Spatially resolved HI studies of dwarf galaxies have provided a wealth of precision data. However these high-quality, resolved observations are only possible for handful of dwarf galaxies in the Local Volume. Future HI surveys are unlikely to improve the current situation. We therefore explore a method for estimating the surface density of the atomic gas from global HI parameters that are conversely widely available. We perform empirical tests using galaxies with resolved HI maps, and find that our approximation produces values for the surface density of atomic hydrogen within typically 0.5dex of the true value. We apply this method to a sample of 147 galaxies drawn from modern near-infrared stellar photometric surveys. With this sample we confirm a strict correlation between the atomic gas surface density and the star formation rate surface density, which is vertically offset from the Kennicutt-Schmidt relation by a factor of 10-30, and significantly steeper than the classical N=1.4 of Kennicutt (1998ARA&A..36..189K). We further infer the molecular fraction in the sample of low surface brightness, predominantly dwarf galaxies by assuming that the star formation relationship with molecular gas observed for spiral galaxies also holds in these galaxies, finding a molecular-to-atomic gas mass fraction within the range of 5-15 per cent. Comparison of the data to available models shows that a model in which the thermal pressure balances the vertical gravitational field captures better the shape of the {Sigma}SFR-{Sigma}gas relationship. However, such models fail to reproduce the data completely, suggesting that thermal pressure plays an important role in the discs of dwarf galaxies.

Cone search capability for table J/MNRAS/480/210/table1 (Externally obtained properties required to compute the surface densities for the KRDJ08 and YJLK14 NIR samples.)