Evaluation of cloud properties in RCA4 CORDEX Africa ensemble

In light of recent advances in climate modelling and downscaling approaches, in the present study we have performed an evaluation of simulations by the regional climate model RCA4 CORDEX Africa region for its ability to recreate cloud fraction, cloud liquid and ice water paths. The RCA4 simulations (driven with 10 different GCMs) are evaluated against different satellite based observations in historical period (1982-2012) over African region. Further, cloud data sets (both satellite products and RCA4 simulations) are evaluated against their ability to represent natural variability (produced by various tele-connections). This would help investigate if climate models can also, to a first order, reproduce these robust features of natural variability. We have used the recently processed cloud property data records from CM-SAF, PATMOSx and ESA-CCI; as satellite observations. Firstly, RCA4 simulations are evaluated against the gridded observational datasets in the area using cloud fraction, cloud liquid and ice water paths as indicative variables. Spatial and temporal changes for the region are then calculated against observations to evaluate RCA4 for said variable. Thereafter, the observations and simulations are tested for their ability to reproduce natural variability, with focus on 3 tele-connections, NINA3.4, NINA3 and IOD indices. Singular Value decomposition is applied over the spatially averaged study region dataset to evaluate the effect of these oscillations.