If you think the snow (icing) will stay on the pipe during the seismic activity and active for the seismic forces, you may add the mass effect of the snow by changing the pipe material density to the new calculated (pipe + snow) density. This is a conversion of the total mass to the pipe geometry which will result with higher density than pipe density. And do load combinations as usual.
If you are concern on the vertical gravity effect only you can model the snow effect with uniform force only and combined with seismic in the load combinations.