Barrier layers are defined as the layer between the pycnocline and the thermocline when the latter are different as a result of salinity stratification. We present a revisited 2-degree resolution global climatology of monthly mean oceanic Barrier Layer (BL) thickness first proposed by de Boyer Montegut et al. (2007). In addition to using an extended data set, we present a modified computation method that addresses the observed porosity of BLs. We name porosity the fact that barrier layers distribution can, in some areas, be very uneven regarding the space and time scales that are considered. This implies an intermittent alteration of air-sea exchanges by the BL. Therefore, it may have important consequences for the climatic impact of BLs. Differences between the two computation methods are small for robust BLs that are formed by large-scale processes. However, the former approach can significantly underestimate the thickness of short and/or localized barrier layers. This is especially the case for barrier layers formed by mesoscale mechanisms (under the intertropical convergence zone for example and along western boundary currents) and equatorward of the sea surface salinity subtropical maxima. Complete characterisation of regional BL dynamics therefore requires a description of the robustness of BL distribution to assess the overall impact of BLs on the process of heat exchange between the ocean interior and the atmosphere.