Japan Meteorological Research Articles

GMDH-based wave prediction model for one-week nearshore waves using one-week forecasted global wave data

Wave forecasting up to 168 h (one-week) ahead is crucial to the operational management of shipping and construction in the marine work site. The present study proposes a method of areal-time One-week Wave Forecast of Nearshore Waves (OWFNW) at 13 stations on the Japanese coast. For each station, two Group Method of Data Handling (GMDH) based wave prediction models for the wave height and period are developed to transform the global wave data into nearshore waves. For each GMDH-based wave prediction model, four models are trained with four different training data sets using the one-week global wave forecast data released in real-time by the Japan Meteorological Agency (JMA), National Oceanic and Atmospheric Administration (NOAA), and European center for Medium-Range Weather Forecasts (ECMWF). The study conducts a series of training and testing GMDH-based wave prediction models at each station, and examines the accuracy of predictions. The best performing GMDH-based wave prediction model selected among the four models is compared to the best global wave forecast data. It was found that the best performance GMDH-based wave prediction model varies for different locations. On average, the best performing GMDH-based wave height model is able to improve the accuracy of nearshore wave predictions up to 38% compared to that of the best global wave height forecast with 168 h lead time. The best GMDH-based wave period model improves the prediction error up to 65% compared to the 168 h ahead error of the best global wave period forecasts. The present methodology to get OWFNW can be applied to other nearshore seas where global wave forecast data and corresponding observed one are available.

Extreme Typhoon Wind Speed Mapping for Coastal Region of China: Geographically Weighted Regression–Based Circular Subregion Algorithm

The coastal region of China is exposed to severe typhoon threats and suffers significant casualties as well as economic losses due to seven to eight landed storms every year. This study developed a geographically weighted regression (GWR) circular subregion algorithm that considers the spatial variation characteristics of typhoon storms to map extreme-typhoon wind speeds. The continuously varying coefficient maps for typhoon-tracking, intensity, and wind field parameter recursive models are determined in the simulation domain using the best track data set from the Japan Meteorological Agency and geographically weighted regression. By introducing a genesis parameter model, each model module is validated independently or together by comparing the records in the best track data set in various circular subregions. The effects of the size of the circular subregion on the convergence of the typhoon wind hazard curve are investigated at two sites before an appropriate size is recommended. Extreme typhoon wind speeds of 10 coastal cities are modeled using the present model coupled with a well-developed analytical wind field model. Several factors that potentially contribute to the remarkable differences between this research and code suggestions as well as other studies are discussed. The estimated return period values of the annual maximum typhoon wind speed at 1,079 grid points in the coastal region of China are interpolated to achieve wind hazard contour maps. Several new findings regarding the spatial characteristics of extreme typhoon winds speed along the coastal region of China are examined.

The Role of Free‐Tropospheric Moisture Convergence for Summertime Heavy Rainfall in Western Japan

AbstractThe role of vertically integrated water vapor flux convergence (IVFC) in heavy rainfall events in western Japan is investigated by dividing the atmosphere into a boundary layer and free troposphere using the Japan Meteorological Agency's mesoscale gridded analysis product and reanalysis product. Rainfall events are defined by peaks of area‐averaged precipitation. The results show an increase in the free‐tropospheric IVFC more than 15 h before rainfall events, contributing to atmosphere moistening. This preceding moistening is favorable for mesoscale convective systems (MCSs) with slantwise ascending deep inflow layers (layer lifting) that produce large precipitation amounts. Appearance frequency of the moist absolutely unstable layer (MAUL) increases around the precipitation peak times, which is especially notable for larger peaks. These results indicate that the free‐tropospheric IVFC contributes to heavy rainfall in western Japan by providing environments favorable for producing and maintaining MAUL, enhancing the organization of MCSs.

Impact of air temperature on occurrence of bath-related cardiac arrest.

