The scaledEVI and scaledLST are calculated as below:
scaledEVI ¼ EVI À 0:1
ð5:32Þ
scaledLST ¼ min LST=30
ð
Þ ; 2:5 À 0:05 Â LST
ð
Þ
ð
Þ
½
ð 5:33Þ
where LST an is the annual mean nighttime LST. Like the GR model, without
predefining the e 0 , the TG model directly calculates e g using both scaled EVI and
LST. LST measures the surface temperature instead of air temperature. LST is an
effective measurement of the physiological activity of top canopy leaves. Moreover, LST has been proven to have a strong relationship with VPD and can be used
as the measurement of drought stress (Sims et al. 2008).
5.3.9 The Eddy Covariance Light-Use Efficiency Model (ECLUE)
The EC-LUE model is driven by four variables: NDVI, PAR, air temperature, and the
Bowen ratio of sensible to latent heat flux (Yuan et al. 2007). The parameters in the ECLUE model are independent across various land cover types. However, the accurate
estimation of the Bowen ratio with remote sensing datasets restricts the application
of the EC-LUE model. FPAR canopy is expressed as the linear function of NDVI:
FPAR canopy ¼ 1:24 Â NDVI À 0:186
ð5:34Þ
e g is estimated as the product of e 0 (2.14 g C m
-2 MJ
-1 APAR) and the
downregulation factor of air temperature (T) or water stress (W):
e g ¼ e 0 Â min T; EF
ð
Þ
ð5:35Þ
EF (evaporative fraction) is defined as a proxy of water condition:
EF ¼
LE
LE þ H
ð5:36Þ
where LE and H are latent heat flux and sensible heat flux (W m
-2 ) from eddy
covariance measurements, respectively. Recently, EF has been estimated as the
ratio of evapotranspiration (ET) to net radiation, and can be derived from the
Remote Sensing-Penman–Monteith (RS-PM) model.
5.4 Validation Efforts Using In Situ Measurements
In-situ observations from eddy covariance flux towers are effective data sources
for the validation of GPP estimations by PEMs. Around six hundred eddy
covariance flux tower sites continuously measure CO 2 , water, and energy fluxes in
various types of terrestrial ecosystems across the world (Xiao et al. 2011).
136
X. Xiao et al.
scaledEVI ¼ EVI À 0:1
ð5:32Þ
scaledLST ¼ min LST=30
ð
Þ ; 2:5 À 0:05 Â LST
ð
Þ
ð
Þ
½
ð 5:33Þ
where LST an is the annual mean nighttime LST. Like the GR model, without
predefining the e 0 , the TG model directly calculates e g using both scaled EVI and
LST. LST measures the surface temperature instead of air temperature. LST is an
effective measurement of the physiological activity of top canopy leaves. Moreover, LST has been proven to have a strong relationship with VPD and can be used
as the measurement of drought stress (Sims et al. 2008).
5.3.9 The Eddy Covariance Light-Use Efficiency Model (ECLUE)
The EC-LUE model is driven by four variables: NDVI, PAR, air temperature, and the
Bowen ratio of sensible to latent heat flux (Yuan et al. 2007). The parameters in the ECLUE model are independent across various land cover types. However, the accurate
estimation of the Bowen ratio with remote sensing datasets restricts the application
of the EC-LUE model. FPAR canopy is expressed as the linear function of NDVI:
FPAR canopy ¼ 1:24 Â NDVI À 0:186
ð5:34Þ
e g is estimated as the product of e 0 (2.14 g C m
-2 MJ
-1 APAR) and the
downregulation factor of air temperature (T) or water stress (W):
e g ¼ e 0 Â min T; EF
ð
Þ
ð5:35Þ
EF (evaporative fraction) is defined as a proxy of water condition:
EF ¼
LE
LE þ H
ð5:36Þ
where LE and H are latent heat flux and sensible heat flux (W m
-2 ) from eddy
covariance measurements, respectively. Recently, EF has been estimated as the
ratio of evapotranspiration (ET) to net radiation, and can be derived from the
Remote Sensing-Penman–Monteith (RS-PM) model.
5.4 Validation Efforts Using In Situ Measurements
In-situ observations from eddy covariance flux towers are effective data sources
for the validation of GPP estimations by PEMs. Around six hundred eddy
covariance flux tower sites continuously measure CO 2 , water, and energy fluxes in
various types of terrestrial ecosystems across the world (Xiao et al. 2011).
136
X. Xiao et al.