DN values, the same DN
values in two images is taken by two different sensors may represent two
different radiance values.
The spectral radiance for particular band was calculated using equation (4)
(Abrams and Hook, 2002):
Lsat = (DN + a) * UCC (4)
Where:
Lsat = at-sensor spectral radiance
DN = digital number (the pixel values in the original image data files)
a = absolute calibration coefficients that are contained in the
Ancillary record of the leader file: 0 for ALOS/AVNIR-2 and -1 for Aster
satellite data
UCC = Unit Conversion Coefficient. This is different for each image
band, and also depends on the gain setting that was used to acquire the
image. UCC for ALOS/AVNIR-2 and Aster are presented in Table 3.1.
42
Table 3.1. Unit conversion coefficient for each band
Band No.
ALOS/AVNIR-2 Aster
ALOS/AVNIR-2
ALAV2A038223770 Aster
1
2
3
4
1
2
3N
-
0.5880
0.5730
0.5020
0.8350
0.676
0.708
0.862
-
Source: LED-ALAV2A038223770-O1B2R_U (2006), Abrams and Hook (2002)
3. Cropping
Cropping is done to form the boundary of area which is used in this
research according to land use map form Quikbird image. Image subset
function in ENVI 4.4 is used to crop the image.
4. Conversion image to raster file
ArcView GIS software can not used image file for analysis and modeling
the data, but Grid file format. The format of image should be converted
from image file format to Grid file format. This process only conversion
the format data, but not value of the data. Save file as to ESRI GRID
function in ENVI 4.4 is used to convert image.
4.5.2 Digital Spatial Analysis
The carbon budget consists of several major processes that describe the
exchange of carbon dioxide between terrestrial ecosystems and the atmosphere.
GPP is the total carbon assimilated by vegetation (Ibrahim, 2006). Satellite remote
43
sensing provides consistent and systematic observations of vegetation and has
played an increasing role in the characterization of vegetation structure and
estimated GPP of vegetation (Xiao et al., 2004). GPP estimation was calculated
using eq. 3.
GPP = ε × fAPAR × PAR = ε × APAR
PAR is actually restricted to just a portion of sunlight's spectrum from 400
to 700 nanometers (nm) which is comparable to the visible range of light that
could be seen by human eye (Horning, 2004). The value of PAR is assumed to be
approximately 0.5 of the incoming solar radiation (Sims et al., 2006; Slamet and
Haryanto, 2006), solar radiation data is got from Indonesian Meteorology and
Geophysics Agency (BMG). Fraction of absorbed photosynthetically active
radiation (fAPAR) by the vegetation cover is related to the NDVI, NDVI has been
widely used for the remote estimation of fAPAR because of its positive linear
relationship with fAPAR (Myneni and Williams, 1994), Ochi and Shibasaki
(1999) tabulated various relationships between fAPAR and NDVI in some Asian
countries, they recommended the relationships is:
fAPAR = -0.08 + 1.075 NDVI (5)
Light use efficiency (ε) (LUE) is a biomespecific value representing
optimal potential of the vegetation for converting PAR to GPP, Light use
efficiency values are similar for all plant types and biomes (Horning, 2004).
Estimation of LUE has, however, proven more problematic. Light use efficiency
can be estimated from mechanistic models based on leaf biochemistry and
44
micrometeorological parameters but these models are complex and generally
require many parameters that cannot directly be estimated from remote sensing
(Running et al., 1999 in Sims et al., 2006). Light use efficiency may be assumed
to be constant under non stressed conditions, but it is affected by stresses,
phonological stages, and the physical environment (Inoue et al., 2008). Ochi and
Shibasaki (1999) recommended the value of ε is 1.5 gC MJ-1 in some Asian
countries. The result of this model is compared with GPP which derive from
MODIS GPP product (MOD17) in Denpasar area derived from
http://daac.ornl.gov/MODIS/modis.html.
Analyses were carried out using ArcView GIS (version 3.2) software with
Spatial Analyst Extensions, including in ArcView GIS (version 3.2) software.
Map calculator function used to spatial calculation analyst (Fig. 4.4).
45
Fig. 4.4. Research Schema
4.5.3 Data presentation
The last results in this research are maps and digital data. For many types
of geographic operations, the end result is best visualized as a map or graph. Maps
are very efficient to communicating geographic information. GIS provides new
and exciting tools to extend the art and science of cartography. Map displays
could be integrated with reports, three-dimensional views, photographic images,
and with multimedia. According to DeBy et al. (2004), the characteristic of map
and their function in relation to the spatial data, in this context the cartographic
visualization process is considered to be the translation or conversion of spatial
data from a database into graphics. Cartographic methods and techniques are
applied during the visualization process. These can be considered to form some
AALLOOSS//AAVVNNIIRR--22 AASSTTEERR//VVNNIIRR
CClliimmaattoollooggyy ddaattaa
LLUUEE CCooeeffffiicciieenntt
GGeeoommeettrriicc CCoorrrreeccttiioonn
CCrrooppppiinngg IImmaaggee
RReedd BBaanndd
NNIIRR BBaanndd
SSppaattiiaall CCaallccuullaattiioonn wwiitthh NNDDVVII eeqquuaattiioonn
SSppaattiiaall CCaallccuullaattiioonn wwiitthh GGPPPP eeqquuaattiioonn
IInnddeexx VVeeggeettaattiioonn MMaapp
TToo
GGPPPP MMaapp
ppooggrraapphhyy MMaapp
Software Use: Process
Input/Output
Where:
EENNVVII 44..44 AArrccVViieeww 33..22
RRaaddiiaannccee CCoorrrreeccttiioonn
CCoonnvveerrssiioonn ttoo GGRRIIDD BByy ssppeeccttrraall rreessoolluuttiioonnss
SSPPAATTIIAALL AANNAALLYYSSTT
46
kind of grammar that allows for the optimal design, production and use of maps,
depend of the application. Output map with GIS application as according to map
consumer desire, for example colors and map scale (As-syakur, 2005). Layout
function in ArcView GIS is used to create a map (Fig. 4.5).
Fig. 4.5. GIS Process
LLaanndd UUssee MMaapp
RRAASSTTEERR DDAATTAA SSTTRRUUCCTTUURREE VVEECCTTOORR DDAATTAA SSTTRRUUCCTTUURREE
CCoonnvveerrssiioonn VVeeccttoorr ttoo
rraasstteerr
LLaanndd UUssee MMaapp SSaatteelllliittee
DDaattaaSSuubb DDiissttrriicctt MMaapp SSuubb DDiissttrriicctt MMaapp
RRiivveerr MMaapp
RRooaadd MMaapp
OOvveerrllaayy wwiitthh ssppaattiiaall ccaallccuullaattiioonn
GGPPPP bbyy LLaanndd UUssee GGPPPP bbyy SSuubb DDiissttrriicctt
GGPPPP MMaapp
OOvveerrllaayy wwiitthh ssppaattiiaall ccaallccuullaattiioonn
OOvveerrllaayy bbyy LLaayyeerr
GGPPPP MMaapp iinn DDeennppaassaarr GGPPPP MMaapp bbyy SSuubb DDiissttrriicctt
VViillllaaggee MMaapp
GGPPPP MMaapp bbyy LLaanndd UUssee
47
CHAPTER V
RESULTS
5.1 Total GPP
The estimation of GPP in this study has been carried out using Light Use
Efficiency approach (LUE Model). Annual values of GPP in land vegetation local
area, Denpasar were estimated using ALOS/AVNIR-2 and Aster data.
The GPP annual value is different with differences satellite data. Annual
value of GPP in ALOS/AVNIR-2 varies from 0.13 gC m-2 yr-1 to 2586.18 gC m-2
yr-1, mean value of annual GPP is 836.23 gC m-2 yr-1. In Aster satellite, minimum
value of GPP is 0.14 gC m-2 yr-1 and maximum value is 2595.26 gC m-2 yr-1 with
mean value of GPP is 776.83 gC m-2 yr-1. Totally GPP per year in Denpasar from
ALOS/AVNIR-2 is 52421.46 tC yr-1 with the area of 6267.56 ha and analysis with
Aster satellite data, total GPP in Denpasar per year is 59355.493 tC yr-1 with the
area of 7647.84 ha (Table 5.1 and Table 5.2). The GPP pixel value distribution
from ALOS/AVNIR-2 satellite data is dominated by low pixel value (< 250 gC m-
2 yr-1) with the area of 1236.62 ha, which is become decrease until the GPP pixel
value is high (2250 – 2587 gC m-2 yr-1) with the area of 17.17 ha. In the other
case, the GPP pixel value distribution from Aster satellite data is dominated by
low pixel value (< 250 gC m-2 yr-1) with the area of 1694.57 ha, which is become
decrease until the GPP pixel value is in high range (2250 – 2595 gC m-2 yr-1) with
the area of 6.59 ha (Table 5.2).