The mortality of the bath-related cardiac arrest (BRCA) is extremely high. While air temperature is reported to be associated with the BRCA occurrence, it is unclear whether daily minimum temperatures or the difference between maximum and minimum air temperatures influences BRCA occurrence the most.A retrospective cohort study of adult patients was conducted between January 2015 and February 2020 at Hirosaki University Hospital Emergency Department. The following data were collected: age, sex, day of cardiac arrest event, location of the event, initial cardiac rhythm, presence of return of spontaneous circulation, and overall mortality (status at 1 month after cardiac arrest event). Based on the day of the event and the location in which the event occurred, daily minimum and maximum temperatures were obtained from the Japan Meteorological Agency database.A total of 215 eligible cardiac arrest cases were identified, including 25 cases of BRCA. Comparing BRCA and non-BRCA, initial shockable cardiac rhythm (4.0% vs 44.7%), presence of return of spontaneous circulation (8.0% vs 34.7%), and overall mortality (96.0% vs 71.6%) differed significantly (P < .05 each). Daily minimum and maximum temperatures showed no significant relationships with BRCA or non-BRCA. Daily minimum temperature was a risk factor of BRCA occurrence after adjusting for age and temperature difference (risk ratio, 0.937; 95% confidence interval, 0.882–0.995).Daily minimum temperature represents a potential risk factor for BRCA occurrence.

Shallow soil effects for contrasting PGAs at nearby strong-motion stations during the 2017, Mj 5.2, southern Akita Prefecture earthquake

On September 8, 2017, an earthquake of Mj 5.2 occurred with the epicenter in southern Akita Prefecture, Japan, at 22:23 local time. According to the Japan Meteorological Agency (JMA), the focal depth was 9 km. Many strong-motion stations of K-NET and KiK-net recorded ground motions from the earthquake. The maximum horizontal vector peak ground acceleration (PGA) of approximately 136 cm/s2 was recorded at one of the KiK-net stations at an epicentral distance of about 8 km. However, despite being 37 km and 53 km far from the epicenter, two stations recorded PGAs of approximately 126 and 113 cm/s2, respectively, similar to that near the epicenter. Even though these PGAs are not rare, we found that the PGAs at the two sites strongly deviated from the median values suggested by a ground motion prediction equation (GMPE), while the nearby sites generally followed the GMPE. Available velocity models showed that shallow shear wave velocities, especially in the top 5 m, were lower (i.e., the soils were softer) at the two sites compared to those at their nearest neighboring sites. We compared the ratios of the PGAs and peak ground velocities (PGVs) at the two sites with respect to their neighboring sites for many earthquakes covering a wide range of magnitudes and azimuths. We found that the PGAs and PGVs at the two sites were systematically larger than those at the adjacent sites. Linear theoretical site amplifications using the available soil models gave peak frequencies around 6-8 Hz at the larger PGA sites. Bandpass-filtered records showed significantly larger PGAs around these frequencies at the larger PGA sites. The above results showed that local site condition is one of the major contributing factors to induce large PGAs. Furthermore, softer sites experience more substantial nonlinear site amplification than the stiffer sites when input motions exceed some threshold PGAs. This latter effect means that the softer sites can produce a variety of ground motion spectra. Nevertheless, the degree of damage to built structures depends on several factors, including the design and quality of construction. We expect that this study contributes to developing improved microzonation maps for earthquake disaster mitigation.

818Prediction of health effects of Asian dust using the MASINGAR in Japan

Abstract Background Health effects of cross-border air pollutants and Asian dust are of significant concern in Japan. Currently, models predicting arrival of aerosols have not investigated the association between arrival predictions and health effects. We investigated the association between subjective health symptoms and data acquired from the Japan Meteorological Agency's (JMA's) the Model of Aerosol Species in the Global Atmosphere (MASINGAR) aerosol model with the objective of ascertaining if the data could be applied for predicting health effects. Methods Subjective symptom scores were collected using self-administered questionnaires and used with JMA model’s surface concentration data to conduct a risk evaluation using multiple linear mixed model, during 2013 to 2015. Altogether, 160 individuals provided 16226 responses. Data regarding climate (temperature, humidity, and atmospheric pressure) and environmental factors (NO2, SO2 and Ox) were used as covariates. We calculated the association between the surface dust concentration and symptoms. Results A strong association was also observed for nasal and cough symptoms (P for trend &amp;lt; 0.001). The differences in scores of nasal symptoms (sneezing and runny) of the highest quartile [Q4] vs. the lowest [Q1] were 0.039 (95% confidence interval (CI): 0.02–0.01, p &amp;lt; 0.05) and 0.046 (95% CI: 0.002–0.02, p &amp;lt; 0.05), respectively. The differences in scores of cough symptoms were 0.036 (95% confidence interval (CI): 0.002–0.01, p &amp;lt; 0.05). Conclusions This study suggests that predictive models for pollutants’ arrival can be used to capability to foresee and possibly prevent the health impact of long range transport of air pollutants, recommending the potential role of aerosol forecast models in health care. MASINGAR is Global Spectral Model (GSM), this have the potential that can contribute in health predictions all over the world. Key messages Asian dust, Health forecast, Allergic symptom

