SEARCH

SEARCH BY CITATION

FilenameFormatSizeDescription
gcb12474-sup-0001-TextS1-2-TableS1-FigS1-9.pdfapplication/PDF9633K

Data S1. GPP SOM.pdf contains.

Data S2. Meteorology from global/reanalysis datasets.

Table S1. Mean annual GPP biases (ΔGPP) in the ISAM-CRUNCEP and the ISAM-NCEP simulations at each site. ΔCRUNCEP = ISAM-CRUNCEP – ISAM-FLUXNET; ΔNCEP = ISAM-NCEP – ISAMFLUXNET.

Figure S1. Geographical distribution of FLUXNET sites used in this study.

Figure S2. Relative differences in mean annual meteorology variables between two global reanalyses datasets used in this study.

Figure S3. Spin-up of deep soil temperature (TSoil50 m) at each site vs. years of spin-up at each site. The initial soil temperature is 1°C (i.e., 274.15 K).

Figure S4. Analysis for tree PFTs (Trop.BET, Trop.BDT, Temp.BDT, NET): (a) Daily GPP vs. daily meteorology for Tavg, Srad and Q. The R2 for each regression fit (line denoted as: All) were calculated using data from all the simulations. (b) Annual GPP vs. annual total Precip for different simulations, along with the respective R2.

Figure S5. Analysis for individual Trop.BET and Trop.BDT sites: Daily climatology of GPP and corresponding Tavg, Srad and Q. Each row of subplots corresponds to an individual site (site name on left corner). Each column shows a respective variable (GPP, Tavg, Srad, Q). For each subplot, the x-axis is the ‘day of year,’ and the y-axis is the respective variable.

Figure S6. Analysis for individual nontree C3 sites: Daily climatology of biases (Δ) in reanalyses-driven climatology of GPP, Tavg, Srad, and Q. Each row of subplots corresponds to an individual site (site name on left corner). Each column shows a respective variable (ΔGPP, ΔTavg, ΔSrad, ΔQ). For each subplot, the x-axis is the ‘day of year,’ and the y-axis is the respective variable. All the biases (Δ) were calculated with respect to the ISAM-FLUXNET counterpart.

Figure S7. Map of C4 fraction in herbaceous plant functional types, for (a) savanna, grass, and pasture, and (b) crop.

Figure S8. Map of vegetated land area mask used for the comparison of GPP estimates. For consistency, nonvegetated (i.e., barren land, desert, glacier, water body) grid cells were removed from all the GPP estimates (ISAM, FLUXNET-MTE, MODIS) during the construction of respective zonal means.

Figure S9. Partitioning of site-level GPP biases into sunlit and shaded components.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.