47
48
Table 5.1. Annual and total value of GPP with differences satellite data
GPP (gC m-2 yr-1) Satellite Data
Max Mean Min Std. Dev.
Total GPP
tC yr-1
ALOS/AVNIR-2 2586.18 836.23 0.13 583.51 52421.46
Aster 2595.26 776.83 0.14 565.03 59355.49
Table 5.2. Total pixels and hectarage of annual GPP value with differences satellite data
ALOS/AVNIR-2 ASTER GPP value
(gC m-2 yr-1) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 123662 1236.62 75314 1694.57
250 - 500 101694 1016.94 59523 1339.27
500 - 750 88378 883.78 49242 1107.95
750 - 1000 76929 769.29 42346 952.79
1000 - 1250 70423 704.23 36175 813.94
1250 - 1500 61249 612.49 30336 682.56
1500 - 1750 52544 525.44 24785 557.66
1750 - 2000 34440 344.40 14900 335.25
2000 - 2250 15720 157.20 6990 157.28
> 2250 1717 17.17 293 6.59
Total 6267.56 7647.84
The map distribution and graphic total pixels distribution value of annual
GPP from ALOS/AVNIR-2 and Aster is shown in Figure 5.1 and Figure 5.2.
49
250000
200000
< 500
500 - 1000150000
1000 - 1500
100000 1500 - 2000
50000
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
Fig 5.1. Annual distribution of GPP from ALOS/AVNIR-2
140000
120000
100000 < 500
500 - 100080000
1000 - 150060000 1500 - 2000
40000
20000
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
Fig 5.2. Annual distribution of GPP from Aster
50
The maximum value of GPP from two satellite data is smaller than the
2707.8 gC m-2 yr-1 is the maximum GPP value derived from MODIS GPP product
(MOD17) in Denpasar area (Fig. 5.3), and smaller than the 2859 gC m-2 yr-1
measured over a tropical peat swamp forest in central Kalimantan-Indonesia
(Hirano et al., 2005) and smaller than to the 3040 gC m-2 yr-1 measured over a
tropical forest in central Amazonia, Brazil (Malhi et al., 1998).
Fig 5.3. Annual distribution of GPP from MODIS product (MOD17)
5.2 The Annual GPP by Land Use in Denpasar
The different land use will effect the different of annual GPP value. In
ALOS/AVNIR-2 satellite data, the maximum value of annual GPP is come from
51
ricefield land use which the value is 2586.18 gC m-2 yr-1 and the minimum value
of annual GPP is 0.13 gC m-2 yr-1 from all land use (Figure 5.4 and Table 5.3).
0
500
1000
1500
2000
2500
3000S
ettle
men
t
Ric
efie
ld
Fore
st(M
angr
ove)
Shr
ub
Per
enni
alpl
ant
Dry
land
Bar
elan
d
Land use
GP
P (g
C/m
2/yr
)
MaxMeanMin
Fig. 5.4. Graphic of annual GPP value with different land use in ALOS/AVNIR-2 satellite data
Table 5.3. Annual value of GPP with differences land use in ALOS/AVNIR-2 satellite data
GPP (gC m-2 yr-1) Land use
Max Mean Min Std. Dev.
Settlement 2511.43 540.49 0.13 440.81
Ricefield 2586.18 1030.08 0.13 618.32
Forest (Mangrove) 2501.92 1123.58 0.13 584.47
Shrub 2427.54 882.12 0.13 589.50
Perennial plant 2456.39 1034.77 0.13 508.46
Dryland 2414.05 893.46 0.13 522.76
Bareland 2489.12 771.56 0.13 531.93
In Aster satellite data, the maximum value of annual GPP is come from
forest (mangrove) land use which the value is 2595.26 gC m-2 yr-1 and the
52
minimum value of annual GPP is from all land use, which value is 0.14 gC m-2 yr-
1 (Figure 5.5 and Table 5.4).
0.00
500.00
1000.00
1500.00
2000.00
2500.00
3000.00S
ettle
men
t
Ric
efie
ld
Fore
st(M
angr
ove)
Shr
ub
Per
enni
alpl
ant
Dry
land
Bar
elan
d
Land use
GP
P (g
C/m
2/yr
)
MaxMeanMin
Fig. 5.5. Graphic of annual GPP value with different land use in Aster satellite data
Table 5.4. Annual value of GPP with differences land use in Aster satellite data
GPP (gC m-2 yr-1) Land use
Max Mean Min Std. Dev.
Settlement 2353.91 492.44 0.14 408.62
Ricefield 2371.86 1020.65 0.14 605.52
Forest (Mangrove) 2595.26 1177.40 0.14 561.12
Shrub 2305.14 794.37 0.14 556.54
Perennial plant 2257.76 989.24 0.14 489.77
Dryland 2261.80 830.61 0.14 453.75
Bareland 2244.33 648.17 0.14 450.25
The totally of annual value of GPP with difference land use in ALOS and
Aster satellite imagery is shown in Table 5.5 and Fig. 5.6.
53
Table 5.5. The totally of annual value of GPP with differences land use in ALOS and Aster satellite
GPP (tC yr-1) Land Use Hectarage
ALOS Aster
Settlement 7179.17 12675.23 15992.84
Ricefield 2616.34 20254.15 22571.65
Forest (Mangrove) 700.69 6255.51 7081.16
Shrub 81.10 469.55 460.96
Perennial plant 961.75 8300.74 8567.52
Dryland 263.26 1888.87 1930.72
Bareland 827.39 2577.41 2750.65
Total 12629.70 52421.46 59355.49
0
5000
10000
15000
20000
25000
Settlem
ent
Ricefie
ld
Forest
(Man
grove
)
Shrub
Perenn
ial pl
ant
Drylan
d
Bareland
ALOSAster
Fig. 5.6. Totally of annual value of GPP with differences land use in ALOS and Aster satellite data
The GPP pixel value distribution in settlement land use from
ALOS/AVNIR-2 satellite data is dominated by low pixel value (< 250 gC m-2 yr-
1) with the area of 763.94 ha, which is become decrease until the GPP pixel value
is high (> 2250 gC m-2 yr-1) with the area of 0.92 ha. In the other case, the GPP
54
pixel value distribution from Aster satellite data is dominated by low pixel value
(< 250 gC m-2 yr-1) with the area of 1164.56 ha, which is become decrease until
the GPP pixel value is in high range (> 2250 gC m-2 yr-1) with the area of 0.07 ha
(Table 5.6). GPP map distribution and graphic total pixels distribution from
ALOS/AVNIR-2 and Aster satellite data in settlement land use is shown in Figure
5.7.
Table 5.6. Total pixels and hectarage of annual GPP value with differences
satellite data in settlement land use
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 76394 763.94 51758 1164.56
250 - 500 54189 541.89 34817 783.38
500 - 750 39311 393.11 23510 528.98
750 - 1000 26042 260.42 15506 348.89
1000 - 1250 18072 180.72 9532 214.47
1250 - 1500 11124 111.24 5625 126.56
1500 - 1750 6181 61.81 2655 59.74
1750 - 2000 2209 22.09 807 18.16
2000 - 2250 561 5.61 127 2.86
> 2250 92 0.92 3 0.07
Total 2341.75 3247.65
55
Fig. 5.7. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in settlement land use
The GPP pixel value distribution in ricefield land use from ALOS/AVNIR-2
satellite data is dominated by pixel with value 0.13 gC m-2 yr-1 until 1750 gC m-2
yr-1 with the area of 1658.60 ha, which is become decrease until the GPP pixel
value is high (> 2250 gC m-2 yr-1) with the area of 7.97 ha. In the other case, the
GPP pixel value distribution from Aster satellite data is dominated by pixel with
value 0.144 - 1750 gC m-2 yr-1 with the area of 1897.61 ha. which is become
decrease until the GPP pixel value is high (> 2250 gC m-2 yr-1) with the area of
90000
80000
70000
60000
50000
40000
30000
20000
10000
0
< 500
500 - 1000
1000 - 1500
1500 - 2000
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
b)a)
d)c)140000
120000
100000 < 500
500 - 100080000
1000 - 150060000 1500 - 2000
40000
20000
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
56
2.66 ha (Table 5.7). GPP map distribution and GPP graphic distribution from
ALOS/AVNIR-2 and Aster satellite data with ricefield land use is shown in
Figure 5.8.
Table 5.7. Total pixels and hectarage of annual GPP value with differences
satellite data in ricefield land use
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 26097 260.97 12556 282.51
250 - 500 24299 242.99 12149 273.35
500 - 750 23188 231.88 11993 269.84
750 - 1000 23017 230.17 12133 272.99
1000 - 1250 23054 230.54 12009 270.20
1250 - 1500 22422 224.22 11428 257.13
1500 - 1750 23803 238.03 12070 271.58
1750 - 2000 19997 199.97 8803 198.07
2000 - 2250 9936 99.36 5030 113.18
> 2250 797 7.97 118 2.66
Total 1966.10 2211.50
57
b)a)
d)c) 2500060000
50000 20000
< 500< 50040000
500 - 100015000500 - 1000
Fig. 5.8. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map
distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and (d) GPP graphic distribution from Aster in ricefield land use
The GPP pixel value distribution in forest (mangrove) land use from
ALOS/AVNIR-2 satellite data is dominated by pixel with value 500 gC m-2 yr-1
until 1750 gC m-2 yr-1 with the area of 365.86 ha. In the other case, the GPP pixel
value distribution from Aster satellite data is dominated by pixel with value 500
gC m-2 yr-1 until 2000 gC m-2 yr-1 with the area of 477.93 ha (Table 5.8). GPP map
distribution and GPP graphic distribution from ALOS/AVNIR-2 and Aster
satellite data with forest (mangrove) land use is shown in Figure 5.9.
30000
20000
10000
0
1000 - 1500
1500 - 2000
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
1000 - 1500
10000 1500 - 2000
5000
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
58
Table 5.8. Total pixels and hectarage of annual GPP value with differences satellite data in forest (mangrove) land use
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 4513 45.13 1814 40.82
250 - 500 5245 52.45 2020 45.45
500 - 750 6144 61.44 2667 60.01
750 - 1000 7718 77.18 3500 78.75
1000 - 1250 8123 81.23 4071 91.60
1250 - 1500 7523 75.23 4029 90.65
1500 - 1750 7078 70.78 3793 85.34
1750 - 2000 5364 53.64 3181 71.57
2000 - 2250 3358 33.58 1490 33.53
> 2250 602 6.02 164 3.69
Total 556.68 601.40
59
b)a)
d)
Fig. 5.9. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in forest (mangrove) land use
The GPP pixel value distribution in shrub land use from ALOS/AVNIR-2
satellite data is dominated by pixel with value < 250 gC m-2 yr-1 – 1250 gC m-2 yr-
1 with the area of 38.81 ha, which is become decrease until the GPP pixel value is
high (> 2250 gC m-2 yr-1) with the area of 0.59 ha. In the other case, the GPP pixel
value distribution from Aster satellite data is dominated by pixel with value < 250
gC m-2 yr-1 – 1250 gC m-2 yr-1 with the area of 45.70 ha, which is become decrease
until the GPP pixel value is high (> 1250 gC m-2 yr-1) with the area of 0.14 ha
(Table 5.9). GPP map distribution and GPP graphic distribution from
16000
14000
12000
10000
8000
6000
4000
2000
0
< 500
500 - 1000
1000 - 1500
1500 - 2000
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
9000c) 8000
7000< 500
6000500 - 1000
50001000 - 1500
40001500 - 2000
3000
2000
1000
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
60
ALOS/AVNIR-2 and Aster satellite data in shrub land use is shown in Figure
5.10.
Table 5.9. Total pixels and hectarage of annual GPP value with differences
satellite data in shrub land use
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 867 8.67 481 10.82
250 - 500 845 8.45 479 10.78
500 - 750 798 7.98 409 9.20
750 - 1000 668 6.68 388 8.73
1000 - 1250 703 7.03 274 6.17
1250 - 1500 567 5.67 217 4.88
1500 - 1750 373 3.73 136 3.06
1750 - 2000 210 2.10 93 2.09
2000 - 2250 230 2.30 96 2.16
> 2250 59 0.59 6 0.14
Total 53.20 58.03
61
10001800
9001600
8001400< 500700< 500
1200500 - 1000600500 - 1000
Fig. 5.10. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in shrub land use
The GPP pixel value distribution in perennial plant land use from
ALOS/AVNIR-2 satellite data is dominated by pixel with value 500 gC m-2 yr-1
until 1750 gC m-2 yr-1 with the area of 600.17 ha. In the other case, the GPP pixel
value distribution from Aster satellite data is dominated by pixel with value 250
gC m-2 yr-1 until 1750 gC m-2 yr-1 with the area of 754.27 ha (Table 5.10). GPP
map distribution and GPP graphic distribution from ALOS/AVNIR-2 and Aster
satellite data in perennial plant land use is shown in Figure 5.11.
500
400
300
200
100
0
1000 - 1500
1500 - 2000
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
10001000 - 1500
8001500 - 2000
600
400
200
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
62
Table 5.10. Total pixels and hectarage of annual GPP value with differences satellite data in perennial plant land use
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 6492 64.92 3202 72.05
250 - 500 8070 80.70 4415 99.34
500 - 750 10267 102.67 5169 116.30
750 - 1000 11601 116.01 5972 134.37
1000 - 1250 13300 133.00 6672 150.12
1250 - 1500 13784 137.84 6525 146.81
1500 - 1750 11065 110.65 4770 107.33
1750 - 2000 4689 46.89 1562 35.15
2000 - 2250 892 8.92 204 4.59
> 2250 53 0.53 1 0.02
Total 802.13 866.07
63
b)a)
d)
Fig. 5.11. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in perennial plant land use
The GPP pixel value distribution in dryland land use from ALOS/AVNIR-2
satellite data is dominated by pixel with value < 250 gC m-2 yr-1 until 1500 gC m-2
yr-1 with the area of 180.72 ha. which is become decrease until the GPP pixel
value is high (> 2250 gC m-2 yr-1) with the area of 0.35 ha. In the other case, the
GPP pixel value distribution from Aster satellite data is dominated by pixel with
value < 250 - 1500 gC m-2 yr-1 with the area of 212.65 ha, which is become
decrease until the GPP pixel value is high range (> 2250 gC m-2 yr-1) with the area
of 0.02 ha (Table 5.11). GPP map distribution and GPP graphic distribution from
14000c)30000
1200025000
10000 < 500< 50020000
500 - 1000500 - 1000 8000
6000
4000
2000
0
1000 - 1500
1500 - 2000
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
15000 1000 - 1500
1500 - 200010000
5000
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
64
ALOS/AVNIR-2 and Aster satellite data in dryland land use is shown in Figure
5.12.
Table 5.11. Total pixels and hectarage of annual GPP value with differences
satellite data in dryland land use
ASTER ALOS/AVNIR-2 GPP val
( Total p e (ha) Total pixel tarage (ha)
ue
grC/m^2/yr) ixels Hectarag s Hec
< 250 2708 27.08 1144 25.74
250 - 500
21 23
3037 30.37 1585 35.66
500 - 750 3232 32.32 1953 43.94
750 - 1000 3279 32.79 2017 45.38
1000 - 1250 3214 32.14 1590 35.78
1250 - 1500 2602 26.02 1162 26.15
1500 - 1750 1820 18.20 647 14.56
1750 - 2000 907 9.07 216 4.86
2000 - 2250 303 3.03 16 0.36
> 2250 35 0.35 1 0.02
Total 1.37 2.45
65
b)a)
d)
Fig. 5.12. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in dryland land use
The GPP pixel value distribution in bareland land use from ALOS/AVNIR-2
satellite data is dominated by pixel with value < 250 gC m-2 yr-1 until 750 gC m-2
yr-1 with the area of 179.37 ha, which is become decrease until the GPP pixel
value is high (> 2250 gC m-2 yr-1) with the area of 0.79 ha. In the other case, the
GPP pixel value distribution from Aster satellite data is dominated by pixel with
value < 250 gC m-2 yr-1 until 1000 gC m-2 yr-1 with the area of 328.97 ha, which is
become decrease until the GPP pixel value is in high range (2000 gC m-2 yr-1 –
2250 gC m-2 yr-1) with the area of 0.59 ha (Table 5.12). GPP map distribution and
4000
c)7000
35006000
3000< 5005000 < 500
2500 500 - 1000500 - 1000
2000
1500
1000
500
0
1000 - 1500
1500 - 2000
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
40001000 - 1500
30001500 - 2000
2000
1000
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
66
GPP graphic distribution from ALOS/AVNIR-2 and Aster satellite data in
bareland land use is shown in Figure 5.13.
Table 5.12. Total pixels and hectarage of annual GPP value with differences
satellite data in bareland land use
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 6556 65.56 4307 96.91
250 - 500 5973 59.73 4014 90.32
500 - 750 5408 54.08 3498 78.71
750 - 1000 4575 45.75 2802 63.05
1000 - 1250 3926 39.26 1996 44.91
1250 - 1500 3198 31.98 1324 29.79
1500 - 1750 2194 21.94 671 15.10
1750 - 2000 1047 10.47 223 5.02
2000 - 2250 440 4.40 26 0.59
> 2250 79 0.79 0 0.00
Total 333.96 424.37
67
b)a)
d)c)14000 9000
800012000
700010000 < 500 < 500
6000500 - 1000 500 - 1000
Fig. 5.13. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in bareland land use
5.3 The Annual GPP by Districts in Denpasar
In ALOS/AVNIR-2 satellite data, the maximum value of annual GPP in
South Denpasar district which the value is 2586.18 gC m-2 yr-1, in West Denpasar
district the maximum value of GPP is 2511.43 gC m-2 yr-1, in North Denpasar
district the maximum value of GPP is 2462.30 gC m-2 yr-1, and in East Denpasar
district the maximum value of GPP is 2449.19 gC m-2 yr-1 (Figure 5.14 and Table
5.13).
