Microsoft Word - 22 BIOTROPIA VOL. 13 NO. I, 2006 : 22 - 36 INTEGRATION OF NPP SEMI MECHANISTIC - MODELLING, REMOTE SENSING AND CIS IN ESTIMATING CO2 ABSORPTION OF FOREST VEGETATION IN LORE LINDU NATIONAL PARK TANIA JUNE0, ANDREAS IBROM2), AND GODE GRAVENHORsr9 "BIOTROP-ICSEA, SEAMEO BIOTROP, Bogor. lndonesia;email: taniajune@biotrop.org: and Laboratory ofAgrometeorology, Bogor Agricultural University, Bogor, Indonesia. ''STORMA (Stability of Rainforest Margin) Scientists hltp:/Avww. storma.de ABSTRACT Net Primary Production, NPP, is one of the most important variables characterizing the performance of an ecosystem. It is the difference between the total carbon uptake from the air through photosynthesis and the carbon loss due to respiration by living plants. However, field measurements of NPP are time-consuming and expensive. Current techniques are therefore not useful for obtaining NPP estimates over large areas. By combining the remote sensing and GIS technology and modelling, we can estimate NPP of a large ecosystem with a little ease. This paper discusses the use of a process based physiological sunshade canopy models in estimating NPP of Lore Lindu National Park (LLNP). The discussion includes on how to parameterize the models and how to scale up from leaf to the canopy. The version documented in this manuscript is called NetPro Model, which is a potential NPP model where water effect is not included yet. The model integrates CIS and the use of Remote Sensing, and written in Visual Basic 6.0 programming language and Map Objects 2.1. NetPro has the capability of estimating NPP of Cs vegetation under present environmental condition and under future scenarios (increasing [CO2], increasing temperature and increasing or decreasing leaf nitrogen level). Based on site-measured parameterisation of VaM* (Photosynthetic capacity), /Jj (Respiration) and leaf nitrogen ONi), the model was run under increasing CO2 level and temperature and varied leaf nitrogen. The output of the semi-mechanistic modelling is radiation use efficiency (?). Analysis of remote sensing data give Normalized Difference Vegetation Index (NDVI) and related Leaf Area Index (LAI) and traction of absorbed Photosynthetically Active Radiation (/M>AK). Climate data are obtained from 12 meteorological stations around die parks, which includes global radiations, minimum and maximum temperature. CO2 absorbed by vegetation (Gross Primary Production, GPP) is then calculated using the above variables and parameters with the following equation: estimating NPP, while ecosystem respiration is set as a function of temperature for estimating NEE. Under present condition, the net absorption of CO> by the vegetation of Lore Lindu National Park (NPP) is 1330.31 gCm"2year"' and at double CO2 and temperature increased of 3.5 "C, it increased by 23 %, reaching 1638.80 gCm'2 year'1. Key words : NPP Semi-mechanistic model, photosynthesis, carbon sequestration, net primary-production, tropical forest 22 BIOTROPIA VOL. 13 NO. 1,2006 INTRODUCTION The global carbon balance has become an issue of great concern during the last decade due to its impact on climate. IPCC (2001) has implicated that the increase in CO2 concentration from 285 ppm in year 1780 to 360 ppm in year 2000 has resulted in 0.6 °C global average temperature increase and it is projected that in the year 2100, the increase will be up to 5.8 °C. The concern over this problem has resulted in international agreements (e.g., Rio 1992; UNFCCC 1994; and Kyoto 1997) to reduce CO2 concentration in the atmosphere. Net primary production (NPP) is an important quantitative characteristic of an ecosystem (Churkina et al. 1999). It refers to the net production of organic carbon by plants in an ecosystem usually measured over a period of a year or more. It is the difference between carbon gain through photosynthesis and carbon loss through respirations. It constitutes the total annual growth increment (both above and below ground) plus the amounts grown and shed in senescence, reproduction