Remote Sensing
- Field-Based Remote Sensing
- Elizabeth Walter-Shea, Mark Mesarch and Denise Gutzmer
- Project Goal
- Because remote sensing data are related to such plant characteristics as the fraction of absorbed light and the
amount of leaf material, we believe remotely sensed crop measurements can provide field, local and regional estimates
of cropland carbon assimilation. Spectral vegetation indices, composed of transforms of red or green and near-infrared
(NIR) reflectances (ρ), have been widely used as indicators of vegetation changes.
The objectives of our research include:
- To improve our understanding of the relation between crop reflectance (ρ), the absorbed fraction
of photosynthetically active radiation utilized by the crop (FPARgreen) and crop CO2 flux.
- To test various remotely sensed methods of estimating FPARgreen (i.e., fraction of
photosynthetically active radiation (PAR) intercepted by the photosynthesizing canopy elements).
- Project Description
- The core questions to be investigated are as follows:
- To what degree are the relationships between the fraction of absorbed photosynthetically active
radiation (FPARgreen) utilized by the crop and vegetation indices linear? How do these
relationships vary with sun angle?
- As dead leaf material accumulates during a growing season, are the (FPARgreen) - vegetation
indices relationships preserved for any geometry? If not, how do the relationships change? Is there any way to
estimate the fraction of dead leaf material with a vegetation index?
- Can we estimate canopy CO2 flux using vegetation indices?
- Progress
- Incoming solar energy and that reflected from the crop and soil were measured continuously during the 2003 growing
season at the three CSP research field sites using instruments located in the center of each field; these instruments,
known as radiometers, measure incoming and reflected sunlight in four different wavebands (Fig. 1).
| |
Fig. 1 A pair of SKYE 4-Channel radiometers (measuring
incoming (a) and reflected (c) light) are mounted in each field along with a digital camera (b) which provides day
to day documentation of changes in the crop. Student, Aaron Jenson, levels the radiometers which are
mounted 17 feet above the soil surface.
| |
Accompanying these measures of incoming and reflected light from the crop were measurements of reflected and
transmitted light from individual leaves in the crop canopy (Fig. 2).
| |
Fig. 2 Student Duane Ward (top) measures reflected
and transmitted light from individual corn leaves while student Marcus Shorney measures within a soybean crop.
Reflected and transmitted light of intact individual leaves (b) are measured using an SE590
spectroradiometer (a) along with a Li-Cor integrating sphere (c) and internal light source (d). Reflectance
(the ratio of reflected light to incoming light) and transmittance (the ratio of transmitted light to incoming
light) are calculated from these measurements. |
|
Reflectance (ρ; ratio of reflected to incoming readiation) was calculated and converted into remote sensing
vegetation indices:
(1) Normalized Difference Vegetation Index (NDVI) =
[( ρNIR - ρred) / ( ρNIR + ρred)] and
(2) [( ρNIR / ρgreen) - 1]
where ρNIR is the near-infrared (NIR) light reflectance (862-874 nm), ρred
is red light reflectance (665-675 nm) and ρgreen is green light reflectance (536-562 nm).
Comparisons of mid-day values of vegetation indices and total FPAR (total fraction of absorbed photosynthetically
active radiation) throughout the growing season show that NDVI is near-linearly related to total FPAR across all three
sites as the canopy develops (Fig. 3, upper curve).
| |
Fig. 3 Mid-day values of the remote sensing normalized difference
vegetation index [NDVI = [( ρNIR - ρred) / ( ρNIR +
ρred)] as a function of the total fraction of photosynthetically active radiation absorbed by the
corn crop (total FPAR) in 2003. ρNIR is reflectance in the near-infrared spectrum (862-874 nm)
and ρred is reflectance in the red portion of the spectrum (665-675 nm).
| |
Once the leaves started turning from green to yellow and brown in color (Fig. 4), a different relationship
resulted (Fig. 3, lower curve).
| |
Fig. 4 Reflectance and transmittance of individual leaves from a
corn field in 2002. Vertical bars represent green (536-562 nm), red (665-675 nm) and NIR (862-874 nm)
portions of the incoming sunlight which correspond to the incoming and reflected sunlight measurements over the
crops as mounted on the masts in the center of each field.
| |
While this relationship was also near-linear, there is little change in total FPAR with large changes in NDVI.
Two distinct relationships describe the comparison between the ratio index [(ρNIR / ρgreen) - 1] and total FPAR as well
(Fig. 5).
| |
Fig. 5 Mid-day values of the remote sensing vegetation index
[(ρNIR / ρgreen) - 1] which is related to green leaf area index (LAI) and the total
fraction of photosynthetically active radiation absorbed by the corn crop (total FPAR) in 2003. ρNIR
is reflectance in the near-infrared spectrum (862-874 nm) and ρgreen is reflectance in the
green portion of the spectrum (536-562 nm).
| |
As with the NDVI relation, NIR-green ratio changed drastically over a small change in total FPAR during the
green-up stage of crop development. Unlike the relation with NDVI, the NIR-green ratio index relationship with
total FPAR was non-linear. By accounting for the brown leaf components in the crop (FPARgreen), the
relationships between vegetation indices and FPARgreen during crop green-up were similar to those
during the leaf yellowing stage (Figs. 6 and 7).
| |
Fig. 6 Mid-day values of the remote sensing normalized difference
vegetation index [NDVI = (ρNIR - ρred) / (ρNIR+ ρred)]
as a function of the green fraction of photosynthetically active radiation absorbed by the corn crop
(FPARgreen) in 2003. ρNIR is reflectance in the near-infrared spectrum (862-874 nm)
and ρred is reflectance in the red portion of the spectrum (665-675 nm).
| |
| |
Fig. 7 Mid-day values of the remote sensing vegetation index
[(ρNIR/ ρgreen)-1] as a function of the green fraction of photosynthetically active
radiation absorbed by the corn crop (FPARgreen) in 2003. ρNIR is reflectance in the
near-infrared spectrum (862-874 nm) and ρgreen is reflectance in the green portion of the spectrum
(536-562 nm).
| |
- Staff
- Members of the Field-Based Remote Sensing research team are associated with the Climate and Bio-Atmospheric
Sciences (CBAS) group within the School of Natural Resources. See the CBAS web page at
http://snrs.unl.edu/cbas/ or
http://snr.unl.edu/okarmcart/.
- Elizabeth Walter-Shea
(Professor)
- Mark Mesarch
(Research / Outreach Specialist)
- Denise Gutzmer (Graduate Student)
|
