KOPRI Repository

Model Inter-Comparison of Gross Primary Productivity and Evapotranspiration in Alaska

Cited 0 time in wos
Cited 0 time in scopus
Title
Model Inter-Comparison of Gross Primary Productivity and Evapotranspiration in Alaska
Other Titles
알래스카 총생상량과 증발상량을 모델로 비교 분석
Authors
Lee, Jane
Lee, Yoo Kyung
Keywords
EvapotranspirationGross Primary Productivity
Issue Date
2016
Citation
Lee, Jane, Lee, Yoo Kyung. 2016. Model Inter-Comparison of Gross Primary Productivity and Evapotranspiration in Alaska. 2016 American Geophysical Union. Moscone Center, San Francisco. 2016.12.12~2016.12.16.
Abstract
Alaska can be divided into two large regions, the arctic tundra and boreal region. The study area is Alaska from 72°N-52°N, and 170°W-140°W, latitude and longitude, respectively.Machine-learning and process-based modeling approaches are two ways to derive regional and global scale gross primary productivity (GPP) and evapotranspiration (ET). Machine-learning approaches upscale in-situ observed fluxes to a large scale with statistical modes, with satellite-derived parameters and other explanatory variables. Process-based models use a series of nonlinear equations to represent land-atmosphere soil system and associated fluxes. In this study, we perform model inter-comparison of gross primary productivity and evapotranspiration products from one process-based model (the Breathing Earth System Simulator) and two independent machine-learning models(Support Vector Regression, and MPI-BGC). The Breathing Earth System Simulator (BESS) is a process-based model that uses MODIS land and atmosphere products to estimate gross primary productivity (GPP) at global scales with 1km resolution. GPP is computed by a carbon-water coupled module, which incorporates a two-leaf longwave radiative transfer model, Farquhar's photosynthesis model and a quadratic Penman-Monteith and energy balance equations. The derived instantaneous estimates of GPP and ET were temporally upscaled to 8-day mean estimated with a simple cosine function. BESS estimated products has been evaluated in agricultural, forest and savanna ecosystems, but not in arctic tundra. The Support Vector Regression (SVR) model, satellite remote sensing data, and disturbance information was combined with 21 eddy covariance towers in Alaska to upscale the estimated carbon and energy balance from 2000 to 2011. The MPI-BGC (Max Planck Institute for Biogeochemistry) is an empirical model using meteorological, satellite and FLUXNET data to estimate global atmosphere-land carbon exchange.
URI
https://repository.kopri.re.kr/handle/201206/7175
Conference Name
2016 American Geophysical Union
Conference Place
Moscone Center, San Francisco
Conference Date
2016.12.12~2016.12.16
Type
Poster
Indexed
포스터
Appears in Collections  
2016-2017, Developing analytical techniques for investigating changing permafrost ecosystems in the Arctic (16-17) / Kim, Mincheol (PN16082; PN17082)
Files in This Item

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse