Anthropogenic emissions of chlorofluorocarbons (CFCs) and other chlorinated species have greatly enhanced stratospheric halogen abundance over the second half of the twentieth century (World Meteorological Organization, 2018), and thereby led to substantial depletion of the ozone layer. The widespread, though indirect, climatic impacts of ODS via ozone depletion, largely linked to the the shift of the Southern Hemispheric circulation, are well documented (Polvani et al., 2011;Previdi & Polvani, 2014;Thompson & Solomon, 2002). But ODS also impact the climate system in ways that are not mediated by ozone, since they are potent well-mixed greenhouse gases (Ramanathan, 1975). Recent studies have shown that the surface warming from ODS may have played a key-role in the observed weakening trends of the Walker circulation (Polvani & Bellomo, 2019), and in the warming of the Arctic and the associated sea ice loss (Polvani et al., 2020). These studies suggest that the direct climate impacts of ODS, as greenhouse gases, may deserve closer study. In this paper, as a first step, we seek a deeper understanding of the purely radiative effects of ODS.Radiative forcing (RF) provides an important metric to evaluate and contrast the climatic impacts of different atmospheric trace gases (Ramaswamy et al., 2019). It is well established that ODS are powerful greenhouse gases (Ramanathan, 1975), with global warming potentials over a 100-year time horizon (GWP100) thousands of times larger than CO 2 (Hodnebrog, Aamaas, et al., 2020). However, the RF of ODS remains subject to some uncertainty. Detailed spectral calculations of the ODS RF using Line-by-Line (LbL) codes, routinely included in IPCC reports (e.g., Hodnebrog et al., 2013), typically employ two single (1-D) atmospheric profiles (tropical/ extratropical) as this is the best compromise between computational expediency and accuracy (errors are less than 1% compared to calculations with better spatial resolution). Moreover, ODS in these models are typically assumed to be uniformly distributed, and assumptions concerning their lifetime are made to try to capture spatial