In Detail: Drought monitoring using the Standard Vegetation Index (SVI)

Spectral vegetation indices are among the most commonly used satellite data products for evaluation, monitoring, and measurement of vegetation cover, condition, biophysical processes, and changes. This recommended practice shows how to apply a multi-temporal analysis of MODIS-based Standard Vegetation Index (SVI) to support drought monitoring and early warning. The method was developed by UFSM in Brazil in cooperation with UN-SPIDER within the SEWS-D project. It has been applied in different countries in the Central American Dry Corridor and in the Dominican Republic in the Caribbean. For more information on the SEWS-D project see the dedicated webpage (SEWS-D project).

The practice on drought monitoring and early warning has been adapted to the programming languages R and Python and gradually updated, for example by including a cloud mask, which has not been considered before. Also, an additional R script is provided to be able to compute the SVI for large areas (>200.000km2).

For related information and drought definitions see the related recommended practice on drought monitoring using the Vegetation Condition Index developed by the Iranian Space Agency in cooperation with UN-SPIDER.

Definition of the SVI

Originally building on the NDVI anomaly concept, the Standardized Vegetation Index (SVI) developed by Peters et al. (2002) describes the probability of variation from the normal NDVI over multiple years of data (e.g., 12 years), on a weekly time step. The SVI is a z-score deviation from the mean in units of the standard deviation, calculated from the NDVI or EVI values for each pixel location of a composite period for each year during a given reference period. The equation below shows the general calculation of the SVI

Equation Standardized Vegetation Index

where zijk is the z-value for the pixel i during week j for year k, VIij is the weekly VI value for pixel i during week j for year k whereby both the NDVI or EVI can be utilized as VI, µij is the mean for pixel i during week j over n years, and σij is the standard deviation of pixel i during week j over n years.

This recommended practice calculates the SVI based on the Enhanced Vegetation Index (EVI) which has some advantages compared to the NDVI such as an improved sensitivity over dense vegetation conditions and is less affected by aerosol influences.

More detailed information on the Standardized Vegetation Index can be found here.

SVI based on EVI or NDVI

The SVI can be calculated based on the Normalized Difference Vegetation Index (NDVI) as well as on the Enhanced Vegetation Index (EVI) depending on the user's preference. Both products can be easily downloaded from AppEEARS. However, the SVI based on the EVI can be preferred in certain circumstances to the NDVI given following advantages:

  • No distortions in the reflected light caused by the particles in the air
  • No distortions in the reflected light caused by ground cover below the vegetation
  • The EVI data product does not become saturated as the NDVI when viewing rainforests and large amounts of chlorophyll

Cloud mask

In order to mask out pixels that are impacted by atmospheric interferences such as clouds as well as snow and ice cover, the Pixel Reliability Quality Assurance (QA) layer of MOD13Q1 is being used. The layer classifies the quality of the vegetation index in the following categories:

"Good" and "marginal" data in the pixel reliability bands are accepted as sufficient quality and will be considered for the analysis. Other filter values than "0" and "1" are used to mask out the corresponding EVI pixels.

Find more information about the MODIS Land Products Quality Assurance here

Input data:

  • MODIS MOD13Q1 EVI or NDVI data: full-time series (2000 until to date) for the specific days of the year that are of interest in the selected geographic region (available from AppEEARS)
  • MOD13Q1 Pixel Reliability data (available from AppEEARS)

Software:

  • R (free statistical software)
  • Python: To run the Python version via the provided Jupyter Notebook you need python installed on your computer. We recommend downloading and installing Anaconda 3 (https://www.anaconda.com/download/) with the python 3.6 version as it includes a lot of useful packages
  • (Modis Reprojection Tool (MRT))

This recommended practice has been developed for applications in drought early warning systems and can be applied globally.

Strengths

  • MODIS data is freely available and easy to access, so is the software (R and MRT) used in this practice.
  • The MOD13Q1 product has already undergone important preprocessing steps like geometric and atmospheric correction. The vegetation indices NDVI and EVI are readily available, which makes the dataset easy to use.
  • MOD13Q1 also contains quality information per pixel, which can be used for quality assessment of the final product.
  • The Pixel Reliability QA layer, also included in MOD13Q1, is used to mask out pixels that are impacted by atmospheric interferences such as clouds which improves the information of the final SVI maps.

Limitations

  • MODIS sensors are mounted on two satellite platforms: Terra (launched in December 1999) and Aqua (launched in May 2002). This means, the time series go back to 2000/2002, which is not very long from a climatologic viewpoint. Since the SVI is using the mean and standard deviation of the time series, it is desirable to use a representative time span. It is, however, possible to extend the time series with AVHRR data provided that a thorough inter-calibration between the two sensor systems is given.
  • The SVI can only provide a relative comparison of the vegetation condition while the assessed deviation from the mean vegetation condition cannot be translated into an absolute deviation of for example the plant height. Neither can the SVI be interpreted for an absolute quantification of agricultural damage.

For more detailed information on the strengths and limitations of the Standardized Vegetation Index, visit the explanatory page of the SVI.