Remote Sensing
- Estimation of Carbon Dioxide Exchange
- Anatoly Gitelson, Don Rundquist, Bryan Leavitt, Galina Keydan, Rick Perk, and Veronica Ciganda
- Project Goal
- Develop quantitative remote sensing techniques to estimate biophysical properties of crop
- fraction of absorbed photosynthetically active radiation
- green vegetation cover
- green biomass
- net ecosystem CO2 exchange (NEE)
- Project Description
- We are investigating the relationship between remotely measured reflectance (sun light reflected by crops)
and tower-based measurements of CO2 fluxes. The project was carried out in three large production
fields;
each field was equipped with tower eddy covariance flux instrumentation and supporting meteorological sensors.
- Spectral
radiometric measurements were made in the visible and near infra-red spectral regions using two hyperspectral
radiometers mounted on “Goliath”, an all-terrain sensor platform (Figure 1). To calculate reflected light, we measured
simultaneously upwelling radiance and incident irradiance. From reflectance spectra we were able to retrieve such
biophysical crop characteristics as biomass and greenness which are a proxy of crop photosynthetic activity and, thus,
CO2 fluxes.
| |
Figure 1. “Goliath”, an all-terrain sensor platform measuring sun
light reflected by crop reflectance spectra in different stages of corn development. |
|
- We also acquired imagery by an AISA hyperspectral imaging system (Figure 2), with 35 spectral bands between 400
and 900 nm. Measurements were made from an altitude of ~1000 m, providing a spatial resolution of ~3 m/pixel. Output
includes distribution of biomass (or greenness) and crop photosynthetic activity (or vigor).
| |
Figure 2. CALMIT’s
AISA Imaging Spectrometer is programmable from 1-286 spectral channels (430-900 nm). |
|
- Progress
- We developed techniques for quantitative assessment of vegetation cover, biomass and CO2 fluxes (Figures
3 and 4) in irrigated and rainfed maize and soybean. The techniques allow us to make synoptic estimates of these
important crop biophysical characteristics, to calculate carbon sequestered by crops, to detect early stages of crop
stress and to predict yield in precision agriculture.
| |
Figure 3. Carbon dioxide flux and its remote estimate (where
ρRed Edge is the light reflectance around 700 nm and ρNIR is the
near-infrared (NIR) reflectance (>750 nm).
| |
Figure 4. Comparison of measured and predicted CO2 fluxes.
| Date |
Irrigated Maize (Field 1) |
Irrigated Soybean (Field 2) |
Rainfed Soybean (Field 3) |
| June 21, 2002 |
 |
 |
 |
| June 27, 2002 |
 |
 |
 |
| July 12, 2002 |
 |
 |
 |
| July 15, 2002 |
 |
 |
 |
| September 7, 2002 |
 |
 |
 |
| September 17, 2002 |
 |
 |
 |
| |
Figure 5. Maps of Net Ecosystem Carbon Dioxide Exchange
for irrigated maize and irrigated and rainfed soybean fields for the 2002 growing season. Data was developed
from AISA hyperspectral imagery. The maps show how crop greenness and CO2 fluxes change with crop
development and maize or soybean phenology. It also quantitatively shows how much higher maize CO2
fluxes are than soybean fluxes and it shows that rainfed soybean sequesters less carbon than irrigated
soybean. |
|
| |
Figure 6. Mid-day tower-based measured NEE vs. estimated
NEE for maize and soybean fields generated from six AISA hyperspectral images taken between June and September,
2002. |
|
- Staff
- Anatoly Gitelson
Team Leader and Professor (SNRS & CALMIT)
- Don Rundquist
Professor (SNRS) and CALMIT Director
- Bryan Leavitt
Technician (CALMIT)
- Galina Keydan
Programmer (CALMIT)
- Rick Perk
Assistant Geoscientist (CALMIT)
- Veronica Ciganda PhD Student (Agronomy and Horticulture & CALMIT)
|
