We use narrowband imaging ( FWHM = 70 Å ) to select a sample of emission line galaxies between 0.20 \lesssim z \lesssim 1.22 in two fields covering 0.5 sq .
deg .
We use spectroscopic follow-up to select a sub-sample of H \alpha emitting galaxies at z \sim 0.24 and determine the H \alpha luminosity function and star formation density at z \sim 0.24 for both of our fields .
Corrections are made for imaging and spectroscopic incompleteness , extinction and interloper contamination on the basis of the spectroscopic data .
When compared to each other , we find the field samples differ by \Delta \alpha = 0.2 in faint end slope and \Delta \log [ L ^ { * } ( \mathrm { { \mathrm { erg s ^ { -1 } } } } ) ] = 0.2 in luminosity .
In the context of other recent surveys , our sample has comparable faint end slope , but a fainter L ^ { * } turn-over .
We conclude that systematic uncertainties and differences in selection criteria remain the dominant sources of uncertainty between H \alpha luminosity functions at this redshift .

We also investigate average star formation rates as a function of local environment and find typical values consistent with the field densities that we probe , in agreement with previous results .
However , we find tentative evidence for an increase in star formation rate with respect to the local density of star forming galaxies , consistent with the scenario that galaxy-galaxy interactions are triggers for bursts of star formation .