Normalized Differential Vegetation Index (NDVI)

Normalized Differential Vegetation Index (NDVI) is one of the most widely used vegetation indices with applications in ecology, forestry, agriculture, wildlife and other disciplines. The NDVI dataset is based on the Moderate Resoluation Imaging Spectroradiometer (MODIS) data. All raster files found in the viewer are based on the methods to fit double logistic curves to NDVI time series in Biscof et al. 2012. The only departures from Bischof et al's methods are: (1) we used 8 day NDVI composite instead of 16 day; and (2) we used snow cover data for each pixel to extend the “winter” period (i.e., floored NDVI values) until there is no snow cover for 2 consecutive weeks.

DLC

dlcparams: fitted parameter values (and their standard errors) of the fitted double logistic regression curve for each pixel. There are 8 layers per year, representing: "xmidS", "scalS", "xmidA", "scalA", "xmidS_SD", "scalS_SD", "xmidA_SD", "scalA_SD". Value ranges: xmidS = 75-228; scalS = 1-20; xmidA = 167-288; scaleA = 1-35.

Moderate Resoluation Imaging Spectroradiometer (MODIS)

The Moderate Resolution Imaging Spectroradiometer (MODIS) is an instrument that collects remotely sensed data used by scientists for monitoring, modeling, and assessing the effects of natural processes and human actions on the Earth's surface.

cleanVals: cleaned 8 day NDVI values. Cleaning includes: flooring during winter, NA snow days after winter is over, running median filter, and scaled between 0 and 1.

fittedVals: fitted values from the fitted double logistic curve on a daily basis for the entire year. Values are scaled between 0 and 1 for the entire time series (i.e., all years combined). One file for each year (i.e., 365 layers in each file). *Note that we don’t store these data anymore because is it such as large dataset! You’ll have to extrapolate out from dlcparams file.

IRGVals: first derivative or Instantaneous Rate of Green-up (IRG) of the scaled fittedVals on a daily basis for the entire year. Only for spring period of curve. One file for each year (i.e., 365 layers in each file). Values are also scaled between 0 and 1. *Note that we don’t store these data anymore because is it such as large dataset! You’ll have to extrapolate out from dlcparams file.

sumIRG: cumulative summation over time of the IRG values of each pixel on a daily basis for the entire year. One file for each year (i.e., 365 layers in each file). Values are positive. *Note that we don’t store these data anymore because is it such as large dataset! You’ll have to extrapolate out from dlcparams file.

deriv2: second derivative of the fitted double logistic curve on a daily basis for the entire year. Only for spring period of curve. Values were calculated from scaled (between 0 and 1) the first derivative or Instantaneous Rate of Green-up (IRG) values. One file for each year (i.e., 365 layers in each file). *Note that we don’t store these data anymore because is it such as large dataset! You’ll have to extrapolate out from dlcparams file.

maxIRG : value of the maximum IRG. Units are change in scaled NDVI per day. Each layer represents a year.

scalS: fitted spring scale parameter of the double logistic curve. This value relates to maxIRG. Each layer represents a year.

xmidS_SD: Standard Error of fitted spring midpoint parameter (i.e., Julian day of peak rate of increase in NDVI) of the double logistic curve. Each layer represents a year.

maxIRGdate: Julian date that maximum IRG occurs. Each layer in the file represents a year.

maxNDVIdate: Julian date that maximum NDVI occurs. Each layer in the file represents a year.

springStart : Julian date that spring starts. This represents the 1st second derivative of the NDVI curve. Each layer in the file represents a year.

springEnd: Julian date that spring ends. This represents the 2nd second derivative of the NDVI curve. Each layer in the file represents a year.

Lquantile and Uquantile : Values of the Lower 0.025 and Upper 0.975 quantile of the 14 year NDVI time series for each pixel used to scale the time series between 0 and 1 for double logistic curve fitting. It is a single raster layer.

csumNDVI : cumulative summation (or integrated) of NDVI over the course of a year. These values were calculated by un-scaling the fittedVals using the Lquantile and Uquantile, then subtracting the minimum value (i.e., flooring the baseline of the NDVI time series to 0), then running cumulate summation of these values. One file for each year (i.e., 365 layers in each file). *Note that we don’t store these data anymore because is it such as large dataset! See csumNDVImax for the sum of the values for each year.

csumNDVImax: summation of csumNDVI (also known as integrated NDVI) . Each layer in the file represents a year

eMODIS

eMODIS provides annual (1 value per year per grid cell) phenological metrics at 250 m resolution from 2001 to present based on an analysis of a time series of the Normalized Difference Vegetation Index (NDVI) for a given year from satellite imagery captured from the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite. DEVISE includes the following variables: Start of season time (day of year denoting the start of measurable photosynthesis activity), Maximum time (day of year where photosynthesis activity peaked), End of season time (day of year denoting the end of measurable photosynthesis activity), Duration (number of days where photosynthetic activity is detected), and Time Integrated NDVI (a metric of cumulative photosynthetic activity for the year).

Start of season time (SOST): Beginning of measurable photosynthesis in the vegetation canopy (day of year). Calculated on a cell by cell basis on a single temporal scale (annually) with the statistics max, mean, and median.

Maximum Time (MAXT): Time of maximum photosynthesis in the canopy (day of year). Calculated on a cell by cell basis on a single temporal scale (annually) with the statistics max, mean, and median.

End of Season Time (EOST): End of measurable photosynthesis in the vegetation canopy (day of year). Calculated on a cell by cell basis on a single temporal scale (annually) with the statistics max, mean, and median.

Duration (DUR): Length of photosynthetic activity (number of days). Calculated on a cell by cell basis on a single temporal scale (annually) with the statistics max, mean, and median.

Time Integrated NDVI (TIN): Canopy photosynthetic activity across the entire growing season (interpolated NDVI). Calculated on a cell by cell basis on a single temporal scale (annually) with the statistics max, mean, and median.