Mountainous terrains are typical over southeast China, with complex and diverse topography, large terrain undulations, rich geographic features, and meteorological variations. Previous studies show that ERA5 meteorological variables are generally accurate with respect to large plains or urban agglomerations, while their applicability to mountainous areas remains inconclusive. In this paper, using high-precision measurements probed by ground-based remote sensing instruments in May–July 2023 at a typical mountainous Shanghuang site in southeast China, the vertical accuracy of the ERA5 reanalysis datasets were comparatively evaluated. Our findings depict that the horizontal wind speeds of the ERA5 reanalysis data show a good performance compared to the Doppler lidar observations. In quantitative terms, ERA5 horizontal wind speeds are about 8% higher than the observed values below a height of 400 m, while above 400 m, an increasing negative bias is observed along as altitude increases. Differing from the horizontal wind speeds, there is a large discrepancy in the vertical wind speeds between the ERA5 and the observations, with a deviation of −150% to 40%. In terms of the thermal variables, the temperature extracted from ERA5 are consistent with the measurements in the low troposphere. Nevertheless, large systematic errors occur at 2000–3000 m, and the overall presentation shows that the errors gradually increase with the increase in altitude. Concerning the relative humidity, the general trend in ERA5 is similar to that observed by the microwave radiometer, but the relative errors from 500 to 2500 m range from 40% to 100%. This study also reveals that ERA5 is poorly representative and requires further improvements during extreme weather events such as rainstorms and typhoons. In particular, the horizontal wind speeds at the middle and lower levels deviate strongly from the observations. Given the importance of atmospheric thermodynamic stratifications in terms of both environmental and climatic issues, the results expand the application of the ERA5 reanalysis datasets in the mountainous areas of southeast China. More importantly, it provides credible reference data for the meteorological predictions and climate modelings in the southeast China mountainous region.

This study provides a comprehensive evaluation of the vertical accuracy of ERA5 reanalysis data for boundary layer height and key meteorological variables, based on high-precision observational data from Baotou, located on the Mongolian Plateau, during the winter (January–March) and summer (July–August) months of 2021. Results indicate that ERA5 exhibits significant biases in horizontal wind speed, with deviations ranging from −5 to 8 m/s at 50 m, primarily driven by sandstorms in winter and convective weather in summer. The most pronounced errors occur below 500 m. Vertical wind speeds are consistently underestimated in both seasons, with biases reaching up to 1 m/s, particularly during active summer convection. ERA5 also struggles to reproduce low-level wind directions accurately. In winter, correlation coefficients range from 0.43 to 0.64 below 200 m and improve to above 0.7 at 500 m. In summer, correlation coefficients are lower, ranging from 0.3 to 0.5 below 200 m, with reduced accuracy at 500 m compared to winter. Temperature deviations increase above 2000 m, with a relative overestimation of 3% at 3000 m. Relative humidity is generally overestimated by 5–20% between 1000 and 2000 m in winter and by 10–30% in summer. For boundary layer heights, ERA5 overestimates daytime mixed-layer heights by up to 2000 m in summer and 500–800 m in winter. In contrast, ERA5 captures nocturnal stable boundary layer heights well during winter. This comprehensive evaluation of the vertical structure accuracy of ERA5 reanalysis data, conducted in a heavily industrialized city on the Mongolian Plateau, offers essential insights for improving meteorological studies and refining climate models in the region. The findings provide valuable reference data for enhancing weather forecasting and supporting climate change research, particularly in complex terrain areas.

The accuracy of ERA5 reanalysis datasets and their applicability in the coastal area of Bohai Bay are crucial for weather forecasting and environmental protection research. However, synthesis evaluation of ERA5 in this region remains lacking. In this study, using a tropospheric wind profile radar (CFL-06L) placed in coastal Huanghua city, the deviations of ERA5 reanalysis data are assessed from the ground to an altitude of 5 km. The results indicate that the wind speed of ERA5 reanalysis data exhibits good consistency from the surface to the tropospheric level of about 5 km, with R² values ranging from 0.5 to 0.85. The lowest mean wind speed error, less than 3 m/s, occurs in the middle layer, while larger errors are observed at the surface and upper layers. Specifically, at 150 m, the R² is as low as 0.5, with numerous outliers around 5000 m. Seasonal analysis shows that the ERA5 wind field performs best in summer and worst in autumn and winter, especially at lower levels affected by circulation systems, high stratus clouds, and aerosols, with errors reaching up to 10 m/s. Further analysis of extreme weather events, such as heavy rain; hot, dry winds; and snowstorms, reveals that the effects of sea-land winds and strong convective systems significantly impact the observation of wind profiles and the assimilation of reanalysis data, particularly under the constrain of boundary layer height. Additionally, we also find that the transition of sea-land breeze is capable of triggering the nighttime low-level jet, thereby downward transporting the aloft ozone to the ground and resulting in an abnormal increase in the surface ozone concentration. The study provides a scientific basis for improving meteorological forecasting, optimizing wind energy resource utilization, and formulating environmental protection policies, highlighting its significant scientific and practical application value.

The variation of polar vortex intensity is a significant factor affecting the atmospheric conditions and weather in the Northern Hemisphere (NH) and even the world. However, previous studies on the prediction of polar vortex intensity are insufficient. This paper establishes a deep learning (DL) model for multi-day and long-time intensity prediction of the polar vortex. Focusing on the winter period with the strongest polar vortex intensity, geopotential height (GPH) data of NCEP from 1948 to 2020 at 50 hPa are used to construct the dataset of polar vortex anomaly distribution images and polar vortex intensity time series. Then, we propose a new convolution neural network with long short-term memory based on Gaussian smoothing (GSCNN-LSTM) model which can not only accurately predict the variation characteristics of polar vortex intensity from day to day, but also can produce a skillful forecast for lead times of up to 20 days. Moreover, the innovative GSCNN-LSTM model has better stability and skillful correlation prediction than the traditional and some advanced spatiotemporal sequence prediction models. The accuracy of the model suggests important implications that DL methods have good applicability in forecasting the nonlinear system and vortex spatial–temporal characteristics variation in the atmosphere.

The Gradient Richardson Number (R_i) is an important parameter for appraising the stability and turbulence exchange at the atmospheric boundary layer (ABL). However, high-resolution measurements of R_i profiles are rarely reported, especially in megacities. In this study, a Doppler wind lidar and a microwave radiometer were simultaneously utilized to measure the 2 km R_i vertical profile in downtown Beijing. These measurements were verified to have high accuracy compared with observations from a 325 m meteorological tower, with root-mean-square errors (RMSEs) of less than 1.66 K, 7.9%, and 1.45 m/s for the temperature, relative humidity, and wind speed (WS) for all altitudes and corresponding Pearson correlation coefficients (R) of 0.97, 0.93, and 0.81. The inter-comparisons of different spatial (25 m, 50 m, 100 m) and temporal resolutions (1 min, 30 min, 1 h) form a 3 × 3 resolution matrix of R_i, in which the 1 h temporal resolution of R_i overestimates the intensity and active area of turbulence. The R_i value retrieved from the 100 m spatial resolution data overestimates these by half as it misidentifies the height of the stable area at the near surface. There are significant differences between the data with a 1 min temporal resolution and a 25 m spatial resolution (defined as the standard resolution of R_i), and the rest of the data in the resolution matrix (defined as data at other resolutions), with an RMSE > 1 and an R < 0.8. The difference between data at the standard resolution and data at other resolutions increases with elevations, which results from frequent weather processes or from water-vapor blocking at higher altitudes. The R_i profiles reveal that the atmospheric layer at altitudes from 100 m to 500 m in daytime is unstable, with R_i < 0, while it is neutral, with 0 < R_i < 0.25, at night-time from 200 m to 400 m. The atmosphere above the ABL in a megacity is rather stable, with R_i > 0.25, whereas below the ABL, it is neutral or unstable, which is due to drastic changes in the WS and temperature that are affected by the topography and surface friction.

In July 2021, an extreme precipitation event occurred in Henan, China, causing tremendous damage and deaths; so, it is very important to study the observation technology of extreme precipitation. Surface rain gauge precipitation observations have high accuracy but low resolution and coverage. Satellite remote sensing has high spatial resolution and wide coverage, but has large precipitation accuracy and distribution errors. Therefore, how to merge the above two kinds of precipitation observations effectively to obtain heavy precipitation products with more accurate geographic distributions has become an important but difficult scientific problem. In this paper, a new information fusion method for improving the position accuracy of satellite precipitation estimations is used based on the idea of registration and warping in image processing. The key point is constructing a loss function that includes a term for measuring two information field differences and a term for a warping field constraint. By minimizing the loss function, the purpose of position error correction of quantitative precipitation estimation from FY-4A and Integrated Multisatellite Retrievals of GPM are achieved, respectively, using observations from surface rain gauge stations. The errors of different satellite precipitation products relative to ground stations are compared and analyzed before and after position correction, using the ‘720’ extreme precipitation in Henan, China, as an example. The experimental results show that the final run has the best performance and FY-4A has the worse performance. After position corrections, the precipitation products of the three satellites are improved, among which FY-4A has the largest improvement, IMERG final run has the smallest improvement, and IMERG late run has the best performance and the smallest error. Their mean absolute errors are reduced by 23%, 14%, and 16%, respectively, and their correlation coefficients with rain gauge stations are improved by 63%, 9%, and 16%, respectively. The error decomposition model is used to examine the contributions of each error component to the total error. The results show that the new method improves the precipitation products of GPM primarily in terms of hit bias. However, it does not significantly reduce the hit bias of precipitation products of FY-4A while it reduces the total error by reducing the number of false alarms.