Cloud radiative kernels quantify the sensitivity of top-of-atmosphere (TOA) radiative fluxes to cloud fraction perturbations within seven cloud top pressure categories and seven optical depth categories (see figure on right). Multiplying the cloud radiative kernels, which are a function of latitude, month, and surface albedo, by changes in cloud fraction (segregated on the same cloud top pressure-optical depth grid) between two climate states yields a quantitative estimate of cloud-induced TOA radiation anomalies. Normalizing these by the change in global mean surface temperature between the two climate states provides a measure of cloud feedback.
Because the kernels are computed using a single radiative transfer code (Fu-Liou), differences in cloud feedbacks among climate models can be unambiguously attributable to inter-model differences in the responses of clouds. Furthermore, because the cloud feedback is computed directly from changes in cloud fields rather than inferred from TOA fluxes, no adjustments are necessary to account for non-cloud-induced radiative flux anomalies. A notable advantage of this technique is its ability to quantify the contribution of 49 different cloud types to the cloud feedback. For more details, see our papers documenting these results (below). Please feel free to email me with any questions.
There are two sets of cloud radiative kernels available for download:
1) The cloud radiative kernels that are appropriate for use with climate model output were developed using zonal mean temperature and humidity profiles averaged across control runs of six CFMIP1 climate models as input to the radiation code. Please refer to Zelinka et al. (2012a,b) for details. Download cloud_kernels2.nc
2) The cloud radiative kernels that are appropriate for use with observations were developed using zonal mean temperature, humidity, and ozone profiles from ERA Interim over the period 2000-2010 as input to the radiation code. Please refer to Zhou et al. (2013) for details.
Update (12/30/16): The previous version of these kernels (obs_cloud_kernels2.nc) had been computed with infrared cloud optical depths that were re-scaled from their visible counterparts. This resulted in LW kernel values that were erroneously small in many bins. This error has been fixed in the current version (obs_cloud_kernels3.nc). The SW kernel is unaffected. Download obs_cloud_kernels3.nc
Sample Python, Matlab, and NCL scripts that use kernels to compute cloud-induced radiation anomalies are available from my github repository. Feel free to email me with questions or for code to perform the amount/altitude/optical depth decomposition of Zelinka et al. (2013).
Zelinka, M. D., S. A. Klein, and D. L. Hartmann, 2012a: Computing and Partitioning Cloud Feedbacks Using Cloud Property Histograms. Part I: Cloud Radiative Kernels, J. Climate, 25, 3715–3735. doi:10.1175/JCLI-D-11-00248.1.
Zelinka, M. D., S. A. Klein, and D. L. Hartmann, 2012b: Computing and Partitioning Cloud Feedbacks Using Cloud Property Histograms. Part II: Attribution to Changes in Cloud Amount, Altitude, and Optical Depth, J. Climate, 25, 3736–3754. doi:10.1175/JCLI-D-11-00249.1.
Zhou, C., M. D. Zelinka, A. E. Dessler, P. Yang, 2013: An analysis of the short-term cloud feedback using MODIS data, J. Climate, 26, 4803–4815. doi: 10.1175/JCLI-D-12-00547.1.
Zelinka, M. D., S. A. Klein, K. E. Taylor, T. Andrews, M. J. Webb, J. M. Gregory, and P. M. Forster, 2013: Contributions of Different Cloud Types to Feedbacks and Rapid Adjustments in CMIP5, J. Climate., 26, 5007–5027. doi: 10.1175/JCLI-D-12-00555.1.