Real-time 30-second-refresh numerical weather prediction (NWP) was performed with exclusive use of 11,580 nodes (~7%) of supercomputer Fugaku during Tokyo Olympics and Paralympics in 2021. Total 75,248 forecasts were disseminated in the 1-month period mostly stably with time-to-solution less than 3 minutes for 30-minute forecast. Japan's Big Data Assimilation (BDA) project developed the novel NWP system for precise prediction of hazardous rains toward solving the global climate crisis. Compared with typical 1-hour-refresh systems, the BDA system offered two orders of magnitude increase in problem size and revealed the effectiveness of 30-second refresh for highly nonlinear, rapidly evolving convective rains. To achieve the required time-to-solution for real-time 30-second refresh with high accuracy, the core BDA software incorporated single precision and enhanced parallel I/O with properly selected configurations of 1000 ensemble members and 500-m-mesh weather model. The massively parallel, I/O intensive real-time BDA computation demonstrated a promising future direction.

A phased-array weather radar (PAWR) with fast scanning and wide elevation coverage capabilities has been developed. The PAWR transmits elevationally broad beams by feeding power to a limited number of its antenna elements, and receives signals reflected by precipitation media using all 128 antenna elements, each of which is connected to an analog-to-digital converter (ADC). After ADC sampling, digital processing of beamforming is applied to the received signals to accomplish receptions simultaneously over multiple angles. The PAWR can, thereby, make observations at 100-m range increments out to 60 km at 1° azimuth angle intervals over a range of 360° and at about 1° elevation-angle intervals from 0°-90° within the short time of 30 s. Reflectivity factors measured by the PAWR were compared with those from a collocated C-band radar, and were found to have a bias of 0.53 dB and a standard deviation of 3.68 dB, which indicates sufficient accuracy for observing precipitation. Furthermore, in a sample observation of convective rain, the PAWR detected a precipitation core 9 min before it reached the ground by its 30-s fast scanning and wide elevation coverage. It, thereby, became evident that the fast scanning and wide elevation coverage capabilities of the PAWR give earlier detection and a higher detection probability, which will enable earlier and more accurate warnings of severe weather phenomena and a better potential of mitigating the damage.

Following the invention of the telegraph, electronic computer, and remote sensing, “big data” is bringing another revolution to weather prediction. As sensor and computer technologies advance, orders of magnitude bigger data are produced by new sensors and high-precision computer simulation or “big simulation.” Data assimilation (DA) is a key to numerical weather prediction (NWP) by integrating the real-world sensor data into simulation. However, the current DA and NWP systems are not designed to handle the “big data” from next-generation sensors and big simulation. Therefore, we propose “big data assimilation” (BDA) innovation to fully utilize the big data. Since October 2013, the Japan's BDA project has been exploring revolutionary NWP at 100-m mesh refreshed every 30 s, orders of magnitude finer and faster than the current typical NWP systems, by taking advantage of the fortunate combination of next-generation technologies: the 10-petaflops K computer, phased array weather radar, and geostationary satellite Himawari-8. So far, a BDA prototype system was developed and tested with real-world retrospective local rainstorm cases. This paper summarizes the activities and progress of the BDA project, and concludes with perspectives toward the post-petascale supercomputing era.

We apply a C-band multi-parameter weather radar in the Okinawa Island to non-weather measurements. Its full-polarimetric and Doppler velocity measurement capability would be very powerful tool for observing land and ocean surface properties, if the target area is located appropriately to the radar. Located at the top of a mountain in the middle of the island, the radar has good view of its surrounding ocean. We make the radar observe ocean surfaces, in different wind conditions. The results show that the radar cross-section is larger when the wind speed is larger, and relatively smaller in the downwind side of the island, as expected. Cross-polarization components HV show similar dependency on wind conditions to HH and VV components. Measured Doppler velocities also show azimuth variation depending on the wind speed and direction.

National Institute of Information and Communication Technology (NICT, formally CRI) has developed a new C-band (5340 MHz) multi-parameter Doppler radar system with a bistatic Doppler radar network to establish the next-generation technology of rain observation for meteorological and hydrological applications such as weather forecasts and run-off analysis in predicting floods. This new radar is named COBRA (CRI Okinawa Bistatic polarimetric RAdar). The weather targets of this system are typhoons, Baiu-frontal rainfall, meso-scale precipitation in subtropical zones, and clear air turbulence. Two transmitter (klystron) units are used for the polarization observation. The transmission polarization for each pulse is selected from six possible polarizations. Additionally, NICT has developed equipment for weather radar that modulates the transmitting signal and demodulates the received signal (i.e., a pulse compression function). This equipment has been added to the existing COBRA system, forming a new system referred to as COBRA+. The COBRA+ system uses two traveling-wave tube amplifier (TWTA) transmitter units.

