The radiative impact of clouds strongly depends on their partitioning between liquid and ice phases. Until recently, however, it has been challenging to unambiguously discriminate cloud phase in a number of important global regimes. CloudSat and CALIPSO supply vertically resolved measurements necessary to identify clouds composed of both liquid and ice that are not easily detected using conventional passive sensors. The capability of these active sensors to discriminate cloud phase has been incorporated into the fifth generation of CloudSat's 2B‐FLXHR‐LIDAR algorithm. Comparisons with Clouds and the Earth's Radiant Energy System fluxes at the top of atmosphere reveal that an improved representation of cloud phase leads to better agreement compared to earlier versions of the algorithm. The RMS differences in annual mean outgoing longwave (LW) radiation gridded at 2.5° resolution are 4.9 W m−2, while RMS differences in outgoing shortwave (SW) are slightly larger at 8.9 W m−2 due to the larger diurnal range of solar insolation. This study documents the relative contributions of clouds composed of only liquid, only ice, and a combination of both phases to global and regional radiation budgets. It is found that mixed‐phase clouds exert a global net cloud radiative effect of −3.4 W m−2, with contributions of −8.1 W m−2 and 4.7 W m−2 from SW and LW radiation, respectively. When compared with the effects of warm liquid clouds (−11.8 W m−2), ice clouds (3.5 W m−2), and multilayered clouds consisting of distinct liquid and ice layers (−4.6 W m−2), these results reinforce the notion that accurate representation of mixed‐phase clouds is essential for quantifying cloud feedbacks in future climate scenarios.

This study revisits the classical problem of quantifying the radiative effects of unique cloud types in the era of spaceborne active observations. The radiative effects of nine cloud types, distinguished based on their vertical structure defined by CloudSat and CALIPSO observations, are assessed at both the top of the atmosphere and the surface. The contributions from single- and multilayered clouds are explicitly diagnosed. The global, annual mean net cloud radiative effect at the top of the atmosphere is found to be −17.1 ± 4.2 W m−2 owing to −44.2 ± 2 W m−2 of shortwave cooling and 27.1 ± 3.7 W m−2 of longwave heating. Leveraging explicit cloud base and vertical structure information, we further estimate the annual mean net cloud radiative effect at the surface to be −24.8 ± 8.7 W m−2 (−51.1 ± 7.8 W m−2 in the shortwave and 26.3 ± 3.8 W m−2 in the longwave). Multilayered clouds are found to exert the strongest influence on the top-of-atmosphere energy balance. However, a strong asymmetry in net cloud radiative cooling between the hemispheres (8.6 W m−2) is dominated by enhanced cooling from stratocumulus over the southern oceans. It is found that there is no corresponding asymmetry at the surface owing to enhanced longwave emission by southern ocean clouds in winter, which offsets a substantial fraction of their impact on solar absorption in summer. Thus the asymmetry in cloud radiative effects is entirely realized as an atmosphere heating imbalance between the hemispheres.

Observational benchmarks of global and regional aerosol direct radiative effects, over all surfaces and all sky conditions, are generated using CloudSat's new multisensor radiative fluxes and heating rates product. Improving upon previous techniques, the approach leverages the capability of CloudSat and CALIPSO to retrieve vertically resolved estimates of cloud and aerosol properties required for complete and accurate assessment of aerosol direct effects under all conditions. The global annually averaged aerosol direct radiative effect is estimated to be 21.9 W m 22 with an uncertainty range of 60.6 W m 22 , which is in better agreement with previously published estimates from global models than previous satellite-based estimates. Detailed comparisons against a fully coupled simulation of the Community Earth System Model, however, reveal that this agreement on the global annual mean masks large regional discrepancies between modeled and observed estimates of aerosol direct effects. A series of regional analyses demonstrate that, in addition to previously documented biases in simulated aerosol distributions, the magnitude and sign of these discrepancies are often related to model biases in the geographic and seasonal distribution of clouds. A low bias in stratocumulus cloud cover over the southeastern Pacific, for example, leads to an overestimate of the radiative effects of marine aerosols in the region. Likewise, errors in the seasonal cycle of low clouds in the southeastern Atlantic distort the radiative effects of biomass burning aerosols from southern Africa. These findings indicate that accurate assessment of aerosol direct effects requires models to correctly represent not only the source, strength, and optical properties of aerosols, but their relative proximity to clouds as well.

Radiative kernels describe the differential response of radiative fluxes to small perturbations in state variables and are widely used to quantify radiative feedbacks on the climate system. Radiative kernels have traditionally been generated using simulated data from a global climate model, typically sourced from the model's base climate. Consequently, these radiative kernels are subject to model bias from the climatological fields used to produce them. Here, we introduce the first observation‐based temperature, water vapor, and surface albedo radiative kernels, developed from CloudSat's fluxes and heating rates data set, 2B‐FLXHR‐LIDAR, which is supplemented with cloud information from the Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO). We compare the radiative kernels to a previously published set generated from the Geophysical Fluid Dynamics Laboratory (GFDL) model and find general agreement in magnitude and structure. However, several key differences illustrate the sensitivity of radiative kernels to the distribution of clouds. The radiative kernels are used to quantify top‐of‐atmosphere and surface cloud feedbacks in an ensemble of global climate models from the Climate Model Intercomparison Project Phase 5, showing that biases in the GFDL low clouds likely cause the GFDL kernel to underestimate longwave surface cloud feedback. Since the CloudSat kernels are free of model bias in the base state, they will be ideal for future analysis of radiative feedbacks and forcing in both models and observations and for evaluating biases in model‐derived radiative kernels.

Aerosol direct radiative effects are assessed using multi‐sensor observations from the A‐Train satellite constellation. By leveraging vertical cloud and aerosol information from CloudSat and CALIPSO, this study reports new global estimates of aerosol radiative effects and the component owing to anthropogenic aerosols. We estimate that the global mean aerosol direct radiative effect is −2.40 W/m2 with an error of ± 0.6 W/m2 owing to uncertainties in aerosol type classification and optical depth retrievals. Anthropogenic direct radiative forcing is assessed using new observation‐based aerosol radiative kernels. Anthropogenic aerosols are found to account for 21% of the global radiative effect, or −0.50 ± 0.3 W/m2, mainly from sulfate pollution (−0.54 W/m2) partially offset by absorption from smoke (0.03 W/m2). Uncertainty estimates effectively rule out the possibility that anthropogenic aerosols warm the planet, although strong positive forcing is observed locally where anthropogenic aerosols reside above clouds and bright surfaces.