or death of short-lived individuals in a stand plus the amounts consumed by herbivores. Seasonal changes in NPP will influence seasonal changes in net ecosystem exchange (NEE; NPP-soil respiration), and it is a principal cause of seasonal changes in atmospheric CO2 (Keeling et al. 1996). Tropical forests are important sink of CO2, representing 59 % of the global carbon pool in forests (Dixon et al. 1994). Although the area of tropical forests is only 22 % of the global forest area, they account for 32 - 43 % of the world's potential terrestrial NPP (Melillo et al. 1993; Field et al. 1998). Because conventional measurements (periodical destructive harvesting) of NPP of all ecosystems at all times are impractical, models are needed to estimate NPP. The model discussed in this manuscript is based on Farquhar and Caemmerer (1982), June (2002) and De Pury and Farquhar (1997). Most part of the model equations are mechanistic and some are semi mechanistic. When using such models in estimating NPP, prediction of environmental effects to NPP become more reliable. This manuscript contained theoretical background in developing the models, at leaf and canopy level; parameterisation and integrating remote sensing (RS) technology for data input and Geographical Information System (GIS) for display. Parameterisation of model was conducted using photosynthesis system ADCLC4AM. The model is run to answer the following question: (i). How much CO2 is absorbed by vegetation in a protected forest like Lore Lindu National Park ; (ii). How would changes in temperature, light, and CO2 concentration in the atmosphere affect the absorption? Model parameters and estimated NPP of Lore Lindu National Park are presented. Integration of the model with remote Sensing data and GIS are done using Map Object 2.1 and Visual Basic 6.0 in an application software NetPro. 23 Integration of NPP semi mechanistic-modelling - T. June et al. MODELLING NET PRIMARY PRODUCTION (NPP) Plant Photosynthesis and Net Primary Production Plants fix carbon dioxide from the atmosphere during the process of photosynthesis. All the carbon fixed during photosynthesis is called gross primary production (or gross primary productivity in terms of rate, for example, tonnes of carbon per hectare per year, usually called GPP). Some of the fixed carbon is used by the plants themselves for metabolic processes (largely respiration) and in this process, carbon dioxide is returned to the atmosphere. The carbon that is not used in respiration remains on the plant and adds to the biomass of the plant. This is net primary production (NPP). NPP represents the net new carbon stored as biomass in stems, leaves or roots of plants. It is the difference between the carbon assimilated during photosynthesis by plant leaves and carbon consumption through respiration by leaves, stems and roots. It is a quantitative measure of plant growth and carbon uptake. The knowledge of NPP distribution provides information on the productivity of croplands, forest and grasslands and thus helps improve management strategies for sustainable development of natural resources. At the national scale, NPP allows the estimation of the contribution of landmass to the global carbon budget which is important in global change studies. In agricultural system and forestry, NPP is usually defined as the increase in the standing biomass plus losses through litterfall and through consumption by herbivore. There are three main ways of estimating net primary productivity - first involves the estimation of biomass production and second through the measurement of gas exchange and modelling or third through CO2 flux measurement using surface tower over forest canopy or agroforestry system or using mast installed with instrument directly measuring CO2 fluxes over short type of vegetation like grassland or crop. NPP modeling and data requirements. An attractive approach for estimating NPP was firstly proposed by Mont'ith (1972; 1977), in which he determined dC/dt (carbon accumulation over