Potential of Regional Ionosphere Prediction Using a Long Short‐Term Memory Deep‐Learning Algorithm Specialized for Geomagnetic Storm Period

AbstractIn our previous study (Moon et al., 2020, https://doi.org/10.3938/jkps.77.1265), we developed a long short‐term memory (LSTM) deep‐learning model for geomagnetic quiet days (LSTM‐quiet) to perform effective long‐term predictions for the regional ionosphere. However, their model could not predict geomagnetic storm days effectively at all. This study developed an LSTM model suitable for geomagnetic storms using the new training data set and redesigning input parameters and hyper‐parameters. We collected 131 days of geomagnetic storm cases from January 1, 2009 to December 31, 2019, provided by the Japan Meteorological Agency's Kakioka Magnetic Observatory, and obtained the interplanetary magnetic field Bz, Dst, Kp, and AE indices related to the geomagnetic storm corresponding to each storm date from the OMNI database. These indices and F2 parameters (foF2 and hmF2) of Jeju ionosonde (33.43°N, 126.30°E) were used as input parameters for the LSTM model. To test and verify the predictive performance and the usability of the LSTM model for geomagnetic storms developed in this manner, we created and diagnosed the 0.5, 1, 2, 3, 6, 12, and 24‐h predictive LSTM models. According to the results of this study, the LSTM storm model for 24‐h developed in this study achieved a predictive performance during the three geomagnetic storms about 32% (10%), 34% (17%), and 37% (5%) better in root mean square error of foF2 (hmF2) than the LSTM quiet model (Moon et al., 2020, https://doi.org/10.3938/jkps.77.1265), SAMI2, and IRI‐2016 models. We propose that the short‐term predictions of less than 3 h are sufficiently competitive than other traditional ionospheric models. Thus, this study suggests that our model can be used for short‐term prediction and monitoring of the regional mid‐latitude ionosphere.

Ambient temperature and ambulance car visits in Japan by age category and severity

BACKGROUND AND AIM: Using the Japanese public domain data, we have reported that a severe heat exposure during the power outage on 10th of September due to the typhoon raised the number of ambulance car visits due to heatstroke disproportionately. Here, we investigated which age groups and which severity categories were affected. METHODS: Because the Typhoon Faxai hit Japan in September 2019, we collected the ambulance car visit data for September from 2010 to 2019 from the Fire and Disaster Management Agency, the Ministry of Internal Affairs and Communications. Corresponding weather data were collected from the Japan Meteorological Agency. Both datasets are publicly available from the respective websites. Because the already-categorized data either by age group or by severity level were available, we set the age groups as 0-6, 7-17, 18-64 and 65+, and severity levels as mild, moderate, severe, and deceased. The comparison was on the difference between the September heatstroke risk in 2019 and that in the preceding years. First, we examined Chiba prefecture, in which the power outage was by far extensive. As the control, we also examined the surrounding 6 prefectures, where the power outage was minimal. RESULTS:As for Chiba, age groups 18-64 and 65+ and mild to moderate severity groups showed much higher risk. The control prefectures did not show increased risk. Deceased cases were rare, but two deaths in Chiba and no death in the other prefectures were identified on 10th of September. CONCLUSIONS:The severe heat exposure during the typhoon-related power outage increased the risk of mild and moderate heatstroke among 18 years and older age group. KEYWORDS: Temperature,Heatstroke, Ambulance car visits

Application of the Modified PLUM Method to a Dense Seismic Intensity Network of a Local Government in Japan: A Case Study on Tottori Prefecture

An application of the PLUM (Propagation of Local Undamped Motion) method to real-time seismic intensity distribution from a dense seismic intensity network maintained by a local government in Japan is demonstrated. The JMA (Japan Meteorological Agency) has employed the PLUM method from March 2018 as a supplement to the traditional EEW (Earthquake Early Warning) which requires hypocenter determination. A dense observation network is desirable for upgrading accuracy and immediateness of EEW including the PLUM method. Seismic intensity meters in Tottori Prefecture, Japan, are suitable for the purpose because they have been improved to broadcast packets with peak ground acceleration and JMA seismic intensity at every 1 s. Also, 34 seismic intensity meters are installed in the target area while only six observation sites are used for the EEW by the JMA. The packet data are received at Tottori University, and the modified PLUM method considers wave propagation with damping from not only the observation points but also all evaluating grid points at a 1 km mesh. Additionally, P-wave amplitudes in preliminary tremors are also introduced to estimate seismic intensities from an empirical relationship. Applying the methodology to past earthquake data, more detailed and rapid evaluations of seismic intensity distribution are achieved. We expect the system will be utilized for earthquake disaster mitigation measures of local governments.

Evaluation of uncertainties derived from meteorological forecast inputs in plume directions predicted by atmospheric dispersion simulations

ABSTRACT Atmospheric transport, dispersion, and deposition models (ATDMs) can support decision-making during nuclear emergencies; however, uncertainties in the ATDM results need to be carefully evaluated. To investigate the uncertainties derived from meteorological forecast inputs, we conducted three-day forecast simulations every day for one year with hypothetical releases of radionuclides (one-hour releases every 6 h) from a nuclear facility. The forecast outputs were compared with the analysis outputs during the same period. The difference between the outputs is treated as the uncertainty in the forecasts and is represented as an angle based on the discrepancy in the plume directions between the analysis and forecast outputs. Using meteorological inputs made by Japan Meteorological Agency, the discrepancy angle (Ang) increased by approximately 10 per day on an annual average basis. Meanwhile, the Ang values were occasionally 4–5 times higher than the annual average during short time periods. Since the Ang time series show seasonal and diurnal changes, the statistical characteristics likely depend on the geographical and meteorological conditions, as well as the types of meteorological inputs. Additionally, a main factor in the uncertainty is the wind-direction difference between the analysis and forecast outputs on scales of more than or less than 100 km.

The unusual case of the ultra-deep 2015 Ogasawara earthquake (MW7.9): Natural time analysis

On 30 May 2015, a powerful earthquake (EQ) (M W 7.9, Japan Meteorological Agency reported magnitude M8.1) struck west of Japan's remote Ogasawara (Bonin) island chain, which lies more than 800 km south of Tokyo. It occurred at 680 km depth in an area without any known historical seismicity. It was the first EQ felt in every Japanese prefecture since intensity observations began in 1884. Here, by applying natural time analysis, which unveils hidden properties in complex time series, we find that almost three and a half months before this powerful EQ, which is the strongest one after the M W 9.0 Tohoku EQ on 11 March 2011, the fluctuations of the order parameter of seismicity were minimized around 17 February 2015. Remarkably, such a behavior has been also observed 1–3 months before all shallow EQs in Japan of magnitude M7.6 or larger that occurred since 1984 until the M W 9.0 Tohoku EQ. We also find that upon the Ogasawara EQ occurrence the change of the entropy of Japanese seismicity upon changing the direction of the time arrow (i.e., under time reversal) exhibited a minimum. This minimum, which may appear when a system approaches a dynamic phase transition, is not equally deep with that observed on 22 December 2010 along with the occurrence of a M7.8 EQ in Bonin islands. The fact that the latter EQ was followed almost 3 months later by the appreciably stronger Tohoku EQ, while the Ogasawara EQ did not, is discussed in the frame of the Lifshitz-Slyozov-Wagner theory of phase transitions and subsequent work by Penrose and coworkers.

Long-term observation of electrical discharges during persistent Vulcanian activity

Very low frequency and wide-band lightning detection networks can detect major volcanic plumes via their intense electrical and lightning activity. However, the high number of non-detected explosive episodes confirmed by direct observations, reveals the limits of these systems in the detection of the more frequent smaller ash-rich explosive events. Here, we use a data-efficient thunderstorm detector to observe electrical discharges generated from July 2018 to January 2020 by the persistent Vulcanian activity of Minamidake crater at Sakurajima volcano in Japan.Two thunderstorm detectors recorded the electrical activity produced by explosions at Minamidake crater from a distance of 3 and 4 km from the active vents. The instruments measured the induced current due to the change in electric field with time within the extremely low frequency range (1-45 Hz). Using a volcanic lightning detection algorithm together with the catalogue of volcanic explosions compiled by the Japan Meteorological Agency (JMA) and Tokyo Volcanic Ash Advisory Center (Tokyo VAAC), the number of electrical discharges, the electrical discharge rate and the total amount of measured voltage were determined for each individual explosive event. In addition, the start of the electrical discharges was compared to the explosion onset as provided by the JMA (with a one-minute time resolution).The sensors detected electrical discharges in 71% of the 724 recorded explosions. Our detection algorithm successfully recognises the presence/absence of electrical discharges with an accuracy of 73%. We find a non-linear positive correlation between the number of discharges and the plume height. Moreover, we find that the maximum electrical discharge rate and the maximum amount of measured voltage by a single discharge also increase with plume height. Fracto- and tribo-electrification appear to be the dominant plume electrification mechanisms. Even for the few explosive events that exceeded the −10°C isotherm, the timescale of electrical activity seems to be too short for ice nucleation to make a significant contribution to the plume electrification. Finally, for 12% of the electrically-active explosive events, discharges were detected by the sensors more than a minute before the JMA explosion onset.Our results show the capability of our detectors in pinpointing the inception of electrified explosive episodes in real-time and in providing an indication of the magnitude of each explosion, demonstrating their effectiveness as a cost- and data-efficient instrumentation for the monitoring of explosive ash emissions at active volcanoes.

Simulation of Storm Surges in Bangladesh Using NWP Models

Tropical cyclone causes enormous damage of property and life in the coastal regions of Bangladesh. Majority of the damage during tropical cyclone occurs because of storm surge. Bangladesh is vulnerable to storm surge flooding due to its complex geometry and location. The huge loss of property and life can be reduced by predicting storm surge during a tropical cyclone. So, an attempt has been made in this study to see the effectiveness of predicting storm surge by Meteorological Research Institute (MRI) storm surge model from Japan Meteorological Agency (JMA). In this study, MRI storm surge model is used to simulate storm surge of five cyclones cyclone Bulbul (November 2019), cyclone Fani (May 2019), cyclone Mora (May 2017), cyclone Roanu (May 2016) and cyclone Komen (July 2015) were simulated using MRI storm surge model to see the effectiveness of the model. The model was run for 48-hours for two different data sets which are, European Centre for Medium-Range Weather Forecasts (ECMWF) data and Global Spectral Model (GSM) data. Simulated storm surge heights found from the model are compared with the reported surge height from Bangladesh Inland Water Transport Authority (BIWTA) and Bangladesh Meteorological Department (BMD). The model is able to simulate storm surge height, mean sea level pressure (MSLP) and wind data of all the selected cases. The time series graph developed using python comparing the simulated results of storm surge height data with observed data for the event cyclone Bulbul showed the simulated data of both NWP datasets were in line with the observed data. And comparison of the maximum storm surge height for all five events shows simulated data for both the data sets were close to the reported data and the difference was well within the margin of error. The results indicate MRI model as a useful tool for storm surge forecasting&#x0D; The Dhaka University Journal of Earth and Environmental Sciences, Vol. 9(1), 2020, P 31-38

How reliable are TIGGE daily deterministic precipitation forecasts over different climate and topographic conditions of Iran?

AbstractAccurate forecasting of precipitation has been a very crucial issue for developing meteorological and hydrological early warning systems. Ensemble forecasts using numerical weather prediction models are promising for these purposes. This research addressed the reliability of global precipitation forecasts of the THORPEX (The Observing System Research and Predictability Experiment) Interactive Grand Global Ensemble (TIGGE) archive over distinct climate and topographic regions of Iran. Besides, post‐processing the raw forecasts and assessing TIGGE datasets' performance in predicting extreme rainfall events, at the thresholds of 10–20, 20–50, 50–100, and higher than 100 mm, are other objectives of this study. Results indicated that the European Centre for Medium‐Range Weather Forecasts (ECMWF) and Meteo France forecasts, with the mean correlation coefficients (CCs) of 0.64 and 0.61, were closer to the ground‐gauge observations, especially in different types of humid zones. On the contrary, in semi‐arid and extra‐arid climate zones, the CC value was relatively low. Furthermore, Meteo France, Korea Meteorological Administration, Japan Meteorological Agency and ECMWF outperformed other models in detecting rainy or non‐rainy days. Also, by increasing the elevation, the correlation between forecasts and in situ observations increased significantly, and the percentage increment of CC for ECMWF and Meteo France models is higher than 38%. Moreover, the skill of predictions for extreme events (higher than 50 mm) decreased meaningfully, while at the thresholds of 10–20 and 20–50 mm, ECMWF and China Meteorological Administration outperformed other models. Finally, post‐processing of raw forecasts using quantile mapping improved the CC and relative bias metrics up to 30% and 51%, respectively, especially in extra‐arid, semi‐arid and Mediterranean climate zones.

