Modeling Gross Primary Production of Savanna Woodlands in Southern Africa Using MODIS Imagery and CO2 Flux Tower Data

C. Jin; X. Xiao; L. Merbold; A. Arneth; E.M. Veenendaal; W. Kutsch

Abstract

Accurate estimation of gross primary production (GPP) of savanna ecosystem is valuable for evaluating the role of Africa in the global carbon cycle. An eddy flux observation network has been established to continuously measure the net CO2 fluxes (NEE) across various savanna vegetation types in Africa (CarboAfrica). Several publications have reported the seasonal dynamics and interannual variation of GPP for the savanna vegetation through partitioning of the measured NEE data. The satellite-based Production Efficiency Models (PEM), which calculate GPP as the product of absorbed photosynthetically active radiation (PAR) and light use efficiency (LUE), have been developed to scale up in situ GPP estimation from the eddy flux towers to regional scale. In this study, the Vegetation Photosynthesis Model (VPM) and the Moderate Resolution Imaging Spectroradiometer (MODIS) data were evaluated for their capacity to model GPP for savanna woodlands at two eddy flux towers in Botswana and Zambia, respectively. These two sites have different woodland types and precipitation pattern (Mopane woodlands vs. Miombo woodlands, semi-arid vs. semi-humid). In the VPM model, GPP is simulated as the product of photosynthetically active radiation (PAR), air temperature, Enhanced Vegetation Index (EVI), and Land Surface Water Index (LSWI). The results show that the simulated GPP by the VPM track well the temporal dynamic of GPP estimated from the eddy covariance measurements at these two sites. In addition, the land surface phenology of savanna woodlands, described by the satellite vegetation indices, especially the water-sensitive satellite indices-LSWI, are proved to match the phenology based on vegetation physiology activity measured by eddy covariance towers. The information of the timing and duration of vegetation growing season is useful for assisting the VPM modeling. Further evaluation of VPM simulations for and other savanna ecosystems is necessary before the VPM model is applied to estimate GPP of savanna ecosystems in Africa.

Original language	English
Publication status	Published - 2012
Event	AGU 2012, fall meeting, San Francisco, California, USA - Duration: 3 Dec 2012 → 7 Dec 2012

Conference/symposium

Conference/symposium	AGU 2012, fall meeting, San Francisco, California, USA
Period	3/12/12 → 7/12/12

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Cite this

@conference{4bde71f8040d4b9bbee671e5214a7bc9,

title = "Modeling Gross Primary Production of Savanna Woodlands in Southern Africa Using MODIS Imagery and CO2 Flux Tower Data",

abstract = "Accurate estimation of gross primary production (GPP) of savanna ecosystem is valuable for evaluating the role of Africa in the global carbon cycle. An eddy flux observation network has been established to continuously measure the net CO2 fluxes (NEE) across various savanna vegetation types in Africa (CarboAfrica). Several publications have reported the seasonal dynamics and interannual variation of GPP for the savanna vegetation through partitioning of the measured NEE data. The satellite-based Production Efficiency Models (PEM), which calculate GPP as the product of absorbed photosynthetically active radiation (PAR) and light use efficiency (LUE), have been developed to scale up in situ GPP estimation from the eddy flux towers to regional scale. In this study, the Vegetation Photosynthesis Model (VPM) and the Moderate Resolution Imaging Spectroradiometer (MODIS) data were evaluated for their capacity to model GPP for savanna woodlands at two eddy flux towers in Botswana and Zambia, respectively. These two sites have different woodland types and precipitation pattern (Mopane woodlands vs. Miombo woodlands, semi-arid vs. semi-humid). In the VPM model, GPP is simulated as the product of photosynthetically active radiation (PAR), air temperature, Enhanced Vegetation Index (EVI), and Land Surface Water Index (LSWI). The results show that the simulated GPP by the VPM track well the temporal dynamic of GPP estimated from the eddy covariance measurements at these two sites. In addition, the land surface phenology of savanna woodlands, described by the satellite vegetation indices, especially the water-sensitive satellite indices-LSWI, are proved to match the phenology based on vegetation physiology activity measured by eddy covariance towers. The information of the timing and duration of vegetation growing season is useful for assisting the VPM modeling. Further evaluation of VPM simulations for and other savanna ecosystems is necessary before the VPM model is applied to estimate GPP of savanna ecosystems in Africa.",

author = "C. Jin and X. Xiao and L. Merbold and A. Arneth and E.M. Veenendaal and W. Kutsch",

year = "2012",

language = "English",

note = "AGU 2012, fall meeting, San Francisco, California, USA ; Conference date: 03-12-2012 Through 07-12-2012",

}

TY - CONF

T1 - Modeling Gross Primary Production of Savanna Woodlands in Southern Africa Using MODIS Imagery and CO2 Flux Tower Data

AU - Jin, C.

