15B.1 Water Vapor Measurement System using Digital Terrestrial Broadcasting Waves

Tuesday, 29 August 2017: 4:00 PM
St. Gallen (Swissotel Chicago)
Seiji Kawamura, National Institute of Information and Communications Technology, Tokyo, Japan; and H. Hanado, T. Kouketsu, H. Ohta, and T. Iguchi

Severe weather phenomena such as localized heavy rainstorms in urban areas are social issues these days. Their dimensions are small in time and space, and it is still difficult to predict when and where they occur. Water vapor is an essential parameter for weather forecast because it is the origin of raindrops. Recently, it becomes known that water vapor near ground surface has an impact for localized heavy rainstorms. However, water vapor is one of the most difficult physical quantities to measure by remote sensing. Main ground-based techniques to measure water vapor is GNSS receivers and microwave radiometer profilers. They can measure integrated water vapor or precipitable water vapor (PWV) in vertical direction. However there are few techniques to measure it horizontally.

We have developed a method to measure water vapor in horizontal direction using digital terrestrial broadcasting waves [Kawamura et al., 2017]. Radio waves are delayed due to water vapor through propagation. Water vapor can be estimated using propagation delay of digital terrestrial broadcasting waves. The basic idea of using propagation delay is the same as that of retrieving PWV by using GNSS. In this study, we estimate water vapor near a ground surface from the horizontal propagation delay of digital terrestrial broadcasting waves. The main features of this observation are, no need for transmitters (small and low cost due to receiving only), applicability wherever digital terrestrial broadcasting is available, and its high time resolution. We have already developed a real-time delay measurement system with a software-defined radio technique, and started to deploy it around Tokyo area. If many small receivers are deployed, 2-D water vapor variations can be monitored with high time and space resolutions. Our target is to improve the accuracy of numerical weather forecast for severe weather phenomena such as localized heavy rainstorms in urban areas through data assimilation. In this presentation, we will introduce a method to measure water vapor using digital terrestrial broadcasting waves, some observed results, and future plans. 

The followings are brief explanation how to measure water vapor using digital terrestrial broadcasting waves. Because the delay due to water vapor is quite small (picoseconds order), very precise measurements are needed for effective observations. Phase fluctuations of local oscillators at radio tower and receivers are essential error factors. We observe phase variations of digital terrestrial broadcasting waves at a single receiving site. If there is a reflector at the opposite side from the radio tower, we can receive direct and reflected waves at this point simultaneously using a single local oscillator. Phase noises of this local oscillator are cancelled out by taking the difference between direct and reflected waves. We can measure a round trip propagation delay between the observing point and the reflector.

There are several standards in digital terrestrial broadcasting all over the world. Each channel in almost all system has bandwidth of about 6 MHz, and we can measure complex delay profiles using embedded known signals. For examples, ISDB-T system is adopted in Japan, which uses Orthogonal Frequency Division Multiplexing (OFDM) for the modulation. In this system, 5617 carriers are used for a single channel within 5.6 MHz bandwidth. In each carrier, scattered pilots (SPs, known signals) are embedded every four symbols. A symbol is the base unit of OFDM modulation, whose length is 1.134 ms. Therefore, complex delay profiles are calculated every 4.536 ms using SPs. Using delay profiles, we can identify direct and reflected waves individually. Phase variations of direct and reflected waves are calculated at each peak of delay profile. We can derive the propagation delay by taking the difference between phase variations of direct and reflected waves.

Figure 1 shows an example of observed results. We receive digital terrestrial broadcasting waves at NICT (Koganei, Tokyo) which is about 29 km westward from the radio tower. One of reflectors (a building) is about 1 km westward from NICT. Observed propagation delay between NICT and the reflector is plotted in red line. We have ground-based meteorological observation at NICT, and we can also calculate propagation delay using these results (black line). The data obtained using digital terrestrial broadcasting waves show good agreement with those obtained by ground-based meteorological observation.

Reference

Kawamura, S., et al. (2017), Water vapor estimation using digital terrestrial broadcasting waves, Radio Sci., 52, 367–377, doi:10.1002/2016RS006191.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner