README

Data

The data folder contains multiple NetCDF (.nc) files used for analysis. These files can be categorized as follows:

Cloud Radiative Kernel (CRK) Files

• gcrk.nc: GEOS RRTMG Cloud Radiative Kernel – computed directly from the GEOS model.
• fcrk.nc: FBCT Cloud Radiative Kernel – derived from the CERES FluxByCldTyp (FBCT) product (Sun et al., 2022).
• zcrk.nc: Zhou et al. (2013) Cloud Radiative Kernel – downloaded from https://github.com/mzelinka/cloud-radiative-kernels/tree/master/data

Each of these CRK files has a shape of (181, 6, 7, 12), where:

• 181 represents latitude (longitudinally averaged).
• 6 corresponds to TAU bins (optical thickness).
• 7 corresponds to CTP bins (cloud top pressure).
• (6×7) follows the ISCCP discretization.
• 12 represents months, starting from May.

GEOS Model Results

Files following the pattern ??jh?.nc are outputs from the GEOS model:

• The suffix p represents the present climate.
• The suffix f represents the future climate scenario (SST+4K).

These files contain arrays of shape (360, 181, 6, 7, 12), where:

• 360 represents longitude.
• The remaining dimensions are the same as those in the CRK files.

The prefix in these filenames indicates the dataset used:

• rg – Räisänen GEOS
• ri – Räisänen ISCCP
• si – SCOPS ISCCP
• rm – Räisänen MODIS
• sm – SCOPS MODIS

Python Scripts

The Python scripts are used for visualization and analysis:

• cf_decomp.py: Implements the Zelinka et al. (2013) decomposition method, breaking down cloud radiative effects into Total, Amount, Altitude, Optical Thickness, and Residual components.
• fig2.py: Generates Figure 2.
• fig3a.py: Generates Figure 3a.
• fig3b.py: Generates Figure 3b.
• figsm1.py: Generates Supplementary Figure 1.
• figsm2.py: Generates Supplementary Figure 2.
• figsm3.py: Generates Supplementary Figure 3.

This research was supported by the use of Python codes developed with assistance from ChatGPT, an AI tool provided by OpenAI.