Global Vegetation

The latest quasi-operational observations of the Earth's vegetation are obtained from the moderate resolution imaging spectroradiometer (MODIS) on board NASA's EOS-Terra (launch: December 1999) and EOS-Aqua (launch: May 2002) satellites (http://eospso.gsfc.nasa. gov/eos_homepage/mission_profiles/index.php). MODIS derives from the following legacy instruments: advanced very high resolution radiometer (AVHRR), high resolution infrared radiation sounder (HIRS), Land Remote Sensing Satellite (Landsat) thematic mapper (TM), and Nimbus-7 coastal zone color scanner (CZCS). MODIS' 36-band spectroradiometer measures VIS and IR radiation with 21 spectral bands within 0.4-3.0 mm and 15 bands within 3-14.5 mm. The instrument's instantaneous field of view (FOV) at nadir is 250 m (two bands), 500 m (five bands), and 1000 m (29 bands). Derived products range from land vegetation and ocean chlorophyll fluorescence to cloud and aerosol properties, fire occurrences, surface temperatures, snow cover on land, and sea ice in the oceans. Table 1 details the spectral bands of MODIS and their key uses. A subset of the spectral bands of MODIS is to be found on the sea-viewing wide field-of-view sensor (SeaWiFS) satellite with eight bands within 0.4-0.8 mm with a spatial resolution of about 1.13 km at nadir. We focus here on satellite platforms for which data time series are available. The best spatial resolution currently available is from the IKONOS satellite (~0.3 m).

Traditionally, for the past 25 or more years, a commonly used measure of global vegetation density or vegetation vigor has been the 'vegetation index' derived from AVHRR. Ratio transforms from visible red (VIS or R) and N-IR bands from remote sensing are widely used for studying different vegetation types and land use. The first channel from the National Oceanic and Atmospheric Administration (NOAA) AVHRR is in the VIS (red) part of the spectrum where chlorophyll absorbs most of the incoming radiation, while the second N-IR channel is in a spectrum region where spongy mesophyll leaf structure reflects most of the light. This contrast between responses of the two bands is represented by the normalized difference vegetation index (NDVI: (CH-2(NIR) -CH-1(VIS))/(CH-2 + CH-1); CH stands for channel), which is correlated with global vegetation parameters such as the fraction of absorbed photosynthetically active radiation (FPAR or fPAR), chlorophyll density, green-leaf area, and transpiration rates. The VIS (red) and N-IR detectors on the AVHRR sensors record radiance in the 0.58-0.68 and 0.725-1.1 mm wavelength regions, respectively. NDVI varies theoretically between -1.0 and +1.0, and increases from about 0.1 to 0.75 for progressively increasing amounts of vegetation and is most directly related to the fPAR absorbed by vegetation canopies, and hence to photosynthetic activity of terrestrial vegetation.

NDVI has been widely used to discriminate between vegetation types and characterize seasonal phenology. Global ecosystem models have used the AVHRR NDVI as the basis to estimate net primary production (NPP) and net ecosystem carbon flux. In most cases, the assumption is made that NDVI can be used as an accurate predictor for fPAR and therefore potential NPP, for many ecosystem types.

With the advent of Terra-MODIS, an enhanced version of the vegetation index, called EVI, has been developed, which uses the additional information obtained from MODIS' expanded range of spectral channels. This additional information enables a better characterization of vegetation in both heavily forested regions such as the Amazon, as well as in semiarid regions such as the Sudano-Sahel (Figure 2).

As the seasons change, the mirror effect of seasonality is seen, with vegetation alternatively blooming and fading, and one hemisphere's vegetation is high while the other is low. A 'global animation' of the seasonal change in vegetation is shown in http://earthobservatory.nasa. gov/Newsroom/EVI_LAI_FPAR/Images/ The biweekly and monthly vegetation index maps have wide usability by biologists, natural resources managers, and climate modelers. Naturally occurring fluctuations in vegetation, such as seasonal changes, as well as those that result from land-use change, such as deforestation, can be tracked. The EVI can also monitor changes in vegetation resulting from climate change, such as expansion of deserts or extension of growing seasons.

MODIS' observations allow scientists to track two 'vital signs' of Earth's vegetation. At Boston University, a team of researchers used MODIS data to create global estimates of the green-leaf area of Earth's vegetation, called leaf area index (LAI) and the amount of sunlight the leaves are absorbing, fPAR ( modismisr/other.html). Both pieces of information are necessary for understanding how sunlight interacts with the Earth's vegetated surfaces - from the top layer, called the canopy, through the understory vegetation, and down to the ground. Figure 3 shows an example of the representation of the vegetation by MODIS-derived LAI and fPAR.

In Africa, rainfall is the most important factor that affects where people and animals live, and influences patterns of plant growth and ecosystem health. Animations of LAI and FPAR images can be viewed at LAI_wdates.mpg (LAI) and http://earthobservatory.nasa. gov/Newsroom/EVI_LAI_FPAR/Images/FPAR_wdates. mpg (fPAR).

They show the cycle of wet and dry seasons in Africa from September 2000 through May 2001 and the corresponding variation in the green-leaf area and how much sunlight the leaves are absorbing over the course of a year. The seasons in the Southern Hemisphere stand in direct opposition to those of the Northern Hemisphere, while meteorological patterns in the Northern Hemisphere roughly mirror those in the Southern Hemisphere. For example, when summer comes in the northern part of Africa in June, the winter (dry season) takes over South Africa, drying out green leaves.

