Fig. 1.
Schematic diagram of the regional terrestrial ecosystem (TECO-R) model for inversion analysis of carbon residence time. NPP, which is modelled by maximum light-use efficiency (ɛ*) (CASA model), is allocated to plant tissues (leaf q L , stem q W and root ) based on allocation coefficients (α L , α W , α R ). Plant tissues enter into fine litter (q F ), coarse litter (q C ) and soil organic carbon pools (q S ) through litterfall. Decomposed litter releases part of the carbon to the atmosphere, and the rest transfers into the soil (θ C , θ F ). Through mechanical breakdown, part of the coarse litter becomes fine litter (η). To reflect the differences in soil profile, roots and soil organic carbon pools are divided into three layers (0–20, 20–50 and 50–100 cm). Each pool has an associated carbon residence time, τ k (k=1, 2, 3).
Table 1. Symbol and definition of parameters, their lower (LL) and upper limits (UL) and other constraints for inverse analysis. Unit is gC MJPAR −1 for ɛ * and years for carbon residence times. Allocation and partitioning coefficients are dimensionless.
Fig. 2.
Comparisons between modelled and observed data lumped by each biome for 13 data sets. Biomes: ENF, evergreen needleleaf forest; DBF, deciduous broadleaf forest; MF, mixed forest; W, woodland; WG, wooded grassland; S, shrubland; G, grassland; and C, cropland. Unit with dimensions is not shown.
Fig. 3.
Inversion results showing the histograms of 22 estimated parameters and cost function with 40 000 samples from M–H simulation for evergreen needleleaf forest (ENF). See Table 1 for parameter abbreviation. Unit with dimensions is not shown.
Fig. 4.
Inversion results showing the histograms of 17 estimated parameters and cost function with 40 000 samples from M–H simulation for Grassland. See Table 1 for parameter abbreviation. Unit with dimensions is not shown.
Fig. 5.
MLEs or means of estimated parameters for eight biomes. Error bars represent standard deviations (SDs) of parameters calculated from 40 000 samples of M–H simulation. Parameters θS1 and θS2 were not shown due to their consistence between biomes (0.05±0.03). See Fig. 2 for biome abbreviations and Table 1 for parameter abbreviations.
Fig. 6.
Spatial patterns of ecosystem carbon residence times (a), its standard deviation (SD, b) and coefficient of variation (CV=SD/mean, c) in terrestrial ecosystems of the conterminous USA.
Fig. 7.
Carbon allocation coefficients (a) and the averaged ecosystem C residence time (b) for eight biomes. See Fig. 2 for biomes abbreviation. Horizontal dash line represents the average ecosystem C residence time in the conterminous USA. Error bars represent standard deviation.
Fig. 8.
The potential of ecosystems C uptakes (a), their coefficient of variance (b) and proportion of soil to ecosystem C uptake in the conterminous USA under the actual NPP increases from Hicke et al. (2002). The positive value means C sink while the negative value mean C source.
Fig. 9.
Yearly ecosystem and soil C uptake in different biomes under the actual NPP increases from Hicke et al. (2002). Error bars represent standard deviation (SD).
