This is a bit strange question. If you take observations from one particular time point, you have cross-sectional data. You could also calculate average cross time for each observation. It depends what you want to study. However, in general, it is not good idea to transform panel data to cross-sectional data, as panel data contains more information.