Fog is formed when water vapor in the atmosphere condenses into miniscule floating water droplets, rendering a visibility of less than 1 km. Fog causes economic damage and poses a hazard to people owing to limited visibility. Fog forecasts depend on the judgment and experience of the forecaster, who bases the forecast on analysis results from meteorological models and satellite images. However, such forecasts are complicated by spatiotemporal changes and fine physical processes. To address this issue, a variety of fog studies have used the Global Navigation Satellite System (GNSS) tropospheric signal delay as an alternative. In this study, we monitored fog and cloudiness by analyzing precipitable water vapor (PWV) extracted from GNSS tropospheric delay. The analysis of the change in PWV during fog generation showed that it is difficult to detect fog from PWV using GNSS; more precise GNSS processing is required. However, it was found that cloudiness affects the change in PWV. In particular, the change in PWV is affected by high-level and middle-level clouds rather than low-level clouds. In other words, we conducted research to detect fog and cloudiness, but it was difficult to monitor them using PWV calculated from GNSS, although the use of GNSS to monitor the high-level and middle-level clouds generated good results. The results of this study will aid in quantitative cloudiness measurements, which at present are mostly performed through non-quantitative visual inspection. Furthermore, these results are expected to be helpful in the development of automatic observation devices for cloudiness measurements, satellite image processing, and weather forecasting.