5000
4000
3000
2000
1000
0
1000 - 1500
1500 - 2000
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
80001000 - 1500
60001500 - 2000
4000
2000
0
2000 - 2600
Total carbon assimilated by vegetation (grC/m^2/yr)
Sum
up
pixe
l
68
0
500
1000
1500
2000
2500
3000
Sou
thD
enpa
sar
Wes
tD
enpa
sar
Nor
thD
enpa
sar
Eas
tD
enpa
sar
District
GPP
(gC/
m2/
yr)
MaxMeanMin
Fig. 5.14. Totally of annual value of GPP by district in ALOS/AVNIR-2 satellite data
Table 5.13. The totally of annual value of GPP by district in ALOS/AVNIR-2 satellite data
District Max Mean Min St. Dev. Total (tC/yr)
South Denpasar 2586.18 870.36 0.13 578.11 22965.17
West Denpasar 2511.43 767.55 0.13 601.86 6301.49
North Denpasar 2462.30 854.38 0.13 603.46 10137.18
East Denpasar 2449.19 802.05 0.13 562.93 12982.13
In Aster satellite data, the maximum value of annual GPP in South Denpasar
district which the value is 2595.26 gC m-2 yr-1, in West Denpasar district the
maximum value of GPP is 2289.78 gC m-2 yr-1, in North Denpasar district the
maximum value of GPP is 2304.26 gC m-2 yr-1, and in East Denpasar district the
maximum value of GPP is 2322.60 gC m-2 yr-1 (Figure 5.15 and Table 5.14).
69
0
200
400
600
800
1000
1200
1400
Sou
thD
enpa
sar
Wes
tD
enpa
sar
Nor
thD
enpa
sar
Eas
tD
enpa
sar
District
GP
P (g
C/m
2/yr
)
MaxMeanMin
Fig. 5.15. Totally of annual value of GPP by district in Aster satellite data
Table 5.14. The totally of annual value of GPP by district in Aster satellite data
District Max Mean Min St. Dev. Total (tC/yr)
South Denpasar 2595.26 831.15 0.14 580.84 25773.41
West Denpasar 2289.78 703.77 0.14 564.56 6873.89
North Denpasar 2304.26 731.22 0.14 538.93 11944.11
East Denpasar 2322.60 764.10 0.14 552.42 14747.20
The totally of annual value of GPP with differences land use in ALOS and
Aster satellite imagery is shown in Fig. 5.16 and Table 5.15.
70
0
5000
10000
15000
20000
25000
30000
South Denpasar
West Denpasar
North Denpasar
East Denpasar
Distric
GPP
(gC/
m2/
yr)
AvnirAster
Fig. 5.16. Totally of annual value of GPP by district in ALOS and Aster satellite data
Table 5.15. The totally of annual value of GPP by district in ALOS and Aster satellite data
Total (tC/yr) District
ALOS Aster
South Denpasar 22965.17 25773.41
West Denpasar 6301.49 6873.89
North Denpasar 10137.18 11944.11
East Denpasar 12982.13 14747.20
Total 52385.97 59338.61
The GPP pixel value distribution in South Denpasar district from
ALOS/AVNIR-2 satellite data is dominated by pixel with value < 250 gC m-2 yr-1
with the area of 461.38 ha, which is become decrease until the GPP pixel value is
high (> 2250 gC m-2 yr-1) with the area of 9.35 ha. In the other case, the GPP pixel
value distribution from Aster satellite data is dominated by low pixel value < 250
gC m-2 yr-1 with the area of 609.17 ha, which is become decrease until the GPP
71
pixel value is in high range (> 2250 gC m-2 yr-1) with the area of 5.72 ha (Table
5.16). GPP map distribution and GPP graphic distribution from ALOS/AVNIR-2
and Aster satellite data in South Denpasar district is shown in Figure 5.17.
Table 5.16. Total pixels and hectarage of annual GPP value with differences
satellite data in South Denpasar district
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 46138 461.38 27074 609.17
250 - 500 40015 400.15 22358 503.06
500 - 750 37129 371.29 19427 437.11
750 - 1000 34962 349.62 18151 408.40
1000 - 1250 32715 327.15 16097 362.18
1250 - 1500 27225 272.25 13197 296.93
1500 - 1750 22851 228.51 9930 223.43
1750 - 2000 14993 149.93 6703 150.82
2000 - 2250 6763 67.63 4627 104.11
> 2250 935 9.35 254 5.72
Total 2637.26 3100.91
72
Fig. 5.17. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in South Denpasar District
The GPP pixel value distribution in West Denpasar district from
ALOS/AVNIR-2 satellite data is dominated by pixel with value < 250 gC m-2 yr-1
with the area of 203.96 ha, which is become decrease until the GPP pixel value is
high (> 2250 gC m-2 yr-1) with the area of 5.20 ha. In the other case, the GPP pixel
value distribution from Aster satellite data is dominated by low pixel value <250
gC m-2 yr-1 with the area of 270.52 ha, which is become decrease until the GPP
pixel value is high (> 2250 gC m-2 yr-1) with the area of 0.11 ha (Table 5.17). GPP
map distribution and GPP graphic distribution from ALOS/AVNIR-2 and Aster
satellite data in West Denpasar district is shown in Figure 5.18.
50000
45000
40000
35000
30000
25000
20000
15000
10000
5000
0
< 500
500 - 1000
1000 - 1500
1500 - 2000
2000 - 2600
Sum
up
pixe
l
Total carbon assimilated by vegetation (grC/m^2/yr)
90000
80000
70000
60000
50000
40000
30000
20000
10000
0
< 500
500 - 1000
1000 - 1500
1500 - 2000
2000 - 2600
Sum
up
pixe
l
Total carbon assimilated by vegetation (grC/m^2/yr)
a) b)
d)c)
73
Table 5.17. Total pixels and hectarage of annual GPP value with differences satellite data in West Denpasar district
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 20396 203.96 12023 270.52
250 - 500 14845 148.45 8240 185.40
500 - 750 11374 113.74 6111 137.50
750 - 1000 8843 88.43 4480 100.80
1000 - 1250 7403 74.03 3710 83.48
1250 - 1500 6407 64.07 3169 71.30
1500 - 1750 5811 58.11 3181 71.57
1750 - 2000 3973 39.73 1940 43.65
2000 - 2250 2473 24.73 551 12.40
> 2250 520 5.20 5 0.11
Total 820.45 976.73
74
b)a)
d)c) 2500040000
3500020000
30000< 500< 500
25000 500 - 1000
Fig. 5.18. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in West Denpasar District
The GPP pixel value distribution in North Denpasar district from
ALOS/AVNIR-2 satellite data is dominated by pixel with value < 250 gC m-2 yr-1
with the area of 238.71 ha, which is become decrease until the GPP pixel value is
high (> 2250 gC m-2 yr-1) with the area of 1.09 ha. In the other case, the GPP pixel
value distribution from Aster satellite data is dominated by low pixel value <250
gC m-2 yr-1 with the area of 375.14 ha, which is become decrease until the GPP
pixel value is high (> 2250 gC m-2 yr-1) with the area of 0.29 ha (Table 5.18). GPP
15000
10000
5000
0
1000 - 1500
1500 - 2000
2000 - 2600
Sum
up
pixe
l
Total carbon assimilated by vegetation (grC/m^2/yr)
500 - 1000
20000 1000 - 1500
1500 - 200015000
10000
5000
0
2000 - 2600
Sum
up
pixe
l
Total carbon assimilated by vegetation (grC/m^2/yr)
75
map distribution and GPP graphic distribution from ALOS/AVNIR-2 and Aster
satellite data in North Denpasar district is shown in Figure 5.19.
Table 5.18. Total pixels and hectarage of annual GPP value with differences
satellite data in North Denpasar district
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 23871 238.71 16673 375.14
250 - 500 19298 192.98 14054 316.22
500 - 750 15851 158.51 11322 254.75
750 - 1000 13377 133.77 8907 200.41
1000 - 1250 12236 122.36 7075 159.19
1250 - 1500 11201 112.01 5954 133.97
1500 - 1750 10777 107.77 5037 113.33
1750 - 2000 8323 83.23 2892 65.07
2000 - 2250 3535 35.35 671 15.10
> 2250 109 1.09 13 0.29
Total 1185.78 1633.46
76
a) b)
d)35000
Fig. 5.19. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in North Denpasar District
The GPP pixel value distribution in East Denpasar district from
ALOS/AVNIR-2 satellite data is dominated by pixel with value < 250 gC m-2 yr-1
with the area of 331.62 ha, which is become decrease until the GPP pixel value is
high (> 2250 gC m-2 yr-1) with the area of 1.53 ha. In the other case, the GPP pixel
value distribution from Aster satellite data is dominated by low pixel value <250
gC m-2 yr-1 with the area of 438.93 ha, which is become decrease until the GPP
pixel value is high (> 2250 gC m-2 yr-1) with the area of 0.47 ha (Table 5.19). GPP
map distribution and GPP graphic distribution from ALOS/AVNIR-2 and Aster
satellite data in East Denpasar district is shown in Figure 5.20.
