In this study, a hybrid-type of day fog detection algorithm was developed based on the optical and textural characteristics of fog top, using the recent geostationary satellite, Himawari satellite/AHI(Advanced Himawari Imager) data.. Supplementary data, such as air temperatures of numerical weather prediction model(LDAPS) on land and simulated clear sky radiance on sea, were used for fog detection. Also, 10 minute data from a visibility meter from the Korea Meteorological Administration were employed for a quantitative verification of the fog detection results. Normalized albedo, an optical property of fog, was utilized to distinguish between fog and other objects such as clouds, land, and oceans. The normalized local standard deviation that represents spatial variability of the fog surface and temperature difference between fog top and air temperature were also assessed to separate the fog from low cloud. Initial threshold values for the fog detection elements were set using hat-shaped threshold values through frequency distribution analysis of fog cases. And the initial threshold values were optimized through the iteration method by minutely changing them using threat score(TS). The validation results showed that the fog detection level was the highest in cases where the fog was strong and prevail. However, the fog detection level differs from case to season. The average POD and FAR for the evaluation cases were 0.72 and 0.70, respectively. Because the FAR is discouragingly high, we analyzed the availability of other channels of Himawari-8/AHI and seasonal threshold values. Among the various sensitivity tests, the BTD(11 um ? 12 um) and BTD(11 um ? 13 um) showed a possibility of reduction of FAR. In addition, sophistication of threshold values as a function of season clearly improved the fog detection level. However, the fog detection level is sensitive depending on the intensity and type of fog. Therefore, to improve fog detection level and stability, it is necessary to correct the systematic errors of LDAPS data and to elaborate the threshold values using more than one year''s case.