CO₂-induced heat source changes over the Tibetan Plateau in boreal summer-Part I: the total effects of increased CO₂

Abstract

The concentration of CO₂ in the atmosphere has been increasing, but its effects on the heat source (HS) over the Tibetan Plateau (TP) are unclear. Aimed at understanding these effects, at first, present study evaluated the CMIP5 (phase 5 of the Coupled Model Intercomparison Project) models and found that their multi-model ensemble (MME) reproduces the spatial pattern of the HS over the TP during June–September (hereafter JJAS) in observation reasonably well. Then, we used the MME to investigate the response of the JJAS HS over the TP to increased CO₂. In response to increased CO₂, the JJAS HS increases significantly. In terms of the response pattern and TP-averaged results, the increase in HS is mainly contributed by the latent heating (LH), which is due to moisture increases (with the lower level stronger than the upper level) and evaporation intensification led by CO₂ change. The leading two intermodel spreads feature a nearly uniform structure and a central-southeastern TP dipole structure, respectively, and account for half of the total intermodel variance. The latent heating is mainly responsible for the spreads. The intensified radiative cooling of the atmosphere slightly dampen the TP-averaged HS increases. Over the TP, when CO₂ increases, the atmospheric column above warms. Accordingly, the net longwave radiation flux out of the atmosphere column enhances, resulting in the intensified radiative cooling over the TP.

Appendix: The responses of shortwave radiation fluxes

In response to increased CO₂, the increase of net shortwave radiation flux into the atmosphere column above the TP is 4.5 W m⁻². The detailed responses of the shortwave radiation fluxes are provided below:

The surface downwelling shortwave radiation flux In response to increased CO₂, this radiation flux decreases during JJAS. The TP-averaged decrease of the CMIP5 MME is 6.8 W m⁻² in the 1%CO₂ results (Fig. 3). The response of the downwelling shortwave radiation at the TP surface features a dipole pattern, with a distinct decrease over the central TP and slight increase over the southeastern and northwestern TP (Fig. 11a). Comparison of the decrease in the all-sky results with the clear-sky results indicates that the inclusion of cloud–radiation feedback does not significantly alter the TP-averaged responses (Fig. 3), but enlarges the uncertainty. The fact that the clear-sky downwelling shortwave radiation uniformly decreases at the TP surface (figure not shown) implies that the inclusion of cloud–radiation feedback slightly modifies the response pattern of the downwelling shortwave radiation at the TP surface. For the reduction in the average of the downwelling shortwave radiation at the TP surface, the intensification of atmospheric absorption of shortwave may be responsible. This intensified absorption is probably led by atmosphere wetting. In response to increased CO₂, the moisture increases due to the atmospheric warming. The water vapor is able to absorb shortwave (Yang et al. 2006).

Fig. 11

The MME response (color shading) of JJAS downwelling shortwave radiation (a units: W m⁻²), upwelling shortwave radiation (b units: W m⁻²), albedo (c units: dimensionless) and snow cover (d units:  %) at the surface for the 1%CO₂ results. The contours represent elevations of 1500, 3000, 5000 and 6000 m; the lattices indicate the response reaching the 95% significance level. Due to lack of data of snow cover, the following models are not participating in the calculation: ACCESS1-0, ACCESS1-3, BNU-ESM, CCSM4, FGOALS-s2, GFDL-CM3, GFDL-ESM2G, GFDL-ESM2M, HadGEM2-ES, IPSL-CM5A-LR, IPSL-CM5A-MR and IPSL-CM5B-LR

The surface upwelling shortwave radiation flux Meanwhile, this shortwave radiation flux decreases more than the downwelling shortwave radiation. In the 1%CO₂ results, the averaged surface upwelling shortwave radiation of the CMIP5 MME over the TP decreases by 12.1 W m⁻² (Fig. 3). The response of the upwelling shortwave radiation features a nearly uniform weakening at the TP surface (Fig. 11b). The weakening is relatively large at elevations above 3000 m. The upwelling shortwave radiation at the surface is mainly the reflection of the downwelling shortwave radiation. The larger change in the upwelling shortwave radiation than that of the downwelling shortwave radiation indicates that the albedo of the TP surface changes. The albedo of the TP also uniformly reduces, with relatively large magnitude above 3000 m (Fig. 11c). The response pattern of snow cover is almost the same as that of albedo, with the spatial correlation coefficient between them being 0.94. This means that the shrinking of snow cover in response to increased CO₂ leads to a reduction in albedo as well as the upwelling shortwave radiation at the TP surface. The inclusion of cloud–radiation feedback may slightly enhance the overall upwelling shortwave radiation at the TP surface, but the enhancement is statistically insignificant (Fig. 3).

The outgoing shortwave radiation flux at the top of the atmosphere Over the TP, this outgoing radiation flux reduces significantly. In the 1%CO₂ results, the TP-averaged magnitude of the CMIP5 MME is 9.6 W m⁻² (Fig. 3). The TP displays a uniform decrease in this radiation, with relatively large magnitude over the western TP at elevations between 3000 and 5000 m (Fig. 12). The response pattern is similar to that of the upwelling shortwave radiation at the TP surface (Fig. 11b), with a spatial correlation coefficient of 0.63 between the two. It indicates that the reduction of the outgoing shortwave radiation is mainly led by the change in the upwelling shortwave radiation at the TP surface. The discrepancies in the detail may be caused by the masking effect of cloud or the atmospheric absorption to shortwave. Overall, the model results suggest that the inclusion of cloud–radiation feedback may not significantly affect the response of the TP-averaged outgoing shortwave radiation (Fig. 3). But it is important to note that the change in radiative fluxes at the top of atmosphere, resulting from the cloud-radiation feedback, is the largest source of uncertainty in the climate response to CO₂ forcing simulated by GCMs because of the unrealistic presentation of cloud processes in the models.

Fig. 12

The MME response (color shading) of JJAS upwelling shortwave radiation (units: W m⁻²) at the top of the atmosphere for the 1%CO₂ results. The contours represent elevations of 1500, 3000, 5000 and 6000 m; the lattices indicate the response reaching the 95% significance level

The downwelling shortwave radiation flux at the top of the atmosphere It is not analyzed because it barely changes (Fig. 3).