30000
25000
20000
15000
10000
5000
0
< 500
500 - 1000
1000 - 1500
1500 - 2000
2000 - 2600
Sum
up
pixe
l
Total carbon assimilated by vegetation (grC/m^2/yr)
45000 c)40000
35000< 500
30000500 - 1000
250001000 - 1500
200001500 - 2000
15000
10000
5000
0
2000 - 2600
Sum
up
pixe
l
Total carbon assimilated by vegetation (grC/m^2/yr)
77
Table 5.19. Total pixels and hectarage of annual GPP value with differences satellite data in East Denpasar district
ALOS/AVNIR-2 ASTER GPP value
(grC/m^2/yr) Total pixels Hectarage (ha) Total pixels Hectarage (ha)
< 250 33162 331.62 19508 438.93
250 - 500 27407 274.07 14842 333.95
500 - 750 23948 239.48 12359 278.08
750 - 1000 19639 196.39 10763 242.17
1000 - 1250 17976 179.76 9244 207.99
1250 - 1500 16343 163.43 7973 179.39
1500 - 1750 13043 130.43 6603 148.57
1750 - 2000 7118 71.18 3329 74.90
2000 - 2250 2947 29.47 1136 25.56
> 2250 153 1.53 21 0.47
Total 1617.36 1930.01
78
Fig. 5.20. (a) GPP map distribution From ALOS/AVNIR-2, (b) GPP map distribution From Aster, (c) GPP graphic distribution from ALOS/AVNIR-2, and
(d) GPP graphic distribution from Aster in East Denpasar District
35000
30000
25000
20000
15000
10000
5000
0
< 500
500 - 1000
1000 - 1500
1500 - 2000
2000 - 2600
Sum
up
pixe
l
Total carbon assimilated by vegetation (grC/m^2/yr)
70000
60000
50000
40000
30000
20000
10000
0
< 500
500 - 1000
1000 - 1500
1500 - 2000
2000 - 2600
Sum
up
pixe
l
Total carbon assimilated by vegetation (grC/m^2/yr)
a) b)
c) d)
79
CHAPTER VI
DISCUSSIONS
The value of GPP was closely related to the value of total solar radiation,
concentration of carbon dioxide and the light use efficiency. The highest value of
GPP is normally closely related to areas of highest vegetation cover. This is
mainly due to the highest reflectance in near infrared region.
According to Xiao et al. (2004), the seasonal dynamics of GPP prediction
from satellite data were similar to those of GPP by observation. In tropical
evergreen forest, Amazon-Brazil, prediction of GPP from MODIS satellite data is
consistent with GPP estimation from the eddy flux tower (Xiao et al., 2005;
Seleska et al., 2003), GPP value prediction from MODIS satellite data is
approximately 2977 gC m-2 year-1 (Xiao et al., 2005). The annual value of GPP
ranges from 1710 to 2635 gC m-2 year-1 estimated from MODIS satellite data is
got from the research in Labanan Concession Area, East Kalimantan-Indonesia
(Nugroho, 2006).
In Denpasar area, the GPP value from ALOS/AVNIR-2 and Aster satellite
data is smaller than the GPP value from MODIS product (MOD17). MODIS
image has a short spectral resolution which the interval spectral resolution of red
band is only 0.05 micrometers and 0.035 micrometers for the NIR band.
Meanwhile, ALOS/AVNIR-2 has longer interval spectral resolution for the red
band until 0.08 micrometers and 0.13 micrometers for the NIR band. In the other
case, Aster has an interval of spectral resolution for the red band until 0.06
79
80
micrometers dan 0.08 micrometers for the NIR band. Smaller interval of spectral
resolution will give higher capability of sensor in detecting object in surface.
According to Kruse (2000) the decreasing of spectral resolution will loose
the ability of map fine spectral to distinguish in detail. Particularly apparent at this
site because there are not any extremely fine spectral differences, coarser spectral
resolution can prohibit discrimination and identification of specific object. The
spectral resolution is a direct function of the material that are trying to identify,
and the contrast between that material and the background materials (De Jong and
Van Der Meer, 2005). As states by Lizarazo (2006) a higher spatial detail does not
mean higher spectral richness and some limitations arise to get accurate classes.
AVNIR-2 TOA reflectances in band 4 AVNIR-2 appear to be lower than the
TOA reflectances from exogenous sensor narrow bands centered at 860 nm. This
could be explained by the significant water vapour and dioxygen absorption
present in this band (Saunier et al., 2006). This different of TOA reflectant also
caused the different of annual GPP value from ALOS/AVNIR-2.
The different land use will effect the different of annual GPP value. Forest
(mangrove) land use has a high mean of annual value GPP with value 1123.58 gC
m-2 yr-1 from ALOS/AVNIR-2 satellite data and 1177.40 gC m-2 yr-1 from Aster
satellite data. The lowest mean annual value of GPP is found in settlement land
use which the value is 540.50 gC m-2 yr-1 from ALOS/AVNIR-2 satellite data and
492.44 gC m-2 yr-1 from Aster satellite.
The different effect land use that caused the different value of GPP is quite a
lot, because the different land use has a different vegetation type, percentage
81
vegetation cover and distribution. The value of vegetation index related to
percentage vegetation cover (Horning, 2004; Inoue et al., 2008). Forest
(mangrove) and ricefield are two type’s land use with a higher mean GPP value,
because the land use having high vegetation cover and homogeneous vegetation
cover. While, the settlement land use has a higher maximum GPP value, but has
lower mean of GPP value because pixel value from vegetation indices make
generalize to other classes of urban land use and activity based on the spectral
values of each pixel. Denpasar is a unique city in Bali, in some place in the center
of the town there are has a holy area. Holy area has a high percentage vegetation
cover, therefore settlement land use have a high maximum value of GPP.
According to Soegaard and Jensen (2003), factories emitting CO2 may be hidden
beneath the same type of roof as found in residential areas, and in the center of
town, the canopy shows great height fluctuations and may cover many different
activities. This creates a problem if aiming at a satellite-based subdivision of
urban areas into generalized classes of urban land use and activity based on the
spectral values of each individual pixel.
The different vegetation type has a different photosynthetic pathway type
(C3 versus C4) also suggest associated differences in GPP production efficiency
(Turner et al., 2002; Black, 2006). Ricefield and shrub assumed dominants by C4
vegetation type and the other land use dominants by C3 vegetation type. The
mean residence time of carbon in C4 species should be remarkably shorter than in
C3 species (Ito and Oikawa, 2004). The leaf-level photosynthesis rate per unit
APAR tends to be greater in C4 than C3 species (Turner et al., 2002; Black, 2006;
82
Bortolot, 2004) under conditions of high light, drought, or high temperature
(Turner et al., 2002). The measurements of Gillon and Yakir (2001) in Cuntz
(2002) indicate that carbonic activity in C4 is more reduced than in C3 plants.
Figure 5.6. shows the high different of total annual GPP with different land
use in ALOS/AVNIR-2 and Aster satellite data, a specially in ricefield and
settlement land use. Ricefield land use is used as an agriculture area with
characteristic crop rotation. The different temporal resolution by the day of
recording ALOS/AVNIR-2 and Aster is influencing for the different annual value
of GPP from those two satellite data.
Settlement land use has a high heterogeneous landscape that influencing the
object reflectance value on earth surface. This problem could be solved by use the
image satellite with higher resolution. With the higher image satellite resolution,
the object detection is get more specific which can make higher accurate of data.
This matter caused the annual mean GPP from ALOS/AVNIR-2 in settlement
land use higher than in Aster (Table 5.3 and Table 5.4), as states by Soegaard and
Møller-Jensen (2003), the urban landscape of the flux measurements get more
complicated due to the surface heterogeneity and the NDVI loses its importance
for scaling the CO2 exchange.
Higher total of annual GPP from Aster in settlement land use is caused by
the different of spatial resolution and the different of spectral resolution between
Aster and ALOS/AVNIR-2. Aster satellite detects vegetation that has a little of
cover around of settlement as pixel that has low value vegetation index and not as
building area (Fig 6.1), this problem caused the annual GPP by Aster satellite is
83
higher than the total annual GPP by ALOS/AVNIR-2 in settlement land use
(Table 5.6). As states by Kerr and Ostrovsky (2003) in Rocchini (2007), satellite
data with better spatial resolution seems to allow to narrow the scale gap between
field and remotely sensed data perceived with coarse resolution satellites.
