Abstract
Rainfall measured by Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar (PR) is important for studying precipitation distribution in the tropical regions. The ground validation of TRMM PR is difficult because the ground sensing systems have different characteristics from TRMM PR in terms of resolution, scale, view aspect and sensing environments. In this paper, we introduce a machine learning system to train ground radars for rainfall estimation using rain gauge data and subsequently using the trained ground radar rainfall estimation to train TRMM PR. This system can build a connection between ground gauge measurements and ground radar observations, and transfer this connection to TRMM PR observations for rainfall estimation. The rain gauge, ground radar and satellite data collected from Melbourne, Florida are used for demonstration purposes. The rainfall estimation product derived from this new system is compared against the TRMM standard products, which shows improvement brought by the new machine learning system.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call