Yes, global models have high enough resolution
No, high resolution LAM models are necessary
No, the vertical extension is too small