Python, matlab, and ncl scripts are provided that demonstrate how to use cloud radiative kernels to compute cloud-induced radiation anomalies and cloud feedback. In addition, a python script is provided that demonstrates how to partition the cloud feedbacks in to components due to changes in cloud amount, altitude, optical depth, and a residual, for all clouds, non-low clouds, and low clouds.
Zelinka, M. D., S. A. Klein, and D. L. Hartmann, 2012: 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, 2012: 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.
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.
Zelinka, M. D., C. Zhou, and S. A. Klein, 2016: Insights from a Refined Decomposition of Cloud Feedbacks, Geophys. Res. Lett., 43, 9259-9269, doi:10.1002/2016GL069917.
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.
The code makes use of the following data:
| Frequency | Name | Description | Unit | File Format |
|---|---|---|---|---|
| monthly mean | clisccp | ISCCP simulator cloud fraction histograms | % | nc |
| monthly mean | rsuscs | upwelling SW flux at the surface under clear skies | W/m^2 | nc |
| monthly mean | rsdscs | downwelling SW flux at the surface under clear skies | W/m^2 | nc |
| monthly mean | tas | surface air temperature | K | nc |
| monthly mean | LWkernel | LW cloud radiative kernel | W/m^2/% | nc |
| monthly mean | SWkernel | SW cloud radiative kernel | W/m^2/% | nc |
Two sets of cloud radiative kernels available at https://github.com/mzelinka/cloud-radiative-kernels/tree/master/data
-
cloud_kernels2.nc: The cloud radiative kernels developed using zonal mean temperature and humidity profiles averaged across control runs of six CFMIP1 climate models as input to the radiation code. These are best for diagnosing feedbacks relative to a modeled pre-industrial climate state. Please refer to Zelinka et al. (2012a,b) for details.
-
obs_cloud_kernels3.nc: The cloud radiative kernels developed using zonal mean temperature, humidity, and ozone profiles from ERA Interim over the period 2000-2010 as input to the radiation code. These are best for diagnosing feedbacks relative to an observed present-day climate state. Please refer to Zhou et al. (2013) for details.
SW and LW cloud feedbacks.
Each feedback is size (MO,TAU,CTP,LAT,LON)=(12,7,7,90,144)
For the provided sample imput data, the code should print the following output, which is the global annual mean LW and SW cloud feedbacks. The values are slightly different in the Matlab and Python versions, possibly due to differences in regridding and in area-weighted averaging.
| Average Cloud Feedback Component | Matlab | Python |
|---|---|---|
| LW | 0.816 | 0.833 |
| SW | 0.414 | 0.402 |
Figures generated by the script are located in the images directory.