time;wnen time = 1 year dC/dt = NPP) as a product of the efficiency of the canopy (e, mol CO2 mol" 1 PAR or in unit gC MJT1), fraction of photosynthetically active radiation, PAR (/APAK) absorbed by the canopy and the daily PAR reaching the top of the canopy (PAR) as: The original approach of Monteith considered the value of e as a constant and based on net CO2 absorption (determined through increased in biomass). On reality, 24 BIOTROPIA VOL. 13 NO. 1,2006  e changes through time due to the changing climatic and plant variables like Leaf Area  Index, nitrogen level, and water status. In order to make the model to be responsive to  the changing environmental condition, or to be used for a climate change prediction  effect, e has to be mechanistically or semi mechanistically modelled using the approach  introduced in June (2002).  The outline of the semi‐mechanistic model to produce the e value introduced in this  manuscript is shown in Figure 1.  01 i\rr scim mcciiaiiiMii-inuueiuug — i. June tfi ui. I light intensity incident on leaf surface (nmolm"2 s"1) J rate of actual electron transport (̂ imol m"2 s"1) •/ma* maximum electron transport rate (̂ imol m'2 s"1) Ke Michaelis-Menten constant for carboxylation by Rubisco (|.ibar) Ka Michaelis-Menten constant for oxygenation by Rubisco (mbar) O ambient partial pressure of oxygen (mbar) R universal gas constant, 8.3144 J mol4 K"1 Rt dark respiration of leaf which continues in the light (nmol m"2 s"1) T leaf temperature (°C) K™« maximum rate of Rubisco activity in the leaf (umol m"2 s"') &„ nitrogen extinction coefficient k light extinction coefficient Af0 leaf nitrogen concentration on top of canopy (mmolm" 2) ^c.kai Total PAR absorbed by the canopy and leaf (umol m"2 s"') NPP Net Primary Production GPP Gross Primary Production NEE Net Ecosystem exchange Rveg Respiration by vegetation (0.45 GPP) Reco Respiration of ecosystem (as a function of temperature) The model used is a C^ photosynthesis model. It is chosen due to the fact that €3 plants dominate 95 % of earth vegetation. It is shown in the model that the responses of C3 leaf photosynthesis to light, temperature and CO2 concentration can be described by the biochemical properties of just two steps in the process, the carboxylation reaction (shown by Kcmax) and the regeneration of the acceptor for carboxylation (shown by J,mu). This mechanistic model, has been widely validated as an accurate predictor of photosynthetic carbon uptake by leaves with variation in environmental conditions. The scaling up to canopy to estimate NPP is done using sun-shade model (De Pury and Farquhar 1997; June 2002). In the simulation, supply of CCK (fj) into the leaf is modelled as 0.7 of ambient CCK (ca). This is a condition where water is not a limiting factor and vapour pressure deficit is around 12.5 mbar. 7max is taken as 2.1 Fcmax. To run the model the following groups of data are needed: (1) Fixed parameters of leaf and canopy photosynthesis for €3 plants (Table 1); (2) Photosynthetic parameters based on measurements ( Table 2) ; (3) Hourly climate data of maximum and minimum air temperature and global radiation. These hourly data are generated from daily data. Ra was determined by extrapolation of a linear regression at the lower end of the PAR response curve (at PAR = 0-100 ^mol m"2 s"1) (Figure 2) and FcmM was estimated from the lower end of the q response curve at q around 100 jibar (Table 2). 