Inventory initialization patch numbering method is inconsistent with set_patchno

[glemieux]

While working on #911, I noticed that the initialize_sites_bv_inventory procedure sets patchno in the reverse order compared to set_patchno procedure in that it starts from the youngest patch and iterates through to the oldest:

fates/main/FatesInventoryInitMod.F90

Lines 491 to 507 in e216ca0

	ipa=1
	total_cohorts = 0
	currentpatch => sites(s)%youngest_patch
	do while(associated(currentpatch))
	currentpatch%patchno = ipa
	ipa=ipa+1
	! Perform Cohort Fusion
	call fuse_cohorts(sites(s), currentpatch,bc_in(s))
	call sort_cohorts(currentpatch)
	! This calculates %countcohorts
	call count_cohorts(currentpatch)
	total_cohorts = total_cohorts + currentPatch%countcohorts
	currentPatch => currentpatch%older
	enddo

Here is set_patchno for comparison:

fates/biogeochem/EDPatchDynamicsMod.F90

Lines 1249 to 1255 in e216ca0

	patchno = 1
	currentPatch => currentSite%oldest_patch
	do while(associated(currentPatch))
	currentPatch%patchno = patchno
	patchno = patchno + 1
	currentPatch => currentPatch%younger
	enddo

This works out to be consistent with everything else in the end given that we call set_patchno after the call to initialize_sites_bv_inventory within init_patches. Additionally, while fuse_patch (called a few lines down) utilizes patchno to check that the patches being fused are different, as long as each patch has a unique number, the order doesn't matter.

That said, we probably should replace the patch number update with a call set_patchno.