A statistical study was conducted on weekday-holiday differences in atmospheric pressure in the central part of Tokyo and Osaka, and on surface wind fields in surrounding areas. The analysis for pressure was based on 33-year data from stations of the Japan Meteorological Agency (JMA). The analysis of winds was based on 44-year data from the Automated Meteorological Data Acquisition System (AMeDAS) of the JMA and 28-year data from the Air Pollution Monitoring System (APMS) of the Tokyo Metropolis. It was found that the pressure in central Tokyo was higher during the daytime on holidays than on weekdays, by approximately 0.04 hPa at 1500 JST. The daytime surface winds had a divergent anomaly within several tens of kilometers of the city center, corresponding to a wind speed anomaly of the order of 0.1 m s⁻¹. Similarly, a pressure anomaly of approximately 0.015 hPa was found in the afternoon on holidays in central Osaka, as well as a divergent anomaly in surface winds in the surrounding area.

This is a case study of March 2023 when Japan experienced significant warming, along with its prediction using the atmospheric reanalysis dataset and Japan Meteorological Agency's Global Ensemble Prediction System (GEPS). A major sudden stratospheric warming event occurred in mid-February 2023 with a shift of the center of the stratospheric polar vortex toward northern Eurasia. In conjunction with the stratospheric polar vortex shift, the tropospheric westerly jet over Japan shifted poleward, leading to significant warming in March 2023. Regression analyses were conducted using the GEPS ensemble forecast initialized on February 15, 2023, which predicted the significant warming phenomenon near Japan. The regressed anomalies of geopotential height from Japan to the mid-latitude North Pacific showed that the strengthening of the displaced stratospheric polar vortex over Siberia was related to the poleward shift of the westerly jet near Japan. This finding suggests that when the stratospheric polar vortex shifts toward Siberia, the subsequent significant warming near Japan can be predicted with reliability. The geographical position of the stratospheric polar vortex is likely to be a key predictor to improve the forecast skill of the weather conditions near Japan in early spring.

In early February 2025, a 12-hour snowfall of 120 cm was observed in Obihiro, located on the Tokachi Plain of southeastern Hokkaido, Japan; it was the highest recorded snowfall in Japan. Concurrently, a marine heatwave (MHW) with pronounced warm sea surface temperature was observed offshore. While MHWs effect on rainfall are documented, their impact on snowfall remains poorly understood. Here, we demonstrated the mechanisms behind the record-breaking snowfall event, including the effects of the MHW. During the heavy snowfall, an extratropical cyclone drove strong easterly winds toward the coastal regions of the Tokachi Plain, and a surface front was located to the south of Obihiro. The easterly winds transported a convectively unstable layer from over the ocean into the front, and the frontal updrafts released the instability. Consequently, convective precipitation systems developed, yielding heavy snowfall at Obihiro on the cold side of the front. Notably, the MHW enhanced the frontal formation and convective instability, increasing precipitation around Obihiro by approximately 50%. This case study demonstrates that MHWs can significantly amplify snowfall under specific atmospheric conditions, advancing our understanding of compound extreme.

In summer 2024, record-high temperatures were observed over Japan, where the area-averaged summer-mean temperature was tied highest with 2023. The extreme heatwave in July was attributable primarily to the poleward-deflected upper-tropospheric subtropical jet (STJ). The low-level North Pacific Subtropical High (NPSH) intensified to the immediate south of Japan under the remote influence of enhanced cumulus convection over the northern Indian Ocean. Persistent anomalous descent and increased solar radiation associated with the NPSH contributed to the record-high temperatures in the southern portion of Japan. In August, the persistent heatwave, particularly over western Japan, was attributable to the poleward-deflected STJ and enhanced convection associated with a lower-tropospheric cyclonic gyre to the southeast of Japan where several typhoons sequentially formed. Meanwhile, heavy rainfall over northern Japan in late July was attributable primarily to developed convective systems organized just to the south the Baiu front under the intensified moist westerly airflow to the north of the markedly-extended NPSH over western Japan. Other factors that could contribute to the extreme heatwave and heavy rainfall in 2024 summer include extreme warmth of the surrounding ocean, global warming, and extremely high zonal-mean temperatures in the midlatitude Northern Hemisphere troposphere during a post El Niño summer.

Sudden localized heavy rainfall events, capable of disrupting daily life and damaging infrastructure, are becoming more frequent. Their nowcasting (very short-term forecast) requires higher spatiotemporal (4D) resolution than conventional radars, and effective 4D methods to extrapolate the vertical development of convective systems. This study evaluates the performance of a new system that generates 10-minute lead-time precipitation nowcasts in real time, which are used by a publicly available smartphone application to issue heavy rainfall warnings. Dense 4D observations from new Multi-Parameter Phased Array Weather Radars (MP-PAWR) in Saitama, Osaka, and Kobe (Japan) are extrapolated using an Artificial Neural Network (ANN4D), which has demonstrated high performance in forecasting the sudden onset of precipitation in Saitama, prior to 2020. The study demonstrates that, despite using the same ANN4D instance, the system generates reliable nowcasts, generalizes well to new locations and years, and that performance is enhanced by a post-ANN4D procedure for mitigating false rainfall predictions. ANN4D outperforms a 4D Eulerian model (TREC4D) in predicting convective rainfall onset, while TREC4D is more efficient for well-developed rainfall. The study identifies minor issues, like the need to expand ANN4D's vertical range, and highlights the next major step: integrating ANN4D with TREC4D to exploit their complementarity.

Based on a cloud resolving model (CRM) simulation, cumulus parameterization (CP) problem for gray zone is revisited. In the CRM simulation of deep cumulus convections in the tropical Pacific, we can see many intermittent kilo-meter scale ascending updraft bubbles (thermals). If we gather these updraft bubbles in a 20 km grid area into one place, they can be represented by a single hypothetical large plume: equivalent plume (EP). The total updraft in the grid is equivalent to the updraft of EP. It is shown that the mean temperature of EP is almost the same as the environment temperature. As the EP is saturated, the mean specific humidity of EP is the same as the saturation specific humidity at the environment temperature. The EP shows that “moist buoyancy” is essential for cumulus convections. With the mean temperature and mean specific humidity of EP, we can calculate the moist static energy and the entrainment rate for EP. Using this entrainment rate and the updraft mass flux at cloud base which is proportional to fractional convective cloud area, we can calculate the updraft mass flux necessary for a CP scheme for gray zone.

This study investigates the uncertainties in future lightning activity over eastern Japan using a convective-permitting model coupled with an explicit-bulk lightning model (BLM). Projected changes to four commonly used lightning flash parameterizations (LFP), the product of Convective Available Potential Energy and precipitation (CP), Updraft Volume (UV), updraft Ice Flux (IFlux), and the McCaul Lightning Flash Algorithm (MLFA), in addition to results from the BLM, are computed by applying the pseudo-global warming method on two highly active lightning cases. Results from LFPs exhibited high uncertainty in future flash counts, with increasing CP and UV, and decreases or minimal change using the IFlux and the MLFA for both cases. The BLM alone exhibits a large decrease in projected flash counts for both cases. The BLM explicitly computes various cloud-electric properties, such as charge separation due to graupel and ice/snow collision, which is widely believed to be an important mechanism for lightning. In future climates, less charge separation per collision and fewer collisions are projected due to changes in cloud ice properties that are not considered in LFPs. The BLM, therefore, offers an alternative assessment of projected lightning activity.

The drainage flow in a valley was observed with meteorological sensors attached on compact drones. We first developed a wind measurement system by mounting an ultrasonic anemometer on a drone and evaluated its performance by comparing the data with those from wind measurements on a meteorological tower and on a pole. With this system, the vertical structure of drainage flow was observed through multiple observation campaigns. By compiling 35 profiles in total, the statistical features of their vertical profiles in temperature, humidity and horizontal winds have been derived. It is found that the flow is characterized by a strong temperature inversion layer with positive humidity anomaly and that the wind profile takes a parabolic shape, with the wind speed taking its maximum around at the center of the layer. This shape seems to agree with the previous findings from measurements at a relatively deep valley while differing from those on a simple plane terrain.

The Nonhydrostatic Icosahedral Atmospheric Model (NICAM)–Local Ensemble Transform Kalman Filter (LETKF) JAXA Research Analysis (NEXRA), a weather research and analysis system integrating NICAM with the LETKF, has been updated. NEXRA combines an atmospheric data assimilation system (NICAM–LETKF) and a five-day deterministic weather forecast model using NICAM. This study compares the previous system version, NEXRA2, with the updated version, NEXRA3. NEXRA3 features a higher-resolution NICAM–LETKF with a 56 km mesh for assimilation and an updated version of the 14 km mesh NICAM with improved physical schemes and optimized source code. We performed a statistical comparison by analyzing one year of data assimilation cycles and conducting five-day deterministic forecasts under both summer and winter conditions. The results demonstrate that NEXRA3 outperforms NEXRA2 in both the NICAM–LETKF analysis and five-day forecasting, particularly in terms of precipitation forecast accuracy. These findings show that the implementation of several model improvements, which have proven highly effective in previous developments, also improve both the data assimilation and forecasting cycles.

The Sea of Japan (SOJ) has exhibited significant sea surface temperature (SST) warming in early winter, particularly in the East Korea Bay (EKB) and the subpolar oceanic front (SF). This study examined the thermodynamic impact of SST anomalies over the EKB and SF regions on the Japan-Sea polar-airmass convergence zone (JPCZ) using high-resolution numerical experiments. Results revealed that the local warm SOJ–SST anomalies played two contradictory roles in modulating the JPCZ. The anomalously warm EKB warmed the atmospheric boundary layer over the downstream region (JPCZ area) where the JPCZ prevailed, thereby decreasing sea level pressure (SLP) through hydrostatic equilibrium. The SLP decrease facilitated low-level wind convergence, intensifying the JPCZ. Enhanced moisture supplies from the sea surface due to strong winds also contributed to the JPCZ precipitation through the dominance of moisture flux convergence. In contrast, the extremely warm SF induced an anomalous surface low to the north of the JPCZ area through boundary-layer warming. Such thermodynamic changes strengthened low-level wind convergence over the SF, whereas they interrupted monsoonal winds that flowed into the JPCZ area, thereby inhibiting the JPCZ precipitation. These findings emphasize that monitoring of the local SOJ–SST warming is crucial for the accurate prediction of the JPCZ.

Subseasonal forecasting for extreme precipitation represents a critical yet challenging frontier in weather prediction, particularly in regions like Vietnam, where monsoons, tropical cyclones, and diverse topography complicate the precipitation patterns. This study explores the integration of machine learning techniques—Random Forest (RF) and Extreme Gradient Boosting (XGB)—into model output statistics to enhance subseasonal extreme rainfall forecasts across Vietnam's seven climatic regions. ECMWF S2S hindcast data for the Madden-Julian Oscillation, monsoon indices, and soil moisture are used to predict rainfall extremes. The models are trained over 2001-2014 and evaluated over 2015-2023 against observational data. Evaluation metrics, including probability of detection, false alarm ratio, critical success index, and Brier skill score, highlight the superior performance of RF and XGB over raw ECMWF forecasts, particularly in North West, North East, Red River Plain, Central North and Central South regions. However, challenges remain in the Central Highland and South regions, where both deterministic and probabilistic skills are weaker. Overall, this study underscores the potential of machine learning to address regional and temporal variability in extreme rainfall prediction, offering a transformative tool for disaster preparedness in Vietnam.