26 B1OTROPIA VOL. 13 NO. 1,2006 To scale up the model result for the whole national park, input data (LAI and /APAR) derived from NDVI (Normalized Difference Vegetation Index) observed from satellite images (Landsat TM) are used as follows: where NIR and RED is the amount of reflected light of visible near infrared, and red wavelengths, respectively; /APAR is fraction of absorbed photosynthetically active radiation. Both equations (3) and (4) are developed for tropical forest. MODEL DEVELOPMENT WITH STUDY CASE LORE LINDU NATIONAL PARK NetPro Model is a prototype of potential net primary production model where water deficit effect is not included yet. The model integrates the use of Remote Sensing in obtaining Leaf Area Index (LAI) through a relationship with NDVI. The mathematical equations showing the linear regression between LAI with NDVI and LAI with /APAR were obtained from tropical forest area. The LAI is used as input to the photosynthesis model, while the fAPM is used as input for Eq. (1) to estimate 27 Integration of NPP semi mechanistic-modelling - T. June et al. Table 1. Fixed photosyiithetic parameters for C3 plants and canopy parameters Parameters Value Sources r* CO: compensation partial pressure in the absence of dark respiration (|*bar) (at 25 42.75 Bemacchi et al. (2001) E activation energy for carboxylation, oxygenation, respiration, rubisco activity and COi compensation point (J mol"') 79430,36380, 46390, 65330,37830 Bemacchi et al. (2001) K, Michaelis-Menten constant for carboxylation by Rubisco (Pa) 40.49 Bemacchi et al. (2001) Ka Michaelis-Menten constant for oxygenation by Rubisco (Pa) 27840 Bemacchi etal. (2001) R universal gas constant (J mol"' K"1) 8.3144 Goudriaan (1977) P« canopy reflection coefficient for diffuse PAR 0.036 Goudriaan (1977) K, diffuse and scattered diffuse PAR extinction coefficient 0.715 Goudriaan (1977) pH reflection coefficient of a canopy with horizontal leaves 0.041 Goudriaan (1977) *' beam and scattered beam PAR extinction coefficient (for random orientation of leaves) 0.69/sinp June (2002) N,, Base level of nitrogen not associated with Photosynthesis (rnmol N m"') 29 Antene/a/. (1995) X. ratio of rubisco capacity to leaf nitrogen content 1.63 June (2002) Pi the leaf reflection coefficient for PAR 0.10 De Pury and Farquhar (1997) T, the leaf transmissivity to PAR 0.05 De Pury and Farquhar (1997) 0, curvature factor of the light response curve 0.7 June (2002) Note: 4-i/ip is sun elevation. 28 BIOTROPIA VOL. 13 NO. 1,2006 Study site: Lore Lindu National Park a. Site Information The NetPro V. 1.0 is run for Lore Lindu National Park, Central Sulawesi The whole national park is located in Lat. 1° 107-1 °50S and Long. 119°50?- 120°20?E (Figure 5). Vegetation that occurs in TNLL includes species dominating the lower area (200-1000 m) such as Mussaendopsis beccariana, Ficus sp., Myristica sp., Pterospermum sp., Canangium odoratum, Arrenga pinata and species dominating the higher area (1000-2500 m, 90 % of TNLL) such as Castanopsis argentea and Lithocarpus sp. Other species includes Podocarpus sp., Elaeorpus sp., Adinandra sp., Litsea sp., Callohylhim sp., Eucalyptus deglupta and Palmae (Kartawinata 1985; Mogea 2002). b. Software used and data input for NetPro Sofware used for data preparation and to run NetPro includes Visual Basic 6.0, ERMapper 6.4, ArcView 3.3 and MapObject 2.1. Data input into NetPro v 1.0 includes (1) shapefile of polygon with different characteristics of NDVI, minimum 29   BIOTROP1A VOL. 13 NO. 1, 2006 Figure 5. Lore Lindu National Park in Central Sulawesi, with red dots showing 12 meteorological stations around the park and red square shows the location the meteorological tower where direct measurement of CO: fluxes are conducted. and maximum temperature and PAR classes of LLNP; (2) daily averaged climate data (Global radiation, maximum and minimum temperature) from year 2001-2005. NDVI is derived from Landsat TM 7 dated 21 August 2001. The climate data, global radiation and daily temperature are zoned into several polygons ( Figure 6). NDVI, Leaf Area Index (LAI), and /APAR resulting from image analysis for year 2001 are shown in Figures 7, 8 and 9. NDVI values are divided into 4 classes and based on the Equation of Ibrahim (2001), LAI ranges from 0 to 10.04. LAI values are used as input to Photosynthesis model to obtain e value. /APAR distribution (ranges from 0 to 65 %) is shown in Figure 9, where these values are used as input to the model to obtain NPP and NEE. Classes of average NDVI, temperature and global radiation are overlayed to form polygons with different characteristic of climate and NDVI. c. Net Primary Production (NPP) and Net Ecosystem Exchange (NEE) of LLNP The value of simulated e changes with changing environmental conditions such as changes in global radiation, temperature, atmospheric CO2 concentration, and nitrogen level of the leaf, and therefore result in varied NPP and NEE values (Table 3). 