Global grids of geophysical data or properties of the Earth are increasingly being used, perhaps because they are easy to access and make calculations on. Where original data are proprietary, researchers can sometimes obtain a form of those data within these grids, albeit at commonly lower resolution. Examples include gravity fields (Balmino et al., 2011; Sandwell et al., 2014), thicknesses of sediment overlying basement (Whittaker et al., 2013; Straume et al., 2019), Earth's topography (Smith and Sandwell, 1997; Becker et al., 2009; Ryan et al., 2009; Weatherall et al., 2015), crustal thickness (Laske et al., 2013), seafloor spreading history (Müller et al., 1997) and magnetic fields (Maus et al., 2009). However, depending on the type of study, when working with such data, we need to know the distribution of the original measurements that were used to compile those grids, hence which grid nodes contain interpolated values and which nodes are close to measurement sites. Ideally, we should be aware of the noise characteristics and resolution of such data also. The term “grid resolution” is commonly used in the literature but is misleading, as it refers to the spacing of nodes in the grid, not the resolution of the contributing data. Resolution can also be affected by processing steps used to create the grid. The article of Saada et al. (2021) describes important calculations and useful observations, though some parts of the study illustrate issues that can be encountered in analysing global grids. Given the increasing use of global grids, we provide this comment to help engender discussion.