conceptually once your leading GF has fallen from 100% to 99% you aren't supposed to be offgassing anymore.
100% in the context of GF low means you're riding the m-values on at least one compartment all the way up.
Yeah, what victorzamora said. I think you may be confusing 100% or GF100 with the point where you
start off-gassing; GF100 is experienced where a tissue is
so supersaturated compared to the instantaneous ambient pressure that it has hit 100% of the maximum 'allowed' in the Buhlmann model at that point in time, or 100% of the M-value. Buhlmann says you cannot safely off-gas
faster than this.
You still off-gas below GF100 by definition so long as you have a gradient between tissues and inspired partial pressure. (You can can even be off-gassing below GF0 as the gradient factor is calculated from ambient pressure not inspired gas inert partial pressure - see below)
GF Low, in the context of how it is programmed in to AFAIK every desktop Buhlmann implementation, influences only where you place the
first stop, not all the way up. The other stops are determined by stepping through the range between GFLo and GFHi depending on how many stops you are going to end up doing between the first stop and surface. Riding the M-values - i.e. maxing out permissible supersaturation at every stop, and hitting GF100 at the start of every stop - would be GF100/100.
The gradient factor you experience during each point of an ascent varies with changing depth, but also with off-gassing during each stop; as you off gas, you get further from the maximum permissible supersaturation at that point in time, so the 'gradient factor' drops until you reduce ambient pressure again by moving up to the next stop.
Dunno if this might help visualise it - this is a representation of the in-the-moment gradient factor experienced over time for an example GF 20/85 ascent, where the first stop, driven by the leading tissue exceeding 20% of permissible supersaturation, occurs at 45m, and ascent to surface is allowed when the leading tissue won't exceed 85% of permissible supersaturation:
Conversely this is the in-the-moment gradient factor experienced over time for our example GF 100/70 ascent after the same depth and bottom time: First stop is triggered when the leading tissue gets to 100% of permissible supersaturation, and final ascent to surface is allowed when the leading tissue at that point won't exceed 70% of permissible supersaturation:
Note that these charts only show the gradient factor (i.e. the % of permissible supersaturation) for the single leading tissue at that point in time. The other tissues all have their own gradient factors at each point in time but these aren't shown because the other tissues are all lower. Early in the ascent, the leading tissue is predominately a faster tissue, and later in the ascent it'll be predominately a medium/slow tissue.
(And obvious disclaimer, all this just explaining how the Buhlmann model works, we shouldn't confuse all this talk of tissues with anything that is actually happening in the body, just that this model seems to be a way of representing an idea of what may be going on in a way that seems to be useful in terms of not actually getting bent, mostly...)
(Edit: And, oh, by the way - in case anyone questions how these charts show the 6m stops going "below GF0" - what this means is the tissues are no longer supersaturated compared to ambient pressure (= 1.6ATA). The example diver is still 'off gassing' even though GF is less than zero because they are breathing oxygen in these examples and therefore the offgassing gradient between tissue and inspired gas is steeper than the gradient between tissues and ambient pressure.)