Fig. 5.1. GPP distribution in ALOS/AVNIR-2 (a) and GPP distribution in Aster (b), compare with vegetation distribution from Quickbird satellite data (c) and
compare to with ground check picture
According to Krue (2000), Spatial resolution is the key to mapping of
detailed scale-dependent variation. Increasing the pixel size (decreasing the spatial
resolution) results in the loss of image detail. There is a tendency to lose small,
discrete occurrences of specific object with larger pixels. The spatial resolution of
the remote sensing data is crucial for discriminating fluxes for the different land
cover types and hence avoiding significant errors due to application of a land
surface model to a mixed pixel containing large contrasts in vegetation cover (Li
et al., 2008).
#S
304400
304400
304450
304450
9036
950 9036950
9037
000 9037000
#S
304400
304400
304450
304450
9036
950 9036950
9037
000 9037000
#S
304400 304450
304400 304450
9036
950 9036950
9037
000 9037000
c) a) b)
84
South Denpasar has a two types land use with a high total GPP influencing
to total GPP value in South Denpasar. Those two types of land use are ricefield
and forest (mangrove). West Denpasar has a low amount of GPP value because in
West Denpasar that covered by a small area of ricefield, especially without
contribution from forest (mangrove) (Table 6.1). According to Li et al (2008),
pixels from satellite imagery are invariably a mixture of vegetation types, land use
types and cover fractions.
Table 6.1. Land use area by district
Hectarage (ha) Land Use South
Denpasar West
Denpasar North
Denpasar East
Denpasar Settlement 2391.33 1729.11 1528.41 1530.66
Ricefield 837.60 363.96 598.16 816.65
Forest (Mangrove) 700.74 - - -
Shrub 56.12 15.69 0.66 8.63
Perennial plant 326.93 104.20 285.23 245.43
Dryland 124.64 38.23 20.10 80.30
Bareland 427.73 130.00 153.10 116.62
West Denpasar and northern area of Denpasar are covered by wider clouds
that effect the estimation of total annual of GPP by those two satellites, Aster and
ALOS/AVNIR-2. The cloud covered in each district is shown in Table 6.2.
85
Table 6.2. Cloud Hectarage by district
ALOS/AVNIR-2 Aster District Hectarage Cloud
hectarage (ha) Percentage
(%) Cloud
hectarage (ha) Percentage
(%) South Denpasar 4864.88 2.11 0.04 102.08 2.10
West Denpasar 2380.95 247.22 10.38 368.55 15.48
North Denpasar 2585.58 437.74 16.93 102.80 3.98
East Denpasar 2798.29 78.47 2.80 81.25 2.90
Total 12629.70 765.54 6.06 654.68 5.18
Basically, the remote sensing data application in GIS analyses is application
the raster data in its analyzing process. The application of raster data in remote
sensing will increasing the GIS ability in data analyze, because the remote sensing
has a such variant raster data which is use in a variant time and area in it’s
different spectral resolution. In the other case, the data raster has a problem, such
as eliminating the detail information by the generalization in one pixel.
Displaying the road and river network on monitor computer is very effective
and efficient tool in observing the relationship between the spatial and physical
attributes of human activity. The integration between road and river of vector data
and GPP value information in raster data could be increasing the accuration of
position and GPP value information’s which is visualized as a map (Appendix 1).
Modern advances in GIS-based cartography make it easier than ever to create
large numbers of maps quickly, using automated techniques. This map is expected
could be use in future planning, especially to control the impact of climate
change.
86
CHAPTER VII
CONCLUSIONS AND SUGGESTIONS
7.1. Conclusion
1. Combination between high resolution remote sensing data with GIS
application can be used to estimated GPP value in urban area like Denpasar,
where more communicative layout result could be given by GIS application.
GIS application also could be increasing the accuration of position GPP value
information’s.
2. Annual value of GPP from ALOS/AVNIR-2 is lower than that of the value
from Aster, because ALOS have a higher spatial resolution and smaller
interval of spectral resolution compared to Aster, where totally GPP per year
in Denpasar from ALOS/AVNIR-2 is 52421.46 tC yr-1 and from Aster is
59355.49 tC yr-1.
3. The high spatial resolution of the remote sensing data is crucial discriminating
different land cover types in urban land cover. With the surface heterogeneous
of land cover, maximum value of GPP from ALOS/AVNIR-2 is smaller than
that of the value GPP from Aster. Meanwhile, the annual mean of GPP value
by ALOS/AVNIR-2 is higher than the annual mean of GPP by Aster, because
ALOS/AVNIR-2 has higher spatial resolution and more significant detection
quality and condition of vegetation than Aster.
4. Different value of GPP is affected by the different of land use where higher
mean value of GPP could be found in forest (mangrove) and ricefield. The
86
87
value of GPP in forest (mangrove) and ricefield by ALOS/AVNIR-2 is
1123.576 gC m-2 yr-1 and 1030.080 gC m-2 yr-1, respectively and by Aster is
1177.396 gC m-2 yr-1 and 1020.648 gC m-2 yr-1, respectively. The lowest mean
value of GPP was found in settlement land use with value of 540.492 gC m-2
yr-1 by ALOS/AVNIR-2 and 492.443 gC m-2 yr-1 by Aster satellite data.
5. The maximum value of GPP by those two satellite data, ALOS/AVNIR-2 and
Aster, is smaller than the maximum GPP value by MODIS GPP product
(MOD17) in Denpasar area, and also that is smaller than the measurement
over a tropical peat swamp forest in central Kalimantan-Indonesia and smaller
than measurement over a tropical forest in central Amazonia, Brazil.
7.2. Suggestion
1. The differences of spatial and spectral resolutions influence accuration of
object detection. The object detection for heterogeneous area such as
settlement land use is recommended to use satellite with high spatial
resolution, meanwhile for homogeneous area such as forest (mangrove) and
ricefield is recommended to use satellite with high spectral resolution.
2. Some research to estimate GPP using ALOS/AVNIR-2 and Aster is needed in
area which has the eddy flux tower, in order to get the more accurate
validation result.
88
REFERENCES
Abrams, M. and S. Hook. 2002. ASTER User Handbook: Version 2. Jet
Propulsion Laboratory/California Institute of Technology
Angel, S., S.C. Sheppard, and D.L. Civco. 2005. The Dynamics of Global Urban Expansion. The World Bank, Transport and Urban Development Department. Washington D.C.-USA.
As-syakur. A.R. 2005. Aplikasi Sistem Informasi Geografi (SIG) Untuk Pemutakhiran Peta Agroklimat Pulau Lombok Berdasarkan Klasifikasi Oldeman dan Schmidt-Ferguson (Skripsi). Denpasar-Indonesia. Udayana University.
Black, S.C. 2006. Estimation of Grass Photosynthesis Rates in Mixed-Grass Prairie Using Field and Remote Sensing Approaches (Thesis). Saskatoon, Saskatchewan-Canada: University of Saskatchewan.
Bortolot, Z.J. 2004. An Adaptive Computer Vision Technique for Estimating the Biomass and Density of Loblolly Pine Plantations using Digital Orthophotography and LiDAR Imagery (Dissertation). the Virginia Polytechnic Institute. Virginia-United States.
Bradford, J.B., J.A. Hicke, and W.K. Lauenroth. 2005. The relative importance of light-use efficiency modifications from environmental conditions and cultivation for estimation of large-scale net primary productivity. Remote Sensing of Environment, 96, 246–255.
Burrough, P.A. 1986. Principles of Geographical Information Systems for Land Resources Assessment. Oxford University Press Inc. New York-USA
CCRS, 2007. Fundamentals of Remote Sensing. A Canada Centre for Remote Sensing. Canada.
Cuntz, M. 2002. A comprehensive global 3D model of δ18O in atmospheric CO2 (Dissertation). University of Heidelberg. Heidelberg-Germany
Daughtry, C.S.T., E.R. Hunt., and P.C. Doraiswamy. 2006. Assessing Carbon Dynamics in Agriculture Using Remote Sensing. USDA-ARS Hydrology and Remote Sensing Laboratory. Beltsville, Maryland-USA.
De Jong, S.M. and F.D. Van Der Meer. 2005. Remote Sensing Image Analysis: Including the Spatial Domain. Kluwer Academic Publishers. Dordrecht
88
89
DeBy, R.A., Y. Georgiadou, R.A. Knippers, M.-J. Kraak, Y. Sun, M.J.C. Weir, and C.J. Van Westen. 2004. Principles of Geographic Information Systems: An Introductory Textbook. The International Institute for Geo-Information Science and Earth Observation (ITC). Netherlands.
Department of Geography, University of Maryland. 2004. Landsat Technical Guide: Global Land Cover Facility. Maryland-USA
ECCM. 2002. ECCM technical note: Carbon sinks in the Amazon – the evidence. The Edinburgh Centre for Carbon Management, Produced for the World Land Trust. Brazil
Edwards, G.E., and D.A. Walker. 2001. Photosynthetic Carbon Assimilation. School of Biological Sciences, Washington State University, Pullman-USA.
