Fig. 1.
Schematic diagram of the terrestrial ecosystem model (TEM). Boxes show state variables: vegetation carbon (CV); vegetation nitrogen (NV); soil organic carbon (CS); soil organic nitrogen (NS); and available soil inorganic nitrogen (NAV). Arrows show carbon and nitrogen fluxes: gross primary production (GPP); autotrophic respiration (R A); litterfall carbon (LC); litterfall nitrogen (LN); nitrogen uptake by vegetation (NUPTAKE); net nitrogen mineralization of soil organic nitrogen (NETNMIN); outside nitrogen inputs (NINPUT) and nitrogen loss from the ecosystem (NLOST) (More details, see Raich et al., 1991, McGuire et al., 1992).
Table 1. Linear relationship between growth temperature and optimum temperatures [referring to (Kattge and Knorr, 2007) Table 3, used in eq. (12) in this paper]
[i] T opt(V cmax) and T opt(J max) are the optimum temperature for plant photosynthesis for V cmax and J max in Farquhar et al. model.
Table 2. Base temperature-related parameters in TEM for various vegetation types in this study
[i] Units are °C.
Fig. 2.
Effects of monthly mean temperatures on GPP, R m and R H. (a) relative response of GPP to air temperature at different acclimation stages. Base thermal condition: t growth=21.8°C, T max=34°C, T opt=30°C, T min=0°C; warmest condition: t growth=34.1°C, T max=38.7°C, T opt=34.7°C and T min=0°C; coolest condition: t growth=20.2°C, T max=33.4°C, T opt=29.4°C and T min=0°C. b that used for calculating shifts of these parameters is set to be 0.385 as the average of b for T opt(V cmax) and T opt(J max). Base growth temperatures are extracted as the annual mean air temperature of the grid cell in our regional dataset in 2000; warmest and coolest grow temperatures are determined by the maximum and minimum annual mean air temperature for the same grid. (b) Response of plant maintenance respiration (R m) to air temperature with fixed Q10=2.0 and variable Q10 that used in TEM with different vegetation carbon pool sizes. Parameter K rb is set to be a typical value as −5.28. (c) Response of heterotrophic respiration (R H) to soil temperature at different soil organic carbon pool sizes.
Fig. 3.
Potential vegetation coverage of forest region and climatic zones in the conterminous United States at a resolution of 0.5°×0.5° (longitudes×latitudes). The mixed forest is defined in TEM as the combination of 50% deciduous forest and 50% coniferous forest.
Fig. 4.
Variations of air temperature in the 1990s and the 21st century under the IPCC SRES climate scenarios. ‘max T’, ‘mean T’, ‘min T’ represent the annual maximum, mean and minimum air temperatures, respectively.
Fig. 5.
Annual maximum air temperature from the monthly dataset of the most extreme and modest climate scenarios (A1FI and B1 respectively) in 2000, and the 2050s as well as 2090s.
Fig. 6.
Temporal variations of GPP, NPP and NEP simulated with the previous and the revised versions of TEM from 1990 to 2100 over the study region and each climatic zone. ‘F’ represents GPP, NPP or NEP. The subscript ‘O’ indicates the results estimated by a previous version of TEM and the subscript ‘T’ indicates the results estimated by the revised TEM with incorporation of the temperature acclimation effect. ‘std’ stands for the standard deviation of using T opt(V cmax) and T opt(J max) and their standard errors in the linear relationship of temperature acclimation.
Fig. 7.
Temporal variations of R A and R H simulated with the previous and the revised versions of TEM from 1990 to 2100 over the study region and each climatic zone. ‘F’ represents R A or R H. The subscripts and ‘std’ have the same meaning in Fig. 6.
Fig. 8.
Temporal variations of VEGC and SOC simulated with the previous and the revised versions of TEM from 1990 to 2100 over the study region and each climatic zone. ‘S’ represents VEGC or SOC. The subscripts and ‘std’ have the same meaning in Fig. 6.
Table 4. Table S1. Estimated carbon dynamics for the 21st century under the IPCC SRES climate scenarios. The subscript ‘O’ indicates the results estimated by the previous version of TEM and the subscript ‘T’ indicates the results estimated by the revised TEM with incorporating the temperature acclimation effect. ‘Mean’ and ‘std’ stand for the average and the standard deviation of the 100-yr data, respectively. Supplementary Materials
Units: Pg C VEGCT 71.99 9.91 70.30 8.44 68.69 6.43 69.51 7.13 VEGCO 64.74 4.60 67.15 5.00 68.12 5.94 68.24 5.85 SOCT 22.20 1.35 23.19 0.68 23.96 0.90 23.66 0.78 SOCO 21.89 1.66 22.95 0.90 23.85 0.81 23.48 0.69
Table 5. Table S2. Euclidean Distances (ED) between the estimated carbon dynamics by the two versions of TEM under the IPCC SRES climate scenarios in the 21st century. Units are Pg C.
Fig. 9.
Seasonal variations of estimated GPP and NEP in the 21st century under the IPCC SRES climate scenarios. The plots are the 100-yr average values and the ‘std’ represents the standard deviation of the 100-yr results.
Fig. 10.
Variations of estimated ratio of plant respiration and photosynthesis (R/P) by the previous and the revised versions of TEM from 1990 to 2100 over the study region and each climatic zone. The subscripts and ‘std’ have the same meaning in Fig. 6.
Fig. 11.
Spatial distribution of the differences between GPP and NEP estimated by the two versions of TEM under the A1FI climate scenario. The differences are calculated by the revised TEM GPP and NEP minus the results estimated by the previous version of TEM.
Fig. 12.
Variations of ET and WUE simulated with the previous and the revised versions of TEM from 1990 to 2100 over the study region and each climatic zone. ‘X’ represents either ET or WUE. The subscripts and ‘std’ have the same meaning in Fig. 6.
Table 3. Differences of the estimated carbon dynamics in the 21st century between using 1-yr time-scale of temperature acclimation, 1-month, and 10-yr time-scales
[i] The numbers are calculated by using 1-yr time-scale results minus the results with other time-scales. Units are Pg C yr−1.