Abstract |
The phased array weather radar (PAWR) was developed to detect a warning sign of torrential rainfall and to predict weather hazard using the three-dimensional dense precipitation data every 30 seconds. The X-band PAWR measures 100 m range resolution data over the observation range of 60 km radius in about 100 elevation angles from 0 to 90 degrees. The data generation rate of this big data of 100 m resolution every 30 seconds is about 100 times that of conventional weather radar equipped with a parabolic antenna. The observation big data will be expected to reveal new features of the rapidly developing severe storm. However, the PAWR data has some limitations such as detecting large droplets only, and including clutter noise, or various errors. The data assimilation (DA) technique is useful to use the incomplete observation big data for predicting heavy rainfall. The DA can bridge observation data including some errors to numerical prediction models. However, unnecessary data such as clutter echoes included in the observation data deteriorates prediction accuracy, so it is important to remove it beforehand. The data quality control must be processed in real time for actual utilization of observation data.
|