LAND SURFACE MODEL

The land surface model SiB2 is used to estimate the transfer of energy, mass and momentum between the atmosphere and the vegetated surface. The atmospheric boundary conditions necessary to force SiB2 include air temperature T_m, vapor pressure e_m, wind speed u_m, precipitation P, shortwave downward radiation R_sd, incoming longwave radiation R_ld, and CO₂ and O₂ concentration c_m and o_m at a reference level, z_m, within the atmospheric boundary layer. In practice, mean values of c_m and o_m be defined (35 and 2090 Pa, respectively) for current atmospheric conditions. Daily air temperature, vapor pressure, wind speed and precipitation data from over 8000 worldwide stations was available at Global Surface Summary of Day Data Version 6 (GSSDD; URL: http://www.ncdc.noaa.gov). The station observation data was interpolated to ten kilometers grids for DBHM model with thin plate splines. The shortwave downward radiation was estimated from meteorological observations following Revfeim (1997) and Yang et al. (2001). The incoming longwave radiation was estimated as (Brunt, 1932; Jiménez et al., 1987):

R_ld = 3#34#4T_a⁴ (1)

where 5#5 is the atmospheric emissivity, 6#6 = 5.6704×10^-8 is the Stefan-Boltzmann constant (W ^. m ^{-2 .} K^-4), and T_a is the air temperature (K). The atmospheric emissivity 5#5 is presented as a function of water vapor pressure:

3#3 = 0.66 + 0.0397#7 (2)

where e_a is the vapor pressure (kPa).

Time invariant vegetation and ground parameters and time varying vegetation parameters of SiB2 were obtained from satellite data or assigned following Sellers et al. (1996). SiB2 land cover is available at USGS Global Land Cover Characterization data (GLCC; URL: http://edcwww.cr.usgs.gov/landdaac/glcc/). Global LAI and FPAR based on Pathfinder Version 3 Normalized Difference Vegetation Index (NDVI) with spatial resolution 16 kilometers mesh could be obtained from Myneni et al. (1997). The Food and Agriculture Organization (FAO) global soil type map (Source: Land and Water Development Division, FAO, Rome) was used to produce the DBHM grid soil properties such as soil water potential at saturation 8#8(m), soil hydraulic conductivity at saturation K_s (m/s), soil wetness parameter b, porosity 9#9 and averaged slope S_s. Serval empirical formulas are suggested to estimate the soil hydraulic and thermal parameters (Cosby et al., 1984):

10#10 = - 0.01×10^{(1.88-0.0131 . P_sand)} (3)

K_s = 7.0556×10^{(-6.884+0.0153 . P_sand)} (4)

b = 2.91 + 0.159 ^. P_clay (5)

11#11 = 0.489 - 0.00126 ^. P_sand (6)

where P_sand and P_clay are percentages of sand and clay in a dry soil, respectively. Soil optical properties are assigned by vegetation type following Sellers et al. (1996).