Greenness of the Conterminous U.S.
Greenness Home |
What's New | What is NDVI | What is AVHRR | View Images | Products/Services | Links | Contact Us |
What's New?
The USGS EROS Data Center (EDC) has been producing greenness information of the conterminous U.S. and Alaska using AVHRR data for over 10 years. In 2001, EDC will begin to produce greenness products with improved atmospheric correction for the effects of water vapor, ozone, and Rayleigh scattering.
What is the benefit of atmospheric correction?
The signal received by an earth observing satellite sensor is affected by different atmospheric effects. The effect the atmosphere will have on the signal depends on the wavelength of the signal, the composition of the atmosphere, and the path from the sun to the surface and to the sensor. Wide field of view (FOV) sensors, such as AVHRR, can show considerable scan-angle effects due to changes in the atmospheric path in the cross-scan direction.
The atmosphere scatters and absorbs radiation between the sun and the earth along the solar path and again between the earth and the satellite sensor along the view path. Absorption is the process by which radiant energy is absorbed and converted into other forms of energy. The most efficient absorbers of solar radiation are water vapor, carbon dioxide, and ozone. The AVHRR channel 1 (0.58 - 0.68 µm) is weakly affected by ozone absorption. Water vapor absorption bands near 0.9 µm and 1.1 µm affect measurements in the near infrared region where reflectance from healthy vegetation occurs. The AVHRR channel 2 (0.75-1.1µm) is considerably affected by water vapor.
The water vapor reduces the observed near infrared reflectance observed at the satellite. In addition, the longer path length from the sun - to the surface - to the satellite, the greater the effect that water vapor has on lowering the observed near-infrared reflectance measured at the satellite. As a result, off nadir AVHRR channel 2 observations are more affected than near nadir observations.
The fact that near nadir observations are less affected has been a major advantage in the historic production of the biweekly greenness products. The biweekly composites are constructed from daily observations. EDC uses the maximum NDVI compositing. In other words, the greenest pixel value over the biweekly period is chosen for the composite.
Choosing the greenest pixel has been shown to compensate for undesirable atmospheric and satellite view angle conditions. For example, the selection of the greenest pixel implies that the observation was acquired under conditions when the atmosphere contained the least amount of water vapor. This is because the presence of a high concentration of water vapor lowers the channel 2 response and effectively lowers the derived NDVI value. Also, the shortest path length, which occurs at nadir, is least affected by water vapor concentration. Hence, in theory, the maximum value compositing process selects near nadir observations with the least affect from water vapor.
While in theory maximum value compositing can minimize atmospheric and viewing effects, other factors need to be considered. Most important is the presence of cloud cover in the observations. In cases where cloud cover is persistent over a region, the only cloud free observation may be off-nadir relative to the region. In this case the use of maximum value compositing does not compensate for view angle conditions or for atmospheric conditions. In these circumstances atmospheric correction is most beneficial.
NVDI values computed from atmospherically corrected data are generally higher (greener) than NDVI from uncorrected data. This is due primarily to the effect of water vapor on AVHRR channel 2. Hence a composite produced from corrected daily observations will have a different characteristics than a composite produced from uncorrected data. The most notable will be the selection of pixels from different days and a higher average NDVI for each composite.
The normalized difference vegetation index (NDVI) is used to produce the greenness products. The NDVI is computed using the formula:
This provides the normalized difference between the near infrared and the red reflectance. For vegetation, this relationship quantifies the health, vigor, and quantity of vegetation on the land surface.
The corrections for ozone and Rayleigh scattering are straightforward (Teillet, 1991). Appropriate Rayleigh scattering correction must include an adjustment for atmospheric pressure, which can be derived from the elevation of the target. Recommended reference values for Rayleigh optical depths for standard pressure and temperature conditions are available (Teillet, 1990a; El Saleous et al., 1994). The local elevation adjustment can be derived using GTOPO30 (Gesch and Larson, 1996) which is the best available global digital terrain data at this time and for this task has sufficient accuracy.
The correction for ozone absorption can be based on concentration values from standard climatic tables with latitudinal and seasonal dependence (Teillet, 1992) or based on actual measurements derived from the Total Ozone Mapping Spectrometer (TOMS) or other appropriate sensors (El Saleous et al., 1994).
Until recently it was not possible to acquire daily global water vapor concentration fields in a time frame suitable for near real time production. The routine availability of daily global water vapor concentration fields from NOAA and NASA now makes it possible to perform a reliable water vapor correction to AVHRR data.
What does this mean to our users today?
We recognize that many users utilize the entire 12-year series of greenness information. The data are used to establish historic trends and relationships that are important in the determination of the "normal" condition and in the analysis of change. So it is important to the USGS to deliver a stable and scientific product.
The goal is to have the entire time series (1989 to present) of greenness products consistently processed using atmospheric correction. Since it is impractical to reprocess 12 years of daily observations to accomplish this goal, the atmospheric correction will be applied to the Channel 1 and Channel 2 data in the existing time-series of composites. Using the date band, we can identify the date of observation of each pixel. The historical daily water vapor and ozone fields for each day will be utilized to correct each pixel. The NDVI will be computed from the corrected channel 1 and 2 data.
During the 2001 growing season we will simultaneously produce the 2001 composites and reprocess the historical data. The processing of the historical data will be done in step with the current 2001 composite period. Hence, the complete time series of corrected data for each compositing period will be made available each week, along with the 2001 composite. In addition, the USGS will provide an updated "average" and "departure from average" for composite periods throughout the growing season. This will allow users to continue to use "departure from average" analysis.
Any comments or questions can be addressed to:
U.S. Department of the Interior | U.S. Geological Survey
URL: http://ivm.cr.usgs.gov/
Page Contact Information: custserv@usgs.gov
Page Last Modified: June 2007
