1. What is PEN (Phenological Eyes Network) ? It is a network of ground observation sites for long-term continuous validation of terrestrial ecological remote-sensing. It started in 2003 in Japan. Most sites are located at Asia-Flux network sites (with carbon and water flux observations). Why long-term and continuous? We want to capture the seasonal and inter-annual change pattern of ecosystem by satellites. Because ecosystem is continuously changing, validation of the satellite data should be also continuous. Automatic observations ADFC (Automatic Digital Fisheye Camera) Device:Nikon Coolpix 4300/4500 + FC-E8 converter Target:sky image... validation of cloud flags canopy image... phenology and LAI HSSR (Hemi-Spherical Spectral Radiometer) Device:Eko MS700 + automatic rotating system Target:sky... PAR, PAR/shortwave ratio canopy... spectrum, NDVI, EVI, PRI, etc. Skyradiometer (sunphotometer) Device:PPEDE POM-01 Target:direct sunlight... optical thickness diffuse sunlight... aerosol etc. canopy directional radiance... BRDF Phenological Eyes Network (PEN): A Ground Network for Satellite Ecological Data Validation Kenlo NISHIDA (University of Tsukuba) kenlo@sakura.cc.tsukuba.ac.jp Global Vegetation Workshop 2006, August 7-10, 2006, University of Montana PEN sites TGFTERC Grass FieldGrass 2003- TKYTakayama Flux SiteDBF 2003- TFSTomakomai Flux SiteDNF 2004- (down) FJHFuji-north Flux SiteDNF 2006- KEWKiryu Expr. WatershedENF 2004- MSEMase Flux Site Rice paddy2005- RHNResearch Institute, HNUrban 2005 SGDSugadaira Flux Site Grass 2005- PEN sites TGFTERC Grass FieldGrass 2003- TKYTakayama Flux SiteDBF 2003- TFSTomakomai Flux SiteDNF 2004- (down) FJHFuji-north Flux SiteDNF 2006- KEWKiryu Expr. WatershedENF 2004- MSEMase Flux Site Rice paddy2005- RHNResearch Institute, HNUrban 2005 SGDSugadaira Flux Site Grass 2005- Manual observation Leaf gas exchange Device:LI6400 Target:physiological phenology Leaf optics Device:FieldSpec + LI1800 integrating sphere Target:optical phenology Leaf in situ observation / sampling Device: SPAD, ruler, etc. Target: leaf number, size, angle, pigment, LMA Litter fall Device: litter trap Target: species composition annual maximum LAI LAI indirect measurement Device: LAI2000, TRAC Target: LAI, FPAR transmit reflect Mase paddy (MSE) litter fall + leaf in situ obs. MODIS (MOD15) overstory understory 0 5 1010 1515 2020 10 0 15 0 20 0 25 0 30 0 2003 2004 Amax (  mol m -2 s -1 ) 0 5 1010 1515 2020 10 0 15 0 20 0 25 0 30 0 2003 2004 Amax (  mol m -2 s -1 ) 0 5 1010 1515 2020 10 0 15 0 20 0 25 0 30 0 DOY Amax (  mol m -2 s -1 ) This study was supported by the S-1 program: “Integrated Study for Terrestrial Carbon Management of Asia in the 21 Century Based on Scientific Advancements” funded by the Global Environmental Research Fund of the Ministry of the Environment of Japan. every 10 minutes every 2 minutes every 20 minutes

2. 01/0102/0101/0102/0103/0104/0105/0106/0107/0108/0109/0110/0111/0112/01 reflectance Ground validation of MODIS in terms of NDVI, EVI, CO 2 -uptake, phenology events, and spectrum canopy (downward) forest floor canopy (upward) All the MODIS data are after cloud-screening based on the MOD09GQK QC flag NDVI EVI NDVI CO 2 uptake Phenology Spectrum leaf fall leaf grow snow melt leaf fall leaf grow snow melt leaf fall leaf grow melt snow Conclusions - MOD09 cloud flag does not detect all the cloudy cases (error is about 7%). - NDVI is well represented by L1B rather than MOD09. This may be due to troubles in the atmospheric correction applied to MOD09. - Change of forest floor (snow, melt, understory) is well captured with NDVI rather than EVI. - MOD15 LAI gives reasonable annual-maximum LAI. But the seasonal change of MOD15 LAI does not correspond well with the ground observation. - MOD09 cloud flag does not detect all the cloudy cases (error is about 7%). - NDVI is well represented by L1B rather than MOD09. This may be due to troubles in the atmospheric correction applied to MOD09. - Change of forest floor (snow, melt, understory) is well captured with NDVI rather than EVI. - MOD15 LAI gives reasonable annual-maximum LAI. But the seasonal change of MOD15 LAI does not correspond well with the ground observation.