NASA Vegetation Index Products

Product User GuidePI - Kamel Didan (University of Arizona)
Product ATBDCo-Is: Compton Tucker (GSFC), Armando Barreto (U of A), Jorge Pinzon (SSAI/GSFC)


Vegetation Indices from the Suomi NPP VIIRS Sensor Extending the EOS-MODIS VI Science Algorithm

Demonstrating continuity of VI product

Showing 1:1 correlation with AQUA NDVI, differences mostly due to cloud, data QA issues,
and snow

This product build on the ~34 year multi-sensor VI record (AVHRR + MODIS + VIIRS) and continues this highly valuable and unique long term VI data record (NDVI, EVI, EVI2), and establishes a future and longer term outlook for VI climate data records and Earth System Data Records(continuity).

The proposed product suite includes a MODIS continuity plan with the EVI-3 index (at all resolutions). However EVI has some issues that were typically addressed by adopting a backup algorithm (the previous soil-adjusted vegetation index [SAVI] then a new 2-band EVI) in addition to the blue band incompatibility and poor signal-to-noise ratio. We plan to introduce the new 2-band EVI and begin the phase out of the 3-band EVI).

We plan to use the MODIS QA constrained view angle maximum values composite method (16-day and monthly), sorting the data based on QA, and then selecting the observations with the smallest view zenith angle. We also plan to introduce a new dynamic compositing algorithm (quasi-daily), to reduce data loss in vegetated areas not prone to clouds and improve vegetation monitoring and change detection/phenology work.


Product Details and Links to LPDAAC

Product Name Product ID
NPP/VIIRS Vegetation Indices 16-day 500m VNP13A1
NPP/VIIRS Vegetation Indices 16-day 1km VNP13A2
NPP/VIIRS Vegetation Indices Monthly 1km VNP13A3
NPP/VIIRS Vegetation Indices 16-day CMG VNP13C1
NPP/VIIRS Vegetation Indices Monthly CMG VNP13C2

 


Validation Plan

Our validation efforts will build on existingl MODIS team validation efforts witha focus on leveraging other efforts, such as the National Ecological Observatory Network (NEON) and the National Phenology Network (NPN). We will utilize the the VIIRS subsets, AERONET sites, and higher resolution data (Landsat ). We will participate in relevant field campaigns whenever appropriate and possible, budget permittin, and will establish and maintain regular interaction with the JPSS cal/val team.

We aim to attain CEOS validation stage 1 within the first three years pending production progress, status of the land surface reflectance product, and other constraints. In addition, we plan to devote some resources and effort to assessment of the effects of aerosols and BRDF. These are critical issues that need to be addressed collectively by the land team.

NDVI Variance

Characterization of the sources of variation in the VI records by:

  • Determining the variance distributions with respect to the values of our VIIRS vegetation indices
  • Quantification of the VI spatial (δxy) and temporal (δt) coherence variances, VI errors resulting from surface reflectance uncertainty (εSR), the seasonal variance from seasonal phenology (δS), and the inter-annual variance as our VIIRS VI records increase with time (δIA).
  • Produce the same variance distributions for the MODIS Aqua and Terra VI records and compare to VIIRS to assist in continuity
  • These metrics will be produced byregion, continent, and globally to serve as a quantitative error metric