Ehlers, M. 2002. Remote Sensing and Geographic Information Systems: Image-Integrated Geographic Information Systems. [Cited 2008, March 18]. Available from: URL: http://www.astm.org/DIGITAL_LIBRARY/STP/SOURCE_PAGES/STP1126.htm
EORC-JAXA. 2008. About ALOS Satellite. [Cited 2008, Mei 18]. Available from: URL: http://www.eorc.jaxa.jp/ALOS/cd_alos/eng/alos1.htm
Escobar, F., G. Hunter, I. Bishop, and A. Zerger. 2002. Introduction to GIS. Department of Geomatics, The University of Melbourne. Melbourne-Australia
ESRI. 1996. ArcView GIS; The Geographic Information System For Everyone. Environmental Systems Research Institute, Inc. New York-USA.
Gers, C.J. 2003. Relating Remotely Sensed Multi-Temporal Landsat 7 ETM+ Imagery to Sugarcane Characteristics. South African Sugar Association Experiment Station. Mount Edgecombe-South Africa
Gitelson, A.A., A. Viña, J.G. Masek, S.B. Verma, and A.E. Suyker. 2008. Synoptic Monitoring of Gross Primary Productivity of Maize Using Landsat Data. IEEE Geoscience and Remote Sensing Letters, Vol. 5, No. 2, 133 - 137
Glenn, E.P., A.R. Huete, P.L. Nagler, and S.G. Nelson. 2008. Relationship Between Remotely-sensed Vegetation Indices, Canopy Attributes and Plant Physiological Processes: What Vegetation Indices Can and Cannot Tell Us About the Landscape. Sensors, 8, 2136-2160
90
Gonga-Saholiariliva, N., J. Xavier, M. Denoyelle, Y. Gunnell, and C. Mering. 2008. Wetland and aquatic surface inventory using ALOS satellite images Examples from Sologne (central France) and Al'Tamin province (Iraq). Balwois. Ohrid-Republic of Macedonia.
Grose, A. 2004. Carbon Exchange Variability Over Amazon Basin Using Coupled Hydrometeorological-Mixed Layer PBL-CO2 Assimilation Modeling System Forced by Satellite-Derived Surface Radiation and Precipitation (Dissertation). Florida-USA: The Florida State University.
Hazarika, M.K. and Y. Yasuoka. 2002. Estimation of Terrestrial Carbon Fluxes by Integrating Remote Sensing with Ecosystem Modeling. Institute of Industrial Science , Univ. of Tokyo. Japan.
Heinsch, F.A., M. Reeves, P. Votava, S. Kang, C. Milesi, M. Zhao, J. Glassy, W.M. Jolly, R.Loehman, C.F. Bowker, J.S. Kimball, R.R. Nemani, and S.W. Running. 2003. User’s Guide GPP and NPP (MOD17A2/A3): Products NASA MODIS Land Algorithm. NASA. USA.
Heiskanen, J. 2007. Remote Sensing of Boreal Land Cover: Estimation of Forest Attributes and Extent (Dissertation). Unioninkatu, Helsinki-Finland: University of Helsinki.
Hirano, T., T. Harada, H. Segah, S. Limin, T. June, R. Hirata, and M. Osaki. 2005. CO2 Exchange of a Tropical Peat Swamp Forest in Central Kalimantan. Proceedings AsiaFlux Workshop 2005. International Workshop on Advanced Flux Network and Flux Evaluation. Fujiyoshida Japan.
Hooda, R.S. and D.G. Dye. 1996. Estimating Carbon-fixation in India based on Remote Sensing Data. Haryana State Remote Sensing Application Centre, HAU Campus. India
Horning, N. 2004. Global Land Vegetation; An Electronic Textbook. NASA Goddard Space Flight Center Earth Sciences Directorate Scientific and Educational Endeavors (SEE). Maryland-USA.
Huete, A., C. Justice, and W. van Leeuwen. 1999. Modis Vegetation Index (MOD 13) Algorithm Theoretical Basis Document. NASA. USA.
Hunt, E.R., J.T. Fahnestock, W.K. Smith, R.D. Kelly, J.M. Welker, and W.A. Reiners. 2002. Carbon Sequestration from Remotely Sensed NDVI and Net Ecosystem Exchange. R. S. Muttiah (ed.), Laboratory Spectroscopy to Remotely Sensed Spectra of Terrestrial Ecosystems, Kluwer Academic Publishers. Dordrecht-Netherlands.
91
Ibrahim, A.B. 2006. An Analysis of Spatial and Temporal Variation of Net Primary Productivity over Peninsular Malaysia Using Satellite Data. Department of Remote Sensing Faculty of Geoinformation Science and Engineering University Technologi Malaysia. Johor-Malaysia.
Inoue, Y., J. Peñuelas, A. Miyata, and M. Mano. 2008. Normalized Difference Spectral Indices for Estimating Photosynthetic Hyperspectral and CO2 Flux Measurements in Rice. Remote Sensing of Environmental, 112, 156 – 172
Ito. A., and T. Oikawa. 2004. Global Mapping of Terrestrial Primary Productivity and Light-Use Efficiency with a Process-Based Model. Global Environmental Change in the Ocean and on Land. 343-358.
JAXA. 2007. ALOS; User Handbook. Earth Observation Research Center. Japan Aerospace Exploration Agency. Japan.
Jenkins, J.P., A.D. Richardson, B.H. Braswell, S.V. Ollinger, D.Y. Hollinger, and M.L. Smith. 2007. Refining light-use efficiency calculations for a deciduous forest canopy using simultaneous tower-based carbon flux and radiometric measurements. Agricultural and Forest Meteorology, 143, 64–79.
Jensen, J.R. 2000. Remote Sensing of the Environmental Earth Resource Perspective. Prentice Hall. New Jersey-USA.
Jongschaap, R.E.E. 2006. Integrating crop growth simulation and remote sensing to improve resource use efficiency in farming systems (Dissertation). Wageningen University. Wageningen-Netherlands.
June. T. 2004. Net Primary Production of Tropical Forest: A Modeling Approach. Seameo Biotrop, Bogor-Indonesia.
Kerle, N., L.L.F. Janssen, and G.C. Huurnrman. 2004. Principles of Remote Sensing. The International Institute for Geo-Information Science and Earth Observation (ITC). Netherlands.
Kruse, F.A. 2000. The Effects of Spatial Resolution, Spectral Resolution, and SNR on Geologic Mapping Using Hyperspectral Data, Northern Grapevine Mountains, Nevada. in Proceedings of the 9th JPL Airborne Earth Science Workshop: Jet Propulsion Laboratory Publication 00-18, p. 261 - 269
La Puma, I.P., T.E. Philippi, and S.F. Oberbauer. 2007. Development of a global evapotranspiration algorithm based on MODIS and global meteorology data. Remote Sensing of Environment, 109, 225–236.
92
Li, F., W.P. Kustas, M.C. Anderson, J.H. Prueger, and R.L. Scott. 2008. Effect of remote sensing spatial resolution on interpreting tower-based flux observations. Remote Sensing of Environment, 112, 337–349.
Liang, S., T. Zheng, D. Wang, K. Wang, R. Liu, S. Tsay, S. Running, and J. Townshend. 2007. Mapping High-Resolution Incident Photosynthetically Active Radiation over Land from Polar-Orbiting and Geo stationary Satellite Data. Photogrammetric Engineering and Remote Sensing. 1085-1089.
Lillesand, F.M., and R.W. Kiefer. 1994. Remote Sensing and Image Interpretation. Thrid Editon. John Wiley and Sons. New York-USA
Lizarazo, I. 2006. Urban Land Cover and Land Use Classification Using High Spatial Resolution Images and Spatial Metrics. Proceedings of the 2nd Workshop of the EARSeL SIG on Land Use and Land Cover. Center for Remote Sensing of Land Surfaces, Bonn, 28-30 September 2006. 292-298.
Lotsch, A. 1996. Biome Level Classification of Land Cover at Continental Scales Using Decision Trees (Thesis). Boston-USA: Boston University.
LPSO, 1998. The Landsat-7 Science Data User's Handbook. Last update: January 10, 2008. Landsat Project Science Office. NASA's Goddard Space Flight Center. Maryland-USA.
Malhi, Y., A.D. Nobre, J. Grace, B. Kruijt, M.G.P. Pereira, A. Culf, S. Scott. 1998. Carbon Dioxide Transfer Over a Central Amazonian Rain Forest. Journal of Geophysical Research, 103, 593-612.
Moriyama, T. 2005. Principles of Remote Sensing. Diktat Lucture in Udayana Universty. Denpasar-Indonesia.