AU - Xiao, X.

AU - Merbold, L.

AU - Arneth, A.

AU - Veenendaal, E.M.

AU - Kutsch, W.

PY - 2012

Y1 - 2012

N2 - Accurate estimation of gross primary production (GPP) of savanna ecosystem is valuable for evaluating the role of Africa in the global carbon cycle. An eddy flux observation network has been established to continuously measure the net CO2 fluxes (NEE) across various savanna vegetation types in Africa (CarboAfrica). Several publications have reported the seasonal dynamics and interannual variation of GPP for the savanna vegetation through partitioning of the measured NEE data. The satellite-based Production Efficiency Models (PEM), which calculate GPP as the product of absorbed photosynthetically active radiation (PAR) and light use efficiency (LUE), have been developed to scale up in situ GPP estimation from the eddy flux towers to regional scale. In this study, the Vegetation Photosynthesis Model (VPM) and the Moderate Resolution Imaging Spectroradiometer (MODIS) data were evaluated for their capacity to model GPP for savanna woodlands at two eddy flux towers in Botswana and Zambia, respectively. These two sites have different woodland types and precipitation pattern (Mopane woodlands vs. Miombo woodlands, semi-arid vs. semi-humid). In the VPM model, GPP is simulated as the product of photosynthetically active radiation (PAR), air temperature, Enhanced Vegetation Index (EVI), and Land Surface Water Index (LSWI). The results show that the simulated GPP by the VPM track well the temporal dynamic of GPP estimated from the eddy covariance measurements at these two sites. In addition, the land surface phenology of savanna woodlands, described by the satellite vegetation indices, especially the water-sensitive satellite indices-LSWI, are proved to match the phenology based on vegetation physiology activity measured by eddy covariance towers. The information of the timing and duration of vegetation growing season is useful for assisting the VPM modeling. Further evaluation of VPM simulations for and other savanna ecosystems is necessary before the VPM model is applied to estimate GPP of savanna ecosystems in Africa.

AB - Accurate estimation of gross primary production (GPP) of savanna ecosystem is valuable for evaluating the role of Africa in the global carbon cycle. An eddy flux observation network has been established to continuously measure the net CO2 fluxes (NEE) across various savanna vegetation types in Africa (CarboAfrica). Several publications have reported the seasonal dynamics and interannual variation of GPP for the savanna vegetation through partitioning of the measured NEE data. The satellite-based Production Efficiency Models (PEM), which calculate GPP as the product of absorbed photosynthetically active radiation (PAR) and light use efficiency (LUE), have been developed to scale up in situ GPP estimation from the eddy flux towers to regional scale. In this study, the Vegetation Photosynthesis Model (VPM) and the Moderate Resolution Imaging Spectroradiometer (MODIS) data were evaluated for their capacity to model GPP for savanna woodlands at two eddy flux towers in Botswana and Zambia, respectively. These two sites have different woodland types and precipitation pattern (Mopane woodlands vs. Miombo woodlands, semi-arid vs. semi-humid). In the VPM model, GPP is simulated as the product of photosynthetically active radiation (PAR), air temperature, Enhanced Vegetation Index (EVI), and Land Surface Water Index (LSWI). The results show that the simulated GPP by the VPM track well the temporal dynamic of GPP estimated from the eddy covariance measurements at these two sites. In addition, the land surface phenology of savanna woodlands, described by the satellite vegetation indices, especially the water-sensitive satellite indices-LSWI, are proved to match the phenology based on vegetation physiology activity measured by eddy covariance towers. The information of the timing and duration of vegetation growing season is useful for assisting the VPM modeling. Further evaluation of VPM simulations for and other savanna ecosystems is necessary before the VPM model is applied to estimate GPP of savanna ecosystems in Africa.

M3 - Abstract

T2 - AGU 2012, fall meeting, San Francisco, California, USA

Y2 - 3 December 2012 through 7 December 2012

ER -

Modeling Gross Primary Production of Savanna Woodlands in Southern Africa Using MODIS Imagery and CO2 Flux Tower Data

Abstract

Conference/symposium

UN SDGs

Fingerprint

Cite this