Real-time and clock-shared rainfall monitoring with a wireless sensor network

Abstract

Rainfall is recognized as one of the most important factors in hydrology. Particularly, investigation of the tempo-spatial variation patterns of rainfall and their effect on environments has attracted more and more attention in recent. Because of the limitations in both human costs and existing rainfall monitoring devices, however, it is very hard for researchers to collect real-time rainfall data from large-scale geographical areas. This paper designs and implements RealRain, the first real-time wireless system of monitoring rainfall, which explicitly serves as an effective and efficient scientific instrument for domain experts to facilitate the measurement of large-scale and real-time rainfall. With the tipping bucket as the rainfall signal detecting device, RealRain employs commercial embedded wireless platforms to record and collect rainfall data automatically. Especially, RealRain involves a tree-based time synchronization scheme. For practical account, we also propose a sample-send-store scheme to improve the data reliability of RealRain under potentially harsh environments. With RealRain, end-users can remotely retrieve the real-time and fine-grained rainfall data under a unified timeline. Finally, we implement a proof-of-concept prototype of RealRain, which is completely driven by rainfall events and can automatically feed the data to users. The preliminary laboratory results and analyses demonstrate the feasibility and the efficacy of RealRain.