AVHRR Global Production Efficiency Model (GloPEM)



The Global Production Efficiency Model (GloPEM) generates global maps of Net Primary Production as well as maps of the many other global variables.


The objective of the research driving development of GloPEM is the ability to use remotely-sensed observations of surface spectral reflectance and thermal emission to model and monitor terrestrial net primary production and gross primary production at the global scale. Global data sets of such variables are being collected by the Oak Ridge DAAC for validation purposes.


GloPEM is the first attempt to utilize the production efficiency concept globally, in which the canopy absorption of photosynthetically active radiation (APAR) is used with a conversion "efficiency" or carbon yield of APAR in terms of gross primary production (GPP). The GloPEM model is thus based on physiological principles, in particular the amount of carbon fixed per unit APAR is modeled rather than fitted using field observations.


The approach is unique in that it uses satellite data to measure both APAR and the environmental variables that affect the utilization of APAR in primary production. The use of satellite measurements gives global, repetitive, spatially contiguous and time specific observations of actual (rather than potential) vegetation conditions. Because all the information is derived from satellite observations, the model is responsive to real events such as El Nino - Southern Oscillation (ENSO), volcanic eruptions and other forms of Global Environmental Change being studied as part of NASA's Earth Science Enterprise


The results of the modeling work show that there are significant possibilities of inferring biological and environmental variables using multispectral techniques that need to be explored if the new generation of satellite remote sensing systems is to be exploited productively.

Components of the Modeling Approach

  • Advanced Very High Resolution Radiometer (AVHRR) images at an 8 km resolution from the AVHRR Pathfinder Project are used to provide global coverage in 10-day time steps. These data are radiometrically calibrated and cloud-screened but are not corrected for atmospheric attenuation.
  • The amount of incident Photosynthetically Active Radiation (PAR) is derived from Total Ozone Mapping Spectrometer (TOMS) ultraviolet observations of cloud cover, which are used to modify incident PAR as derived from a clear sky model. This is the only variable used in the model that is not directly inferred from the AVHRR instrument.
  • Surface reflectance properties in the visible and infrared wavelengths are converted to spectral vegetation index (SVI) values which are linearly related to the fraction of incident PAR absorbed by terrestrial vegetation (Fpar). When combined with incident PAR this provides a measure of canopy light absorption (APAR).
  • The minimum value of visible reflectance in the annual observation period is related to the amount of standing above-ground biomass, taking into account the effects of solar zenith angle and cloud shadows.
  • Surface radiometric temperature (Ts) and atmospheric column precipitable water vapor amount (U) are derived from thermal measurements in different spectral wavelength bands (the "split-window" approach).
  • The regression relationship between a moving window array of SVI and Ts values (termed TvX) is used to derive an estimate of ambient air temperature (Ta) by extrapolating to a high SVI value (~0.9) that represents an infinitely thick canopy. It is assumed that canopy and air temperature are equivalent at this point.
  • The atmospheric water vapor amount (U) is extrapolated to the surface and used to estimate surface humidity and dewpoint temperature. When combined with Ta this is used to calculate vapor pressure deficit (VPD).
  • An index of soil surface moisture is derived from changes in the slope of the TvX relationship through time. The potential amount of carbon fixation per unit APAR is calculated from the quantum yield of photosynthesis and a
  • climatological mean air temperature to differentiate between photosynthetic pathways (C3, C4).
  • Potential carbon fixation is reduced by "stress" factors related to plant physiological control (i.e., Ta, VPD, soil moisture) to derive actual carbon fixation in the form of gross primary production (GPP).
  • Respiration related to the growth and maintenance of biomass is subtracted from GPP to derive global net primary production (NPP).

Alterations to GloPEM at the GLCF

The Global Land Cover Facility (GLCF) processed the GloPEM data collection in order to provide users with easier access. The original GloPEM data is in the form of IEEE 32-bit float binary data without geographical header. The GLCF processed all the data into GeoTIFF format which is readily useful in most commercial image processing software without needing extensive manual input. The GLCF also produced pseudo-color previews using an advanced signal processing concept: LLOYDS Quantization, which guarantees lower data loss than any other currently available pseudo color coding scheme. The color previews were then assembled into gif animations showing the year-round trends.

Copyright © 1997-2016 University of Maryland. All rights reserved
University of Maryland National Aeronautics and Space Administration GOFC_GOLD