31     Integration of NPP semi mechanistic-modelling - T. June et al. CONCLUSIONS This framework of model in estimating NPP using remotely sensed NDVI combined with semi-mechanistic modeling is first introduced by the author at the Scientific Meeting on Climate and Weather Prediction, Center for Climate and Atmospheric Science Applications, National Institute of Aeronautics and Space (LAPAN) in Bandung, Indonesia on 31th July 2002 where the proceedings was published in early 2003. The idea was then used by several undergraduate and post graduate students for thesis researches, using a constant value of radiation use efficiency (e) under the author's supervision for study sites in Sumatera and Sulawesi. This research idea is further extended to include a mechanistic estimation of the radiation use efficiency, and in 2004 it was approved by The Integrated Research Award secretariat of The Indonesian Institute of Sciences to be funded for the period of 2004-2006. However with limited availability of budget during that period, the funding was then cancelled after being approved by the reviewer panel. Through SEAMED BIOTROP DIPA2004, the development of the framework was realized and the prototype of the application software NETPRO is produced. Model parameterization was conducted in Lore Lindu National Park early 2005 supported by STORM A SFB 552 (Stability of Rainforest Margin) Project. This work still needs improvement in display and modeling. It also needs inclusion of water effect, although application for Lore Lindu National Park the assumption that water is not a limiting factor can be used. To be applicable to different site and type of vegetation, site specific parameterization (Vcmax, e) and site variable measurement and estimation like leaf nitrogen, LAI, /APAR. climate data (diurnal global radiation, minimum and maximum temperature, diffuse radiation, vapour pressure deficit) are required. It is also important to develop relationship between NDVI and LAI and /APAR from different type of vegetation. Direct measurement of CO2 flux and e from eddy correlation and micrometeorology techniques for validation of model are needed. Application of the model framework to other national parks in Indonesia will be an ongoing effort. REFERENCES Anten, N.P.R., F. Schieving, and M.J.A. Werger. 1995. Patterns of light and nitrogen distribution in relation to whole canopy carbon gain in C3 and C4 mono- and dicotyledonous species. Oecologia, 101:504-513. Bernacchi, C.J., E.L. Singsaas, C.Pimentel, A. R. Portis and S.P. Long, 2001. Improved temperature response functions for models of Rubisco-limited photosynthesis. Plant Cell and Environ., 24:253-259. Churkina, G., S.W. Running, and A.L. Schloss, 1999. Comparing global models of terrestrial net primary productivity: the importance of water availability. Global Change Biol., 5 (suppl): 46-55. 34 BIOTROP1A VOL. 13 NO. 1, 2006 De Pury, D. G. G. and G. D. Farquhar 1997. Simple scaling of photosynthesis from leaves to canopies without the errors of big-leaf models. Plant Cell and Environ., 20:537-557. Dixon, R. K., S. Brown, R. A. Houghton, A. M. Solomon, M.C. Trexler, and J. Wisniewski, 1994. Carbon pools and flux of global forest ecosystems. Science, 263: 185-190. Farquhar, G. D. and , S. von Caemmerer 1982. Modelling of photosynthetic responses to environmental conditions. Physiological plant ecology. //. Encyclopedia,of plant physiology, New Series. O. L. lange, P.S. Nobel, C.B. Osmond and H. Ziegler. Berlin, Springer-Verlag. Farquhar, G. D. and S. C. Wong 1984. "An empirical model of stomatal conductance. Aust J. Plant Physi., 11:191-210 Farquhar, G. D., S. von Caemmerer and J. A. Berry 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta, 