Development of low-cost meteorological observation system based on wireless network for poor-visibility occurred by snowstorm

Hyper-densely monitoring for poor-visibility occurred by snowstorm is needed to make an alert system, because the snowstorm is difficult to predict from the observation only at a representative point. There are some problems in the previous approaches for the poor-visibility monitoring using video analyses or visibility meters; these require a wired network monitoring (a large amount of data: 10MB/sec at least) and the system cost is high (10,000 at each point). Thus, the risk of poor-visibility has been mainly measured at specific point such as airport and mountain pass, and estimated by simulation two dimensionally. To predict it two dimensionally and accurately, we have developed a low-cost meteorological system to observe the snowstorm hyper-densely. We have developed a low-cost visibility meter which works as the reduced intensity of semiconducting laser light when snow particles block off. Our developed system also has a capability of extending a hyper-densely observation in real-time on wireless network using Zigbee; A/D conversion and wireless data sent from temperature and illuminance sensors. We use a semiconducting laser chip (5) for the light source and a reflection mechanism by the use of three mirrors so as to send the light to a non-sensitive illuminance sensor directly. Thus, our visibility detecting system ($500) becomes much cheaper than previous one. We have checked the correlation between the reduced intensity taken by our system and the visibility recorded by conventional video camera. The value for the correlation coefficient was -0.67, which indicates a strong correlation. It means that our developed system is practical. In conclusion, we have developed low-cost meteorological detecting system to observe poor-visibility occurred by snowstorm, having a potential of hyper-densely monitoring on wireless network, and have made sure the practicability.