You might be able to use Option 1 but here are the couple of constraints.

Both the baseline and actual cases must include the month of July.The LEED Reference Guide states that to capture seasonal variations, the preferred period for irrigation water use analysis is at least one full year.Option 2, the theoretical calculation, is still available to teams that have sub-metered irrigation so this might be a good alternative if you only have partial data.