Kilonova and Optical Afterglow from Binary Neutron Star Mergers. I. Luminosity Function and Color Evolution

In the first work of this series, we adopt a GW170817-like, viewing-angle-dependent kilonova model and the standard afterglow model with the lightcurve distribution based on the properties of cosmological short gamma-ray bursts afterglows to simulate the luminosity functions and color evolution of both kilonovae and optical afterglow emissions from binary neutron star (BNS) mergers. {We find that $\sim10\%$ afterglows are brighter than the associated kilonovae at the peak time, most of which are on-axis or nearly-on-axis. These kilonovae would be significantly polluted by the associated afterglow emission. Only at large viewing angles with $\sin\theta_{\rm v}\gtrsim 0.20$, the EM signals of most BNS mergers would be kilonova-dominated and some off-axis afterglows may emerge at $\sim5-10$\,day after the mergers.} At brightness dimmer than $\sim23-24$\,mag, according to their luminosity functions, the number of afterglows is much larger than that of kilonovae. Since the search depth of the present survey projects is $<22$\,mag, the number of afterglow events detected via serendipitous observations would be much higher than that of kilonova events, consistent with the current observations. {For the foreseeable survey projects (e.g., Mephisto, WFST, LSST) whose search depths can reach $\gtrsim23-24$\,mag, the detection rate of kilonovae could have the same order of magnitude as that of afterglows.} We also find that it may be difficult to use the fading rate in a single band to directly identify kilonovae and afterglows among various fast-evolving transients by serendipitous surveys. However, the color evolution between optical and infrared bands can identify them, since their color evolution patterns are unique compared with those of other fast-evolving transients.