Prediction of daily heavy precipitation in a lowland area using GPS and surface weather data

Recent progress in GPS technology has enabled us to estimate the total amount of water vapor in an atmospheric column with infinite height as 'GPS-derived precipitable water vapor' (GPS-PWV). In the present study, we analyzed the GPS and meteorological data obtained in Saga Plain, in northern Kyushu, Japan. An attempt was made to predict heavy precipitation (≥50 mm d^-1) on a daily basis using the analyzed relations between surface temperature and GPS-PWV. Several features were revealed that were associated with the daily heavy precipitation at the study area. An index was made to predict heavy precipitation on a daily basis. Applying the index to observations, we obtained acceptable results of the prediction. It is suggested that GPS is useful for predicting heavy precipitation events.