Characterizing and Understanding cloud-precipitation-radiation-dynamical interactions using CloudSat/CALIPSO and other A-Train observations

The most fundamental characteristics of Earth's climate lies in the manner that radiation balance is achieved and perturbed through various climate processes and feedbacks and external forcings. Recognition of this first-order principle is exhibited by the relatively early development, and continuation, of accurate satellite measurements of top of the atmosphere (TOA) radiation (e.g. SRB, ERBE, CERES) and their use to constrain and evaluate GCM representations of climate. While it is generally understood that GCMs seek, and for the most part tend to achieve global energy balance consistent with these observations and their uncertainties, it is also recognized that there remains considerable biases in radiation in space and time and associated with the relevant inner workings of the (modeled) climate system (e.g. circulation, clouds, water vapor). A key element, in global climate modeling framework, of obtaining an accurate top of the atmosphere (TOA) radiation budget is the representation of clouds, which for GCMs and earth radiation budget considerations can be roughly broken down into realistic vertical distributions of cloud cover, cloud water mass and cloud particle sizes for radiative heating calculations. In this study, we provide a robust evaluation and analysis of the cloud ice and liquid water content, TOA and surface radiation budgets of the CMIP5 20th century simulations, and compare their fidelity to those from CMIP3. To account for observational radiation uncertainty, we utilize a number of contemporary satellite measurements for the IWC/IWP (e.g., CloudSat+ CALIPSO), LWC/LWP (CloudSat; AMSR-E; MODIS), TOA and surface (e.g. CERES-EBAF; CloudSat) and observationally/satellite-constrained model calculations for the surface (e.g. CERES product, CloudSat product). Despite their obvious relevance and application to GCM cloud studies, the considerable care and caution required in order to make judicious comparisons between the GCM representations of (typically only the) clouds and the satellite observations that are an inherent combination of the clouds and falling hydrometeors (e.g., rain or snow) as well as convective core cloud mass clearly evident in the cloud water mass simulated in CMIP3/CMIP5. We evaluate the GCM fidelity in representing TOA outgoing longwave radiation (RLUT) and reflective shortwave radiation (RSUT), as well as surface values of net shortwave and longwave (RSDS). Based on examination of 17 CMIP5 and 12 CMIP3 GCMs over the period 1970-2005 and 1970-1999, respectively, we find that there is no significant improvement in the magnitude and spatial structure of the biases in either the TOA or the surface. In fact, while the multi-model global mean RLUT bias is about the same between CMIP3 and CMIP5, the size of the cancelling errors across the globe is bigger in CMIP5 than CMIP3. Persistent and systematic biases across most of the models and the model ensemble means are underestimated RSUT, overestimated RSDS and RLUT which are consistent to the underestimated IWP/LWP in the extra-tropics and storm tracks and significantly underestimated IWP+LWP in tropical convective active regions over ITCZ/SPCZ, Warm Pool, Indian Monsoon as well as South America, Central Africa. The systematic biases of the RSDS, RLUT and RSUT are in phase with the maximum precipitation regions in the tropics, suggesting that at least a part of this persistent bias stems from GCMs ignoring the effects of precipitating and/or convective core ice and liquid in their radiation calculations.