S. R. Faruqi:

nā = an expression of insistence

The use of nā in the refrain is extremely fine. The kind of appropriateness that Mir shows in everyday usage has hardly been vouchsafed to anybody else. Then in the whole verse, how finely similes and images have been grappled together [dast-o-garebāñ huʾe haiñ]! A whirlwind rises up high; thus the speaker has called it a 'head-lifter' (that is, arrogant). Water flows along the ground; thus he's called it a 'head-burier'.

Then there's the enjoyable fact that in destruction and effectiveness the rank of a flood is considerably higher than that of a whirlwind. When a whirlwind passes through, then no trace of it remains behind; nor does a whirlwind have the breadth and prolongation that a flood does. After a flood passes through, its effects remain. The 'desert' can be taken to be the 'desert of life', and the 'desert of passion' as well.

Mir has expressed this theme a number of times. From the second divan [{906,8}]:

dekheñ pesh āve hai kyā ʿishq meñ ab to jūñ sail
ham bhī us rāh meñ sar gāṛe chale jāte haiñ

[let's see what comes before us in passion; now, like a flood
even/also we, with heads buried/sunk, go along in that road]

Also from the second divan [{1020,9}]:

dekh sailāb us bayābāñ kā
kaisā sar ko jhukāʾe jātā hai

[look at the flood of that desert--
how, with its head bowed, it goes along]

From the third divan:

{1206,9}.

[Compare {1327,5}.]

FWP:

SETS == MIDPOINTS; WORDPLAY
MOTIFS == DESERT
NAMES
TERMS == REFRAIN

The phrase 'like a flood' [jūñ sail] is perfectly positioned to be read either with sar gāṛe ('with the head buried like a flood') or with chalā jā ('go along like a flood').

Lowering or 'sinking' the head can be a sign of humility or sorrow, as contrasted to the 'arrogance' of raising the head up high. But it can also show absolute determination and resolve-- one 'plants' or 'embeds' the head (see the definition above). Only then, paradoxically, does the submerged or 'sunk' head permit the kind of movement that a flood can generate; and of course, floods in deserts often roar down dry channels and so are particularly potent. Such a flood (of tears?) can indeed do more damage than a whirlwind (of sighs?)

Note for grammar fans: SRF's text has mānind bagole ke , while the kulliyat has kī instead. Platts tries to pin down the subtleties; see the definition above. Apparently it's a case similar to that of the adverbial vs. adjectival participle forms, which also often end up used more flexibly than the official rules would dictate.