Generating Global NPP Map for Estimating Agricultural Productivity
Analysis
Data
NDVI is derived from NOAA AVHRR (the Advanced Very High Resolution Radiometer) 8 km data set. The data is 10-days composite originally created from daily data and using Goode Homolosine projection. The orginal scaled digital NDVI (0-255 )
Geophysical NDVI=(digital NDVI-128)*0.008
PAR data set was provided by Prof.Dennis G.Dye (Dept. of Geography). Boston University) and was estimated by 370 nm reflectivity data from the Nimbis-7 Total Ozone Mapping Spectrometer (TOMS). (Dye et. al.,1993) the data set contains monthly PAR value [MJ/m2] and spatial coverage is 66N0 to 66S0 with a spatial resolution of 10 by 10.
Data Processing
1. Generating NDVI monthly composite
select the maximum value of NDVI 10 day composite of the same month (which is called the Maximum Value Composite, MVC), that is because lower values are considered to be measured when clouds cover the surface from the sun.
2. Transferring the NDVI data from Goode Homolosine projection to longitude/latitude projection.
3. Interpolation of the PAR data from 1 deg grid cell to 8 km grid cell
The resolution of PAR data is about 100 km. So interpolate it spatial and make 8 km resolution data set. The method of interpolation is bi-linear interpolation method using
four neighboring points.
4.Calculation of f
APAR
In this study. we used the f
APAR -NDVI relationship equation:
f APAR =-0.025+1.25*NDVI
which was derived from corrected and filtered NDVI with relatively high accuracy. (Ruimy,et al,1994)
5. Calculation of NPP
Use the formula below.
NPP=eS(fAPAR * PAR)
in this research, conversion efficiency is treated as constant, globally
e =1.5[g/MJ]
as shown by Goward et.al(1991).
Then, we will compare the accumulated NPP over each year to FAO production statistics. The procedure is also illustrated in Figure 2
Figure 2: Schematic Diagram of this study
