POES Science Applications

AVHRR Vegetation Applications

Vegetation monitoring- AVHRR data provides an opportunity for studying and monitoring changes in surface vegetation conditions in different ecosystems around the gobe from agricultural assessments and land cover mapping of forests, grasslands, tundra, etc.  

The Global  Vegetation Index (GVI) is an AVHRR applications that makes use of the global area coverage (GAC) data to make a product called the normalized difference vegetation index (NDVI).  GAC data are processed daily and then composited on a weekly schedule to produce global scale vegetation maps which illustrate vegetation.  

Using the near infrared channel (channel two) on the AVHRR sensor, chlorophyll pigment has a signficant spectral reflectance, "green" vegetation has about a 60% reflectance in the spectral range from 0.7-1.3 micrometer.  By comparison, in the visible channel, vegetation has about a 20% reflectance (0.4-0.5 micrometer range).   Recall that differential reflectance of these two bands is one of the ways that land surface cover (vegetation) was detected using the Landsat MSS and Thematic Mapper sensors.  The differential reflectance (difference in radiance between these two channels) can be used to classify land cover, estimate agricultural acreage, and even assess plant stress/health.

The NDVI is defined by the following equation:

NDVI= (CH2-CH1)/(CH2+CH1)

Where CH1 and CH2 are channels 1 and 2 (visible and near-infrared, respectively) on the AVHRR sensor.   Using this channel combination, clouds, water, and snow/ice surfaces have higher reflectance in the visible than in the near-infrared, so the NDVI values for these features is negative.  Surfaces like soils and rock have very similar reflectance values in both bands, and therefore do not give a signficant response, however, vegatation has higher reflectance in the infrared, and the NDVI values for vegetation range from 0.1 to 0.6, with larger values indicative of greater spatial density and more "greeness" of the plant canopy.  The presence of dust or sub-pixel cloud elements will increase the scattering in the visible relative to the infrared, and therefore the apparent reflectance of underlying vegatated surfaces will be reduced.  

If you are interested in further information on the algorithm, and the different versions of correction that have gone into improving the NDVI product, or if you intend to include NDVI in your final class project, you should go to this link to the NOAA Global Vegetation Index User's Guide.   Below are a set of sample images (as thumbnails) which link to the User's Guide set of examples.

2nd Generation Weekly NDVI - 1996 Series, Polar Stereo Northern Hemisphere
Series of 4 IThe following section contains 24 thumbnail images of selected Second and Third Generation GVI data arranged in a matrix of 6 rows x 4 columns. The months of February, May, August and November are represented in each row of images. The first two rows contain the same data but are displayed in different map projections. The images are the scaled Normalized Differential Vegetation Index (NDVI) data in Polar Stereographic (Northern Hemisphere only) and Mercator projections, respectively. The third and fourth rows contain Third Generation NDVI data B-level (weekly) and D-level (climatology) images, respectively, for 1996. The fifth and sixth rows contain the Third Generation Precipitable Water Index (PWI) data B-level and D-level images, respectively, for 1996. Detailed information on GVI, NDVI, PWI, Second Generation, Third Generation and the different data levels can be found in the NOAA Global Vegetation Index Users Guide. Later in this section, these same 24 images are displayed slightly differently. Each row of four images fits on one screen, so that the data can be compared between the different months. If a user wants to examine an image in more detail, they can click anywhere on the image and display the full size image.

Images processed by Ralph E. Meiggs, Physical Scientist, NCDC.

mages Displayed
2nd Gen Weekly NH
FEBRUARY		 2nd Gen Weekly NH
MAY		 2nd Gen Weekly NH
AUGUST		 2nd Gen Weekly NH
NOVEMBER
2nd Generation Weekly NDVI - 1996 Series, Mercator
Series of 4 Images Displayed
2nd Gen Weekly Mercator
FEBRUARY		 2nd Gen Weekly Mercator
MAY 		2nd Gen Weekly Mercator
AUGUST		 2nd Gen Weekly Mercator
NOVEMBER
3rd Generation B-level NDVI -1996 Series, Plate Carrée
Series of 4 Images Displayed
3rd Gen B Level NDVI
FEBRUARY		 3rd Gen B Level NDVI
MAY		 3rd Gen B Level NDVI
AUGUST		 3rd Gen B Level NDVI
NOVEMBER
3rd Generation D-level mean NDVI - 1996 Series, Plate Carrée
Series of 4 Images Displayed
3rd Gen D Level NDVI
FEBRUARY		 3rd Gen D Level NDVI
MAY		 3rd Gen D Level NDVI
AUGUST		 3rd Gen D Level NDVI
NOVEMBER
3rd Generation B-level PWI - 1996 Series, Plate Carrée
Series of 4 Images Displayed
3rd Gen B Level PWI
FEBRUARY		 3rd Gen B Level PWI
MAY		 3rd Gen B Level PWI
AUGUST		 3rd Gen B Level PWI
NOVEMBER
3rd Generation D-level mean PWI - 1996 Series, Plate Carrée
Series of 4 Images Displayed
3rd Gen D Level PWI
FEBRUARY		 3rd Gen D Level PWI
MAY		 3rd Gen D Level PWI
AUGUST		 3rd Gen D Level PWI
NOVEMBER

More information on AVHRR datasets, including the vegetation indices can be found through the NASA Distributed Data Archive, found at the following link:

Global Land Biosphere Data and Resources

NOTE:  These are some  additional links you might begin to explore in looking for applications which address the science issue you are researching for your final class project.

back to Syllabus EVSC494