A Vehicle-Based Observation System for Collecting Near-Ground Meteorological Big Data: A Case Study in the Tokyo Metropolitan Area

Abstract

Extreme and intense atmospheric phenomena occur in complex urban areas, but stationary ground-based meteorological observations cannot capture the detailed meteorological conditions across urban blocks. Vehicle-based mobile observation provides a method to collect high spatiotemporal data in cities. We developed a small IoT observation system to mount vehicles easily and measure air temperature, humidity, pressure, wind speed and direction, together with local time and location. To evaluate this system, we conducted a case study of mobile observation in Tokyo almost daily for one and a half months. Mobile observation data had occasional missing and erroneous, so data extraction and processing were applied based on weather conditions and vehicle's speed. Using the quality-controlled data, we confirmed that given the mobile observation was within 2 km of the fixed-point observation, air temperature and humidity from mobile observation was highly correlated with reliable fixed-point observation (RMSD < 1°C and < 0.34 g/kg). Thus, the mobile observation system potentially provides those datasets comparable to conventional ground meteorological datasets. The spatial distributions of air temperature and humidity exhibited distinctive changes between urban blocks, influenced by land use and urban characteristics. This system enables long-term, extensive vehicle-based observations and aids in detecting extreme atmospheric phenomena in urban areas.