This study investigates the causes of biases in the rain rate estimates by the TMI standard algorithm. Retrieval error is separated from rain/no-rain classification error. As it is well known, the retrieval error is sensitive to storm height. It is interesting that the observed brightness temperature at 85GHz shows clear diurnal variation for given storm height and surface rain rate. It can be explained by the diurnal variation of the land surface brightness temperature (related with the land surface physical temperature) under inhomogeneous rainfall or the diurnal variation of the freezing height (related with the air temperature).

The Dual-frequency Precipitation Radar (DPR) installed on the Global Precipitation Measurement (GPM) core satellite is being developed by JAXA and NICT. This work describes the mission objectives, the precipitation measurement method and techniques, and the construction of the DPR.

This paper outlines the development of the dual-frequency precipitation radar (DPR) to be flown on the Global Precipitation Mission's spacecraft. I. INTRODUCTION In the Global Precipitation Mission (GPM), a dual- frequency precipitation radar (DPR) is planned to be flown on the spacecraft. The spacecraft serves as a high quality reference platform for training and calibrat- ing the rain retrieval algorithms used with the passive mi- crowave radiometers on the other constellation satellites. The dual-frequency radar is expected to provide accu- rate estimates of rainfall rate as well as drop size distribu- tion (DSD) parameters from the combination of Ku- and Ka-band radar returns. This paper outlines the present status of the DPR development. Following this introduc- tion, we discuss the critical issues that affect the designing of the DPR. II. DPR REQUIREMENTS A. Relevance of the DPR to GPM The relevance of the DPR to GPM lies in the radar's ca- pability of measuring storm structure, rainfall rates, drop- size distribution (DSD), path-integrated attenuation, and other useful parameters that cannot be obtained by pas- sive sensors. In the latest design, the DPR is composed of Ku-band and Ka-band channels. The Ku-band radar is approxi- mately the same as the TRMM Precipitation Radar (PR) with some improvements. The Ka-band radar provides high sensitivity to light rain and snow. The combination of data from two channels will provide accurate estimates of drop-size distribution parameters. The Ka-band radar will sample the echo data in two different modes simulta- neously. One is a high-sensitivity mode for light rain and snow detection, and the other is a matched-beam mode in which the sampling volumes of Ka- and Ku-band radar channels are matched for collecting dual-frequency echoes from the identical targets. The data collected in the lat- ter mode are used for the estimation of DSD parameters. In the matched-beam mode, a range resolution of 250 m is employed, while in the high-sensitivity mode, a range resolution of 500 m is planned. The current radar design adopts active phased array antennas in both radar chan- nels to make full use of TRMM experience. The DPR will provide three-dimensional information of hydrometeor distribution with high spatial resolution. Such data are very valuable for the study of storm struc- ture. The accurate rainfall estimates from the DPR are expected to be used for calibrating the corresponding esti- mates from the radiometer on the core satellite. The major importance of the DPR, however, lies in the fact that it can provide the regional and seasonal statistics of storm structure together with DSD parameters. Since rain re- trieval algorithms for passive microwave radiometers have to assume a vertical structure of storm either determinis- tically or statistically, reliable storm structure information is crucial for the accuracy of rain estimation. The statis- tics from the DPR can be used as a database in radiometer algorithms to reduce the uncertainties of the storm mod- els. How to utilize the information from radar data is a challenging issue. A possibility of improving the database used in a TMI rain retrieval algorithm by using TRMM's PR data is currently under examination. The DPR has three main roles in GPM. It will provide three-dimensional information of rain structure. The Ku- band radar is similar to, but not exactly the same as, the TRMM Precipitation Radar (PR). It is improved from the PR. The improvement is necessary because of three rea- sons. Firstly, the proposed orbit of the GPM core satellite is about 400 km and higher than the TRMM's orbit. This necessitates the improvement of the sensitivity to com- pensate for the increased range loss. The designed trans- mitting power is increased to 1000 W from PR's 500 W. (the actual Tx power of the PR turned out to be about 800 W.) Secondly, the orbital inclination is about 65 de- grees and larger than the TRMM's 35 degrees. Because of the oblate shape of the Earth, the altitude of the satel- lite changes more than 20 km at a 65-degree orbit which is much larger than 10 km at a 35-degree orbit. If we use a constant pulse repetition frequency (PRF) like the TRMM PR, we have to use a rather small PRF to absorb this large variation of rain echo range from the radar. The low PRF will result in a low signal-to-noise ratio. To max-

Communications Research Laboratory (CRL) has developed a new C-band (5340-MHz) multi-parameter Doppler radar system with a bistatic Doppler radar network. The purpose is to establish the next-generation technology of precipitation observation for meteorological and hydrological applications, such as weather forecasting and run-off analysis for predicting floods. This new radar is named COBRA (CRL Okinawa Bistatic polarimetric RAdar). The weather targets of the system are typhoons, Baiu-frontal rainfall, meso-scale precipitation in subtropical zones, and clear-air turbulence. Two transmitter (klystron) units are used for to observe the polarization characteristics of precipitation particles. The transmission polarization for each pulse is selected from six possible polarizations. The return signal is measured simultaneously by two receivers, one for the horizontal polarization and the other for the vertical polarization.