APPLICABILITY OF VECTOR AUTOREGRESSIVE MODEL TO ESTIMATION OF INTERNAL MOISTURE CONDITION OF CONCRETE EXPOSED TO ORDINARY NATURAL ENVIRONMENT

To investigate the applicability of VAR-X model to estimating moisture condition of concrete exposed to ordinary natural environment, concrete internal humidity data reported previously were analyzed with weather data from Japan Meteorological Agency. The first 90 days were used for parameter estimation and the second 60 days were used to validate the model and predicted values. The results are summarized below. (1) Within the scope of this study, VAR-X model was applicable to the humidity data in the non-rainfall environment. VAR-X model was also able to predict the internal humidity for 60 days based on the estimated parameters and weather data (temperature and humidity), and the RMSE ranged from 0.87 to 2.04%. (2) It was judged that the VAR-X model was not appropriate to be applied to the humidity data of rainfall environment. The main reason for this is that the impact of rainfall was underestimated by assuming constant coefficients in the model. (3) The mechanisms of humidity fluctuation in concrete are discussed, which include the absorption and release of water and the temperature dependence of concrete water content. Based on the mechanism, difference between rainfall and non-rainfall environment was pointed out and the conditions under which the VAR-X model can be applied are discussed. (4) Methods of utilization and improvement of VAR-X model were discussed for cases where the moisture transfer characteristics of concrete changed and where the data increased or decreased in high humidity region. The advantages of time series analysis, including VAR-X model, were also pointed out in comparison with conventional physical models. VAR-X model can predict humidity fluctuations in concrete using weather data when certain conditions are met. In the future study, the model will be modified and its application to the durability evaluation of various materials will be investigated.

Acceleration of rises in the level of rivers induced by rising air temperatures in a cold climate

The effects of climate change on water discharge in rivers in cold climates were investigated. To quantify the impacts of air temperature rises on the promotion of snowmelt and associated acceleration of a rise in the level of a river, 10 rivers on Hokkaido, northwestern Japan were chosen. Available data of daily water discharge for more than half a century by the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) were correlated with air temperatures observed at the 8 weather stations operated by the Japan Meteorological Agency (JMA). The analyses show that annual average air temperatures have risen at all 8 sites on Hokkaido for the period from 1954 to 2018. The trends for the 8 sites show a range from 0.119 °C/decade to 0.250 °C/decade with an average of 0.191 °C/decade. Annual snowfall has increased at Sutsu. The trend over the period from 1954 to 2018 was 45.2 cm/decade. Otherwise, annual snowfall has not changed significantly in the period (1954–2018). Accelerations of the rise in the level of the river induced by air temperature rises have been observed in the Teshio River, Ishikari River, Saru River, and Mu River. A delay in the rise in the level of the river caused by an increase in snowfall and the weak warming trend from January to April has been observed in the Shiribetsu River. Although air temperatures have risen year after year, a significant change in annual pattern of daily flow has not been observed in the Syokotsu River, Yubetsu River, Tokoro River, and Abashiri River located in eastern Hokkaido. This can be induced by the weak warming trends in April which may be due to the cold current of the sea of Okhotok.

