Satellite and aerial imagery have been used extensively for mapping the abundance and distribution of giant kelp (Macrocystis pyrifera) in southern California. While tides and currents have been shown to affect the amount of floating kelp canopy on the water surface, there have been no quantifications of how these processes can bias remotely sensed kelp estimates in this region. We used unmanned aerial vehicles (UAVs) to map fine-scale changes in canopy area due to tidal height and current speed at both Palos Verdes, CA and Santa Barbara, CA. Additionally, we collected a biweekly time series of kelp canopy area in Palos Verdes over the course of a year to monitor fine-scale, intra-seasonal changes in canopy coverage. Our automated method for detecting kelp canopy in color and multispectral UAV imagery was highly accurate (over 84% and 98%, respectively) in classifying exemplary kelp and water pixels across a range of weather, ocean, and illumination conditions. Increases in tidal height of 1 m reduced the amount of floating kelp canopy by 15% to 32%. Current speeds are generally low in southern California and had no statistically significant effect on apparent bed size. However, increases in current speeds of 0.1 m/s reduced the amount of floating kelp canopy by over 31%. Tidal height and current speed can introduce significant variability to estimates of kelp abundance, but the magnitude of this variability is region specific. The biweekly time series displayed intra-annual variability typical of giant kelp, with a relatively gradual decline in kelp canopy in late summer corresponding with high sea surface temperatures, a rapid decline in the winter associated with wave disturbance, and a gradual recovery in the spring of the following year. The time series was also able to capture intra-seasonal changes in kelp canopy area that would have gone undetected in an annual or quarterly dataset, namely an increase in area to about half of the maximum in the late fall before wave events began.