Systematic Search for FFPs in KMTNet Full-Frame Images. I. Photometry Pipeline

To exhume the buried signatures of free-floating planets (FFPs) with small angular Einstein radius $\theta_{\rm E}$, we build a new full-frame difference image pipeline for the Korean Microlensing Telescope Network (KMTNet) survey based on the newly optimized pySIS package. We introduce the detailed processes of the new pipeline, including frame registration, difference image analysis, and light curve extraction. To test this pipeline, we extract 1-year light curves for 483,068 stars with $I \lesssim 17$ and conduct a model-independent search for microlensing events. The search finds 36 microlensing events, including five new events and six events discovered by other collaborations but missed by previous KMTNet searches. We find that the light curves from the new pipeline are precise enough to be sensitive to FFPs with $\theta_{\rm E} \sim 1~\mu$as. Using the new pipeline, a complete FFP search on the eight-year KMTNet images can be finished within six months and then yield the FFP mass function. The new pipeline can be used for a new KMTNet AlertFinder system, with significantly reduced false positives.