Regional Solar Irradiance Forecast for Kanto Region by Support Vector Regression Using Forecast of Meso-Ensemble Prediction System

From the perspective of stable operation of the power transmission system, the transmission system operators (TSO) needs to procure reserve adjustment power at the stage of the previous day based on solar power forecast information from global horizontal irradiance (GHI). Because the reserve adjustment power is determined based on information on major outliers in past forecasts, reducing the maximum forecast error in addition to improving the average forecast accuracy is extremely important from the perspective of grid operation. In the past, researchers have proposed various methods combining the numerical weather prediction (NWP) and machine learning techniques for the one day-ahead solar power forecasting, but the accuracy of NWP has been a bottleneck issue. In recent years, the development of the ensemble prediction system (EPS) forecasts based on probabilistic approaches has been promoted to improve the accuracy of NWP, and in Japan, EPS forecasts in the mesoscale domain, called mesoscale ensemble prediction system (MEPS), have been distributed by the Japan Meteorological Agency (JMA). The use of EPS as a machine learning model is expected to improve the maximum forecast error, as well as the accuracy, since the predictor can utilize various weather scenarios as information. The purpose of this study is to examine the effect of EPS on the GHI prediction and the structure of the machine learning model that can effectively use EPS. In this study, we constructed the support vector regression (SVR)-based predictors with multiple network configurations using MEPS as input and evaluated the forecast error of the Kanto region GHI by each model. Through the comparison of the prediction results, it was shown that the machine learning model can achieve average accuracy improvement while reducing the maximum prediction error by MEPS, and knowledge was obtained on how to effectively provide EPS information to the predictor. In addition, machine learning was found to be useful in improving the systematic error of MEPS.

ESTIMATION OF SUBSURFACE STRUCTURES AND GROUND MOTION CHARACTERISTICS IN THE AREAS DAMAGED IN THE 2016 CENTRAL TOTTORI PREFECTURE EARTHQUAKE IN JAPAN

On 21 October 2016, an earthquake with a magnitude of 6.6 (determined using the Japan Meteorological Agency’s [JMA] seismic network) occurred in the center of Japan’s Tottori Prefecture. In the city of Kurayoshi and in the towns of Yurihama and Hokuei, JMA seismic intensities of lower 6.0 were observed, and structural damage to housing was concentrated in a limited number of areas. This research conducted microtremor observations around damaged areas and used previous studies to estimate subsurface structures and ground motion characteristics. Low-velocity layers with an S-wave velocity of between 80 and 200 m/s were estimated at all observation sites and a tendency is present for the thickness to increase towards the coastal plains and decrease inland. In an evaluation of the relationship between ground motion characteristics and earthquake damage, researchers found that the occurrence of damage correlated to significant changes in layer thickness. Researchers believed that the damage correlates to the 2-D or 3-D effects of subsurface structures

Design Philosophy for Buildings’ Comfort-Level Performance

The data reported by Japan Meteorological Agency (JMA) show that the fatal casualties and severe injuries are due to heavy shaking during massive earthquakes. Current earthquake-resistant building standards do not include comfort-level performance. Hence, a new performance design philosophy is proposed in this research to evaluate the quantitative effect of earthquake-induced shaking in a building. The earthquake-induced response accelerations in a building are analysed, and the response accelerations related with the characteristic property of the building are used to evaluate the number of Seismic Intensity Level (SIL). To show the indispensability of the newly proposed comfort-level design philosophy, numerical simulations are conducted to evaluate the comfort level on different floors in a building. The results show that the evaluation of residents’ comfort levels should be considered in the current earthquake-resistant building design codes.

Assimilation of Himawari-8 imager radiance data with the WRF-3DVAR system for the prediction of Typhoon Soudelor

Abstract. Himawari-8 is a next-generation geostationary meteorological satellite launched by the Japan Meteorological Agency. It carries the Advanced Himawari Imager (AHI) on board, which can continuously monitor high-impact weather events with high frequency in space and time. The assimilation of AHI radiance data was implemented with the three-dimensional variational data assimilation system (3DVAR) of the Weather Research and Forecasting Model for the analysis and prediction of Typhoon Soudelor (2015) in the Pacific typhoon season. The effective assimilation of AHI radiance data in improving the forecast of the tropical cyclone during its rapid intensification has been realized. The results show that, after assimilating the AHI radiance data under clear-sky conditions, the typhoon position in the background field of the model was effectively corrected compared with the control experiment without AHI radiance data assimilation. It is found that the assimilation of AHI radiance data is able to improve the analyses of the water vapor and wind in a typhoon's inner-core region. The analyses and forecasts of the minimum sea level pressure, the maximum surface wind, and the track of the typhoon are further improved.