149, 78-90. Farquhar, G. D. and S. von Caemmerer 1982. Modelling of photosynthetic responses to environmental conditions. Physiological plant ecology. II. Encyclopedia of plant Physiology, New Series. O. L. Lange, P.S. Nobel, C. B. Osmond and H. Ziegler. Berlin, Springer-Verlag. Field, C. B., M. .1. Behrenfeld, J. T. Randerson, and P. Falkowski 1998. Primary production of the biosphere: integrating terrestrial and oceanic components. Science, 281: 237-240. Hunt Jr., E. R., J. T. Fahnestock, R. D. Kelly, J. M. Welker, W. A. Reiners, W. K. Smith 2002. Carbon sequestration from remotely-sensed NDVI and Net Ecosystem Exchange. In. R. S. Muthiah (ed.). From Laboratory Spectroscopy to Remotely Sensed Spectra of Terrestrial Ecosystems, p 161-174. Kluwer Academic Publishers, Dordrecht, Netherlands. Goudriaan, 3. 1977. Crop micrometeorology: a simulation study. PUDOC, Wageningen. Goudriaan, J and H.H. van Laar 1994. Modeling Potential Crop Growth Processes. Kluwer Academic Publishers. Dordrecht/Boston/London. Ibrahim, 2001. Environment and Development in Coastal Region and in Small Island. Assessing Mangrove Leaf Area Index and Canopy Closure. 1PCC. 2001. Climate Change 2001. The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press. June, T. 2002. Environmental Effects on photosynthesis of Cj plants: scaling up from electron transport to the canopy (Study case: Gtycine max L. Merr). Environmental Biology, Research School of Biological Sciences. Australian National University. Canberra. June, T. 2003 Modelling Net Primary Productivity (NPP) using C3 photosynthesis model: Theoretical Approach. Proceeding to the Scientific Meeting on Climate and Weather Prediction. Pusat Pemanfaatan Sains Atmosfir dan Iklim LAPAN) Bandung, Indonesia 31 July 2002 Keeling, C. D., J.F.S Chin,, and T.P. Whorf 1996. Increased activity of northern vegetation inferred from atmospheric CO2 measurement. Nature, 382,146-149. Lind, M. and R. Fensholt 1999. The spatio-temporal relationship between rainfall and vegetation development in Burkina Faso. Dan. J. Geogr., 1999, 2, 43-56. Lloyd, J., J.Grace, A.C. Miranda, P. Meir, S.C. Wong, H.S. Miranda, I. R Wright, J.H.C Gash, and J. Mclntyre 1995. A simple calibrated model of Amazon rainforest productivity based on leaf biochemical properties. Plant Cell and Environ., 18:1129-1145. Kartawinata, K. 1985. The Tropical Rainforest: Phytogeography, Ecology, and Altitudinal Zonation. In. Remote Sensing in Vegetation Studies. Report of Training course on Remote Sensing Techniques Applied to Vegetation Studies. Bogor, 4 November-13 December, 1985. 35 Integration of NPP semi mechanistic-modelling - T. June et al Melillo, J.M., A.D McGuire, D.W. Kicklighter, B. Moore III, C.J. Vorosmarty and A.L. Schloss 1993. Global climate change and terrestrial net primary production. Nature, 363:234-240. Monteith, J. L. 1972. Solar radiation and productivity in tropical ecosystems. J Appl. Ecol., 9:747-766. Monteith, J. L. 1977. Climate and the efficiency of crop production in Britain. Phill. Trans. R. Soc. LondB., 281:277- 294. Ochi, S. and R. Shibasaki 1999. Estimation of NPP based Agricultural Production for Asian Countries Using Remote Sensing Data and CIS. Institute of Industrial Science. University of Tokyo. Japan. Pinter P. J. 1992. Solar angle independence in the relationship between absorbed PAR and remotely sensed data for alfalfa. Remote Sensing of Environ., 46, 19-25. Prince, S. D. and S. N. Coward 1995. Global Primary Production: A Remote Sensing Approach. J. Biogeogr., 22,815-835. Ruimy, A., B. Saugier, and G. Dedier 1994. Methodology for the Estimation of Terrestrial NPP from Remotely Sensed Data. J. Geophy. Res., Vol. 99, (3): 5263-5283. Von, Caemmerer, S.J.R. Evans, G.S. Hudson and T.J. Andrews (1994). The kinetics of Rubisco inferred from measurements of photosynthesis in leaves of transgenic tobaco with reduced Rubisco content. Planta, 195: 33-47. 36 22.pdf 23.pdf 24.pdf 25.pdf 26.pdf 27.pdf 28.pdf 29.pdf 30.pdf 31.pdf 32.pdf 33.pdf 34.pdf 35.pdf 36.pdf