What are vegetation indices used for?
Vegetation indices derived from Earth observation satellites are important for a wide range of applications such as vegetation monitoring, drought studies, agricultural activities, climate and hydrologic modeling. Vegetation monitoring plays an important role in drought early warning systems, which help to anticipate the risk of food crises at local and global scale.
How are vegetation indices measured from space?
Optical satellite sensors measure the solar radiation reflected from targets on the ground. Multispectral optical sensors are multichannel detectors with a few spectral bands. Each channel is sensitive to radiation within a narrow wavelength band such as the blue, green, red, near infrared, or short wave infrared band. The reflection of radiation by vegetation shows low values in the blue and red band, slightly higher values in the green band, very high values in the near infrared band, and low to high values in the shortwave infrared bands (depending on the wavelength). Very characteristic for the vegetation spectrum is the steep increase of reflectance from red to near infrared, the so called “red edge”. Multispectral sensors do not “see” the whole electromagnetic spectrum but their bands cover parts of the spectrum that are characteristic for different land cover types.
Depending on the intended applications, the design of multispectral sensors differs in number of bands, bandwidth, and wavelengths covered. The sensor RapidEye, for example, was designed for vegetation studies and agricultural applications, and has therefore been especially equipped with a separate red edge band, which is quite unique. The MSi sensor onboard Sentinel-2 will also be equipped with a three new bands in the red-edge region. Commonly used sensors for vegetation studies include the Advanced Very High Resolution Radiometer (AVHRR) onboard NOAA and MetOp, the Moderate-resolution Imaging Spectroradiometer (MODIS) onboard Terra and Aqua, or the VEGETATION 1 and 2 sensors onboard SPOT 4 and SPOT 5.
Spectral indices make use of the characteristics in the spectrum of the respective material of interest. For example, the Normalized Difference Vegetation Index (NDVI) makes use of the red and near infrared bands since the energy reflected in these wavelengths is clearly related to the amount of vegetation cover on the ground surface. Not all spectral indices can be applied to all sensors. When selecting data to calculate indices, the basic requirement is that the the sensor covers the bands used in the index. The index DataBase (IDB) by Henrich et al. gives a comprehensive overview on which indices match with which sensors.
The IDB lists 66 different vegetation indices. In general terms, vegetation indices are dimensionless radiometric measures, which combine information from different channels of the electromagnetic spectrum to enhance the vegetation signal. They measure the spatial and temporal variation of the plant’s photosynthetic activity. Due to their simplicity, vegetation indices are widely used. The most commonly used vegetation indices include: Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Difference Vegetation Index (DVI), and others.
The below images show a comparison of NDVI and EVI based on MODIS for Central America. The data are freely available as MODQ1 product (cf. links to the data below). In the upper image pair data of the wet season in July 2012 show that the NDVI saturates in high biomass areas. The lower image pair with data of the dry season in February 2001 shows that EVI shows lower values in the dry areas, which makes it more difficult to distinguish different biomass. One advantage of EVI is that it includes the blue band making the index less sensitive to atmospheric effects.
In a review paper on vegetation indices and biophysical variables (Clevers 2014, doi:10.1007/978-94-007-7969-3_22), the author states that “Various studies have compared many indices. The performance of the various indices always is different, depending on the specific data sets used for the study, resulting each time in different indices as being the best one. […] One should always consider the theoretical background of an index, its validity range and its purpose, and then use one index as much as possible rendering results that are mutually comparable spatially and temporally”.
Vegetation indices are highly correlated with biophysical variables such as the Leaf Area Index (LAI), Canopy Water, fraction of absorbed photosynthetically active radiation (fAPAR),.and vegetation cover fraction (fCover) the latter two being Essential Climate Variables (ECVs). Radiative transfer models like the PROSAIL model are used to determine the biophysical variables based on reflectance or vegetation indices.
How can I access vegetation indices?
Some vegetation indices are available as free geoinformation items with global coverage and frequent updates, for example:
STAR Global Vegetation Health Products (NOAA) (link to the data)
Agricultural Stress Index System (ASIS, FAO) (link to the data)
Locust Watch (IRI) (link to the data)
Global Agricultural Drought Monitoring and Forecasting System (GADMFS-CSISS) (link to the data)
MODIS Vegetation products (NASA) (link to the data)
NDVI Central Asia (FEWS NET-USGS, USAID) (link to the data)
NDVI Middle East (FEWS NET-USGS, USAID) (link to the data)
NDVI Central America (FEWS NET-USGS, USAID) (link to the data)
NDVI Africa (FEWS NET-USGS, USAID) (link to the data)
Vegetation indices, especially the freely available time series of NDVI and EVI, are commonly used in drought monitoring. The Iranian Space Agency, one of UN-SPIDER Regional Support Offices, has prepared a recommended practice on drought monitoring, which is used for monitoring the impacts of meteorological drought on natural vegetation, i.e. rain fed, range land, and forest. Based on maximum value composites of MODIS NDVI (250m spatial resolution), the Vegetation Condition Index (VCI) is calculated to assess whether the state of the vegetation in a current month is better or worse compared to same month of the previous years. The values range between zero and one: a value of one indicates that the NDVI value of the current month is the maximum of the time series; a value of zero indicates that the NDVI value of the current month is the minimum of the time series; and all values in between indicate where the observed value is situated between the maximum and minimum. Detailed step-by-step procedures how to prepare the MODIS data, how to calculate the VCI, and how to visualize the results are available here on the Knowledge Portal.