Abstract
Vegetation indices (VIs) have been shown to be a proxy of green leaf
area index (gLAI); however, it has not been verified whether the
relationships VI vs. gLAI are the same, as well as VIs retaining their
accuracy, for various crop types for estimating gLAI. The goal of this
study was to (1) determine if the best VIs used in previous studies for
gLAI estimation in maize and soybean may be applicable for potato and
wheat and vice versa, and (2) determining the cause of a hysteresis
between green up and reproductive stages for the VI vs. gLAI
relationship. Spectral measurements of wheat and potato were obtained in
Israel and of maize and soybean in the USA. In Israel, remote estimates
of gLAI were compared with in-situ canopy transmittance measurements of
irrigated potato and wheat under various nitrogen treatments from
2004-2007 for a total of 15 field-years. In eastern Nebraska, USA,
remote estimates of maize and soybean gLAI data were compared with
destructive gLAI determination in two irrigated/rainfed maize/soybean
rotation sites and in one irrigated site under continuous maize. These
data were collected during eight years (2001-2008) for a total of 24
field-years. For all four crops, the ten VIs examined showed
similarities in relationships between VIs and gLAI with the exception of
Red-edge Inflection Point (REIP) and the MERIS Terrestrial Chlorophyll
Index (MTCI). REIP and MTCI have very different relationships with maize
and soybean gLAI in green up and reproductive stages, thus, they require
re-parameterization during the season. This study outlines the two major
factors that influence the VI vs. gLAI relationship in the green up and
reproductive stages. While the results suggest that relationships VI vs.
gLAI are quite close for all four crops, different methodologies in
determining the ground-truth measurements of gLAI prevent us to confirm
whether algorithms calibrated for one crop can be used with no
re-parameterization for other crops. These concerns aside, we found that
normalized difference VIs (NDVI, Green NDVI, etc.) were capable of
estimating accurately gLAI below 2 and ratio VIs (Simple Ratio, CIgreen,
etc.) were best for gLAI above 2. For all four crops, green and red-edge
chlorophyll indices appear to be the most accurate for gLAI estimation.
| Original language | English GB |
|---|---|
| State | Published - 3 Dec 2012 |
| Event | American Geophysical Union, (AGU) Fall Meeting 2012 - San Francisco, San Francisco, United States Duration: 3 Dec 2012 → 7 Dec 2012 |
Conference
| Conference | American Geophysical Union, (AGU) Fall Meeting 2012 |
|---|---|
| Country/Territory | United States |
| City | San Francisco |
| Period | 3/12/12 → 7/12/12 |