HR: 16:35h
AN: B52E-03    [PDF]
TI: Integration of MODIS-LAI with Sim-CYCLE to Estimate the Net Primary Productivity
AU: * Hazarika, M
EM: manzul@iis.u-tokyo.ac.jp
AF: Institute of Industrial Scien ce, Room No. Ce-509, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo, 153-8505 Japan
AU: Yasuoka, Y
EM: yyasuoka@iis.u-tokyo.ac.jp
AF: Institute of Industrial Scien ce, Room No. Ce-509, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo, 153-8505 Japan
AU: Ito, A
EM: itoh@jamstec.go.jp
AF: Frontier Research System for Global Change, 3173-25 Showa-machi, Kanazawa-ku, Yokohama, 236-0001 Japan
AU: Dye, D
EM: dye@jamstec.go.jp
AF: Frontier Research System for Global Change, 3173-25 Showa-machi, Kanazawa-ku, Yokohama, 236-0001 Japan
AB: Net primary production (NPP) is a key component of the global carbon cycle. Among the pools and fluxes that make up the cycle, NPP accounts most of the annual carbon fluxes between the atmosphere and biosphere. To improve our understanding and estimating the NPP accurately at various spatial and temporal scales, there is a need for integration of multiple, complementary and independent methods and datasets. In this context, an ecosystem model plays an important role in synthesizing such desperate time/space data into single coherent analysis of terrestrial carbon fluxes. However, such models require a large number of biophysical and climatic parameters over a large area and their applicability to operational level is often limited by the fundamental drawback of large input parameter requirements. This is especially true for the parameters describing vegetation composition and structure, such as leaf area index (LAI), which is generally highly variable over space and time, and difficult to measure with conventional methods. However, recent advances in remote sensing, both in terms of new sensors (e.g., MODIS) as well as algorithm development for data processing have shown a promising future and now it is possible to do near real-time monitoring of the important biophysical parameters of vegetations like LAI to input into an ecosystem model. The current study aimed at the possibility of integrating remote sensing data to a processed based ecosystem model driven by conventional data. The ecosystem model selected for this study is Sim-CYCLE due to its portability for integrating data from different sources. Sim-CYCLE model is based on the dry-matter production theory and integrates ecophysiological findings into a simple scheme of plant growth to achieve the scaling-up from single-leaf to canopy level. The LAI derived from MODIS sensor (MODIS-LAI) was integrated with Sim-CYCLE and the model was renamed as MOD-Sim-CYCLE. Global annual NPP was estimated as 59.6 Gt C yr-1 by MOD-Sim-CYCLE, whereas it was 62.7 Gt C yr-1 in case of Sim-CYCLE. An intercomparison of LAI and NPP for world's major ecosystems was made for MOD-Sim-CYCLE and Sim-CYCLE. Differences both in magnitude and seasonality were observed in between MODIS-LAI and Sim-CYCLE simulated LAI for each biome. However, MOD-Sim-CYCLE LAI showed seasonality very accurately, synchronising with NPP and provided information on actual phenological conditions of the vegetations on the ground. A high correlation was also found between observed NPP and MOD-Sim-CYCLE NPP.
DE: 1640 Remote sensing
SC: Biogeosciences [B]
MN: 2003 Fall Meeting