Satellite Applications for Detecting Vegetation Phenology - Satellite-based Applications on Climate Change - page 256

influences by utilizing the more atmosphere-sensitive blue band to correct the red
band for aerosol influences (Kaufman and Tanre 1992 ). The EVI has been found to
perform well in heavy aerosol, biomass burning conditions (Miura et al. 1998 ).

NIR Red

NDVI ¼ (17.1)

þ

NIR Red

NIR Red

EVI ¼ G (17.2)

NIR þ C 1 Red C 2 Blue þ L

where L (

¼

1) is the coefficient for canopy background adjustment, C1 (

¼

6) and

C2 ( ¼ 7.5) are aerosol resistance coefficients, and G ( ¼ 2.5) is a gain factor (Huete

et al. 2002 ).
LAI is broadly defined as the amount of leaf area (m 2 ) or the number of
equivalent layers in a canopy per unit ground area (m 2 ) (Knyazikhin et al. 1999 ).
The LAI is a state parameter needed by large-scale ecosystem models describing
the exchange flux of water vapor and CO 2 across the global biosphere-atmosphere
interface (McWilliam et al. 1993 ). The LAI product derived from the Moderate
Resolution Imaging Spectroradiometer (MODIS) reflectance monitors seasonal
variation in LAI at 1-km nadir resolution every 8 days. MODIS LAI provides a
more physically meaningful threshold for defining phenology events (such as onset
of greenness) than other vegetation indices (Kang et al. 2003 ).
Many validation efforts have beenmade to demonstrate the capability and accuracy
of deriving vegetation phenology from satellite measurements. Schwartz et al. ( 2002 )
compared three satellite-derived start-of-season measures and matched the results to
field data collected at the Harvard forest in Massachusetts. They concluded that each
method does a modestly accurate job of tracking the general pattern of surface
phenology. Ahl et al. ( 2006 ) compared field measurements of springtime forest
canopy phenology onset and maturity with estimates calculated from the MODIS-
derived vegetation products in northern Wisconsin forest. Their results showed that
MODIS products captured the general phenological development of the canopy
although they overestimated the leaf area during the over story leaf out period and
predicted onset of greenness and maturity earlier than that from field observations.

17.2.2 Approaches for Deriving Phenology from
Satellite Measurements

Many satellite phenology detection approaches have been addressed since early
90s (Table 17.1 ). The threshold for NDVI has been applied for phenological
classification of terrestrial vegetation (Lloyd 1990 ), modeling seasonal variation
of vegetation (Fischer 1994 ), and detecting characteristic of vegetation phenology
(Markon et al. 1995 ). They assume that a single threshold is applicable across land
covers. However, variation in background reflectance of different vegetation types
makes this a tenuous assumption. Therefore, it is not possible to establish a single,
meaningful threshold that signifies the onset (or end) of vegetative activity for the