Muzein, B.S. 2006. Remote Sensing and GIS for Land Cover-Land Use Change Detection and Analysis in the Semi-Natural Ecosystems and Agriculture Landscapes of the Central Ethiopian Rift Valley (Dissertation). Dresden-Germany. University of Dresden.
Myneni, R.B., and D. L. Williams. 1994. On the Relationship between FAPAR and NDVI. Remote Sensing of Environment, 49, 200-211.
Nugroho, N.P. 2006. Estimating Carbon Sequestration in Tropical Rainforest Using Integrated Remote Sensing and Ecosystem Productivity Modeling: A Case Study in Lebanan Concession Area, East Kalimantan, Indonesia (Thesis). International Institute for Geo-Information Science and Earth Observation (ITC). Netherlands.
93
Ochi, S., and R. Shibasaki. 1999. Estimation of NPP based agricultural production For Asian countries using Remote Sensing data and GIS. Institute of Industrial Science , Univ. of Tokyo. Tokyo-Japan.
Osawa, Y. 2004. Optical and Microwave Sensors on Japanese Mapping Satellite – ALOS. ALOS Project Team, Japan Aerospace Exploration Agency (JAXA). Tsukuba-Japan.
PCI Geomatics. 2005. The OrthoEngine ALOS Module Add-On. [Cited 2008, Mei 22]. Available from: URL: http://www.pcigeomatics.com/products/alos.html
Prakash, A. 2001. Geographical Information Systems - An Overview. Indian Institute of Information Technology. India
Raju, P.L.N. 2005. Fundamentals of Geographical Information System. Geoinformatics Division, Indian Institute of Remote Sensing. Dehra Dun – India
Rautiainen, M. 2005. The Spectral Signature of Coniferous Forests: The Role of Stand Structure and Leaf Area Index (Dissertation). Unioninkatu, Helsinki-Finland: University of Helsinki.
Richards, J.A., and X. Jia. 2006. Remote Sensing Digital Image Analysis; An Introduction. 4th Edition. Springer-Verlag. Berlin-German
Rocchini, D. 2007. Effects of Spatial and Spectral Resolution In Estimating Ecosystem Α-Diversity By Satellite Imagery. Remote Sensing of Environment, 111, 423–434.
Running, S.W., R. Nemani, J.M. Glassy. and P.E. Thornton. 1999. Modis Daily Photosynthesis (PSN) and Annual Net Primary Production (NPP) Product (MOD17): Algorithm Theoretical Basis Document. NASA. USA.
Sabins, F.F. 1977. Remote Sensing Principles and Interpretation. W.H. Freeman and Campany. San Franscisco-USA.
Saleska, S.R., S.D. Miller, D.M. Matross, M.L. Goulden, S.C. Wofsy, H.R. da Rocha, P.B. de Camargo, P. Crill, B.C. Daube, H.C. de Freitas, L. Hutyra, M. Keller, V. Kirchhoff, M. Menton, J.W. Munger, E.H. Pyle, A.H. Rice, and H. Silva. 2003. Carbon in Amazon Forests: Unexpected Seasonal Fluxes and Disturbance-Induced Losses. Science, 302, 1554–1557.
Saunier, S., Y. Rodriguez, A. Mambimba, P. Goryl, P. Grimont, M. Bouvet, F. Viallefont, R. Santer, and G. Chander. 2006. Consolidated Verification Report: AVNIR-2. GAEL Consultant.
94
Short, N.M. 2006. Remote Sensing Tutorial. NASA. USA Government. [Cited 2007, December 27]. Available from: URL: http://rst.gsfc.nasa.gov/
Sims, D.A., H. Luo, S. Hastings, W.C. Oechel, A.F. Rahman, and J.A. Gamon. 2006. Parallel Adjustments in Vegetation Greenness and Ecosystem CO2 Exchange in Response to Drought in a Southern California Chaparral Ecosystem. Remote Sensing of Environmental, 103, 289 – 303.
Slamet S, L and A. Haryanto. 2006. Estimasi Emisi CO2 Dari Kebakaran Hutan (Sebuah Simulasi Dan Aplikasi Dengan Menggunakan Visual FoxPro). Proseding Semiloka Teknologi Simulasi dan Komputasi serta Aplikasi. BPPT. Jakarta-Indonesia.
Soegaard, H. and L. Møller-Jensen. 2003. Towards a spatial CO2 budget of a metropolitan region based on textural image classification and flux measurements. Remote Sensing of Environment, 87, 283–294.
SOVZOND. 2007. ALOS satellite vehicle has implemented the survey of the Moscow region territory with the spatial resolution of 2.5 m. [Cited 2008, Mei 22]. Available from: URL: http://www.sovzond.ru/en/about/news/2937.html
Star, J., and J. Estes. 1990. Geographic Information Systems; An Introduction. Prentice-Hall. Englewood Cliffs. New Jersey – USA.
Still, C.J., J.T. Randerson, and I.Y. FUng. 2004. Large-scale plant light-use efficiency inferred from the seasonal cycle of atmospheric CO2. Global Change Biology. 10, 1240–1252
Tang, J. 2007. The Analysis of Spatial-Temporal Dynamics of Urban Landscape Structure: A Comparison of Two Petroleum-Oriented Cities (Dissertation). Texas-USA: University-San Marcos.
Tu, K. 2000. Modeling Plant-Soil-Atmosphere Carbon Dioxide Exchange Using Optimality Principles (Dissertation). Santa Cruz. USA: University of California.
Turner, D.P., S.T. Gowerb, W.B. Cohenc, M. Gregorya, and T.K. Maiersperger. 2002. Effects of Spatial Variability In Light Use Efficiency on Satellite-Based NPP Monitoring. Remote Sensing of Environment, 80, 397– 405.
USGS. 2007. Geographic Information Systems. [Cited 2008, Jan. 28]. Available from: URL: http://erg.usgs.gov/isb/pubs/gis_poster/
Xiao, X., Q. Zhang, S. Saleska, L. Hutyra, P. de Camargo, S. Wofsy, S. Frolking, S. Boles, M. Keller, and B. Moore. 2005. Satellite-Based Modeling of Gross Primary Production in a Seasonally Moist Tropical Evergreen Forest. Remote Sensing of Environment, 94, 105–122
95
Xiao. X., D. Hollinger, J. Aber, M. Goltz, and Q. Zhang. 2004. Satellite-based Modeling of Gross Primary Production in an Evergreen Needle Leaf Forest. Remote Sensing of Environment, 89, 519-534.
Yüksel, A., A.E. Akay, and R. Gundogan. 2008. Using ASTER Imagery in Land Use/cover Classification of Eastern Mediterranean Landscapes According to CORINE Land Cover Project. Sensors, 8, 1237-1251.
Zhang, C. 2006. Monitoring Biological Heterogeneity in a Northern Mixed Prairie Using Hierarchical Remote Sensing Methods (Thesis). Saskatoon, Saskatchewan-Canada: University of Saskatchewan.
Zhao, T. 2007. Changing Primary Production and Biomass in Heterogeneous Landscapes: Estimation and Uncertainty Based On Multi-Scale Remote Sensing and GIS Data (Dissertation). Michigan: University of Michigan.
http://daac.ornl.gov/MODIS/modis.html.
96
APPENDIX 1 GPP MAP DERIVED FROM
SATELLITE DATA AND GIS TECHNIQUE APLICATION
97
98
99
100
101
102
103
104
105
106
107
About the Author
Abd. Rahman As-syakur, S.P., M.Si. was born in Dompu, December 4th, 1981. Begin early school at TK Pertiwi Dompu (1985-1987), and elementary school at SDN 1 Dompu until fifth year (1987-1992) and graduated at MI An Nur Dili, East Timor (1992-1993). Continuing high school at MTs An Nur Dili (1993-1996) and SMUN 1 Dili (1996-1999). Obtaining Agriculture Degree with concentration Soil Science from Udayana University (1999-2005). Master degree of Environmental Science Udayana University at 2009 for one year and seven months (1.7) year period study. Begin his career as a researcher in
Environmental Research Center Udayana University. Following workshop on Parallel Events COP 13/CMP-3 UNFCCC, Bali (December, 3-4th 2007), JAXA Workshop on Oceanography, Fishery and Remote Sensing (March, 26th 2008), International Symposium on Remote Sensing and Ocean Science in South East Asia (March, 27th 2008), Workshop on International University/Institute consortium for climate change and natural disaster (March, 28-29th 2008), Final Workshop on JAXA Pilot Project of Utilizing ALOS Data in Bali, Indonesia (March, 10th 2009), Training course on ALOS Interferometry and Polarimetry conducted by LAPAN and JAXA (July, 22nd-24th 2008), and Training course on ALOS PRISM DEM and PALSAR Interferometry conducted by LAPAN and JAXA (February, 17th-19th 2009). In his experiences study, had published two scientific journals there are in Environmental Journal and Journal of Science, and also one of scientific proceeding in Indonesian Remote Sensing Society (MAPIN).
top related