Version 4.8 NPP/GPP Global sums(posted: 01 Feb 2007)

Here are the global sums for the NTSG in-house version 4.8 MOD17A3 (The version in the granule ID is 205).

Year	GPP (Pg C/yr)	NPP (Pg C/yr)
2000	110.963804	53.420217
2001	111.259097	53.422021
2002	107.495863	50.864508
2003	108.146107	51.404846
2004	108.902730	52.165953
2005	107.648816	50.932493

Version 4.8 NPP/GPP Products now available.(posted: 31 Aug 2006)

NTSG's enhanced MODIS NPP/GPP product has been updated to version 4.8 (granuale ID uses collection # 205). The hdf granuales are available on our ftp server: MOD17A2, MOD17A3.

Four Years of MOD17 Annual NPP(posted: 27 Feb 2004)

All the images have been regenerated with the latest version of Enhanced NPP (NTSG in-house MOD17A3, collection 4.5, product version number 105 for those who are counting). There are four years of data and images: 2000-2003. It should be noted that since MOD15 data starts on year day 57 of 2000, the periods prior to this were calculated as the average of 2001-2003. If you are using MOD17A2 granules, please pay attention to the QC field, which will indicate that no FPar/Lai values were produced for the early 2000 periods.

The only algorithmic difference between this version of NTSG Enhanced MOD17 and the version distributed at AGU is that this one has tighter input QC tolerances.

The following is a table of global GPP and NPP for 2000 to 2003 produced by Maosheng Zhao.

	GPP(PgC)	NPP(PgC)
2000	108.42	56.06
2001	110.76	57.74
2002	107.82	55.53
2003	107.50	54.80

2000 and 2001 were La Niņa years. 2002 and 2003 were weak El Niņo years.

Enhanced Global Annual NPP Images Available(posted: 05 Dec 2003)

All of the images found here have been regenerated using an enhanced MOD17A3 algorithm. It differs from the official collection 4 MOD17 in that it temporally interpolates missing/unreliable MOD15 inputs based on QC and smoothes/interpolates the DAO inputs. More information can be found in the new version (2.0) of the MOD17 User's Guide. Along with each data raster (image) you will find a matching QC raster. The QC field for MOD17A3 is defined as the percent of interpolated MOD15 inputs. Also, you will find histograms of both the data and QC fields. There are 12 histograms for each raster image. There is a histogram of all values, as well as a histogram for each land cover class. The granules from which these images were generated can be found here on our ftp site.

DAO Grid map issue in all MOD17 Collection 4(posted: 29 Aug 2003)

A new DAO Ver 4.0.2 was recently incorporated into MOD17, and it has now been discovered that the new DAO data was being incorrectly mapped into the MOD17 projection. All MOD17 collection 4 data produced so far is affected and the inconsistency has now been fixed in the MOD17 code. The cell footprint in the MOD17 images will change from 1.0 x 1.0 deg to 1.0 x 1.25 deg. We consider quality of the Collection 4 MOD17 data to not be impacted by this shift in DAO mapping, because the DAO cell boundaries were artificial anyway.