Summer: May 21-July 2000 EVI (Terra-MODIS)

Summer: May 21-July 2000 EVI (Terra-MODIS)

Winter: Nov 21-July 21 2000 EVI (Terra-MODIS)

Winter: Nov 21-July 21 2000 EVI (Terra-MODIS)

Enhanced vegetation index I_Z3HBB

Figure 2 The images show EVI during two different seasons. Vegetation ranges from 0, indicating no vegetation, to nearly 1, indicating densest vegetation. Gray areas indicate places where observations were not collected. The EVI has increased sensitivity within very dense vegetation, and it has built-in corrections for several factors that can interfere with the satellite-based vegetation mapping, like smoke and background noise caused by light reflecting off soil ( EVI_LAI_FPAR/). Credit: NASA/GSFC/University of Arizona.

Enhanced vegetation index I_Z3HBB

Figure 2 The images show EVI during two different seasons. Vegetation ranges from 0, indicating no vegetation, to nearly 1, indicating densest vegetation. Gray areas indicate places where observations were not collected. The EVI has increased sensitivity within very dense vegetation, and it has built-in corrections for several factors that can interfere with the satellite-based vegetation mapping, like smoke and background noise caused by light reflecting off soil ( EVI_LAI_FPAR/). Credit: NASA/GSFC/University of Arizona.

NDVI time series data sets. The AVHRR and predecessor instruments have yielded long time series of NDVI data which have been used widely in many studies worldwide. Time series of NDVI data sets span several satellites and hence are prone to noise or error if not corrected for varying solar zenith angle due to orbital drift, differences in satellite sensors on board different spacecraft, sensor degradation, atmospheric absorption, equatorial crossing time, among others. Data input for atmospheric correction include aerosol optical depth, atmospheric water vapor, and ozone and other gas absorption. The physical products that are used to obtain the NDVI synthesis are also corrected for system errors such as misregistration of the different channels and calibration of all the detectors along the line-array detectors for each spectral band. An excellent comparison of the various NDVI data sets, such as AVHRR/NDVI-PAL (Pathfinder Land Program), Global Inventory Monitoring and Modeling Study (GIMMS-NDVI, and Systeme pour l'Observation de la Terre 4 (SPOT-4) VGT-NDVI), spanning from about 1981 to the present, is found elsewhere.

As an alternative to traditional approaches using predefined classification schemes with discrete numbers of cover types to describe a geographic distribution ofvege-tation over the Earth's land surface, Defries et al. applied a linear mixture model to derive global continuous fields of percentage woody vegetation, herbaceous vegetation, and bare ground from 8 km AVHRR. Linear discriminants for input into the mixture model are derived from 30 metrics representing the annual phenological cycle, using training data derived from a global network of scenes acquired by Landsat. The results suggested that the method yields

December 2000

December 2000

Fraction of photo-synthetically active L_eaf_areaindex(LAI) ___radiation (fPAR)_.

Figure 3 Examples of leaf area index (LAI) and fraction of photosynthetically active radiation (fPAR) derived from MODIS for Africa during December 2000. LAI is defined as the one-sided green-leaf area per unit ground area in broadleaf canopies, or as the projected needle-leaf area per unit ground area in needle canopies. fPAR is the fraction of photosynthetically active radiation absorbed by vegetation canopies. Color code: LAI -colors range from low LAI (0.0-0.1 is yellow) to mid-range LAI (between 2.0 and 3.0 is red) to high LAI (shades of purple); fPAR -low fPAR is in yellow (0.0-0.1, mid-range fPAR is in blues and red/brown (~0.2-0.4), and high fPAR is in shades of purple with light purple being the highest at 1.0. Images are from http:// Original graphics credit: John Weier with design by Robert Simmon, Boston University ( Three image sequences showing September 2000, December 2000, and April 2001 are to found in modismisr/laifpar/lai.afr.jpg and modismisr/laifpar/fpar.afr.jpg.

reliable products that overcome apparent problems with artifacts in the multiyear AVHRR data set due to calibration problems, aerosols and other atmospheric effects, bidirectional effects, changes in equatorial crossing time, and other factors.

Land surface and vegetation classification. For global studies, the land surface and vegetation are classified into broad categories that represent large-scale aspects that can be monitored from space as well as used in land surface models that are coupled to other models of the atmospheric general circulation and climate. A typical classification would be as described by Defries et al. in 2002.

Broadleaf evergreen forest

Coniferous forest and

and woodland


Broadleaf deciduous forest

High-altitude deciduous

and woodland


Mixed forest and woodland

Wooded grassland


Shrubs and bare ground


Bare ground



There are several variations to the above. For example, alternative versions add water bodies, cropland, urban and built-up, and barren, to render the classification system more compatible with those of the International Geosphere-Biosphere Programme (IGBP). The software and land-cover classification system developed for the Food and Agricultural Organization (FAO) and the United Nations Environment Programme (UNEP) are well described by Di Gregorio. The above also represents the typical land surface/vegetation classification system used in global climate system models. These land surface models compute the exchange of energy, water, momentum, and carbon between the biosphere and the atmosphere. They also account for the hydraulic and thermal properties of different soil types. More complex models are used to represent subpixel distributions and species composition. Conservation strategies for managing biodiversity have traditionally assumed that species distributions change relatively slowly, unless they are directly affected by human activity. However, there is increasing recognition that such strategies must include the effects or impacts of global climate change. Satellite-derived NDVI can be most useful for the development and validation of biome models. At more regional and local scales, in situ data are usually needed.

Was this article helpful?

0 0
10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook

Post a comment