Documenting the effectiveness of the phenology metrics is difficult given the current scarcity of ground verification data at the appropriate scale (i.e., vegetation-canopy phenology). However, current efforts by the National Phenology Network to build and organize ground observation databases will go far toward addressing these shortcomings. In the meantime, two examples of coincident ground observations and phenology metrics can serve as a validation demonstration.
In the first example (see graph to the right), ground observations of grass height are compared to AVHRR NDVI and derived phenology metrics. Partners from The Nature Conservancy (TNC) made observations of grass height in the Tallgrass Prairie Preserve, Oklahoma, on a weekly basis for two years (1995 and 1996; green bars). We extracted AVHRR NDVI data over the same sites and same time period (black line), smoothed the data (red line), and extracted phenology metrics (SOS as a yellow diamond and EOS as a blue diamond). Grass height and NDVI data closely mirror each other and the phenology metrics correspond well to periods of increasing (SOS) and decreasing (EOS) growth. Interestingly, TNC personnel noted a period of drought (red triangle) during the summer of 1996 that corresponds to a period of decreased NDVI and subsequent reduced Time-Integrated NDVI.
In a second example (see graph right), AVHRR NDVI was plotted against gross primary production (GPP) over an Agriflux site in Mandan, North Dakota, for the year 2000. Rising NDVI (blue line) corresponded well to carbon uptake (magenta line) and the SOST phenology metric (blue circle) corresponded well to increasing carbon uptake at the beginning of the growing season.
Below are other science projects associated with this project.
Methods for Deriving Metrics
Methods for Deriving Metrics
Challenges in Deriving Phenological Metrics
Deriving Phenological Metrics from NDVI
Data Smoothing - Reducing the "Noise" in NDVI
Documenting the effectiveness of the phenology metrics is difficult given the current scarcity of ground verification data at the appropriate scale (i.e., vegetation-canopy phenology). However, current efforts by the National Phenology Network to build and organize ground observation databases will go far toward addressing these shortcomings. In the meantime, two examples of coincident ground observations and phenology metrics can serve as a validation demonstration.
In the first example (see graph to the right), ground observations of grass height are compared to AVHRR NDVI and derived phenology metrics. Partners from The Nature Conservancy (TNC) made observations of grass height in the Tallgrass Prairie Preserve, Oklahoma, on a weekly basis for two years (1995 and 1996; green bars). We extracted AVHRR NDVI data over the same sites and same time period (black line), smoothed the data (red line), and extracted phenology metrics (SOS as a yellow diamond and EOS as a blue diamond). Grass height and NDVI data closely mirror each other and the phenology metrics correspond well to periods of increasing (SOS) and decreasing (EOS) growth. Interestingly, TNC personnel noted a period of drought (red triangle) during the summer of 1996 that corresponds to a period of decreased NDVI and subsequent reduced Time-Integrated NDVI.
In a second example (see graph right), AVHRR NDVI was plotted against gross primary production (GPP) over an Agriflux site in Mandan, North Dakota, for the year 2000. Rising NDVI (blue line) corresponded well to carbon uptake (magenta line) and the SOST phenology metric (blue circle) corresponded well to increasing carbon uptake at the beginning of the growing season.
Below are other science projects associated with this project.