I have not worked with the model, but in principle it should be possible. If there is any routine that is restricted to run only if the time step is smaller than a certain criterion, then you can simply use the same value for your forcing during each 24 hour period. It is not elegant but practical. Be aware though of the effects: averaging and flattening of peaks for time scales below ( and maybe also above) your time step. If that is important depends on the time and spatial scale of your research.