Atmospheric Science Letters最新文献

Persistent heavy rainfall produced by western North Pacific (WNP) tropical cyclones (TCs) can lead to widespread flooding and landslides in Asian countries. On July 2021, unprecedent rainfall amount occurred when Typhoon In-fa passed through the highly populated eastern China. While the associated synoptic features have been analyzed, the extreme characteristics and return periods of rainfall induced by In-fa remain unexplored. Analyses of rainfall data from a WNP TC database of the China Meteorological Administration (CMA) show that Typhoon In-fa not only produces record-breaking rainfall accumulations at individual surface stations, but generates unprecedent rainfall amounts for the whole area of eastern China. Quantitatively, 2, 4, 11, 24 and 55 stations are exposed to once in 200-, 100-, 50-, 20- and 10-year extreme TC rainfall accumulations, respectively, and total rainfall at 75 stations reaches a record high since 1980. Overall, the return period is up to ~481 years for the total rainfall amount accumulated in eastern China during the 1980–2019 baseline. The extremely long rainfall duration is identified as key to the torrential rains in the Yangtze River Delta before In-fa changes its direction of movement from northwestward to northeastward, while the extreme rain rate plays a dominant role in the northern areas afterwards. Probabilities of occurrence of such an unprecedented TC rainfall event have increased in most (~75%) of the eastern China during the period of 2000–2019 compared with those during 1980–1999. Our study highlights the likely increase in risk of extreme TC-induced rainfall accumulations which should be considered in disaster risk mitigation.

The precipitation in the Qilian (QMA) and Tienshan (TMA) mountain areas is one of the main sources of subsurface and surface water in northwestern China (NWC). Based on two datasets, CN05.1 and station-observed daily precipitation, we found that summer precipitation in 1979–2020 exhibited an increasing trend in NWC. The results of rotation empirical orthogonal function (REOF) analysis also separated the increased precipitation patterns in the QMA and TMA from the other REOF modes; the proportion of the precipitation of these areas to the total NWC summer precipitation substantially increased (0.12%⋅year⁻¹ and 0.03%⋅year⁻¹, respectively). According to the moisture budget, the evaporation changes in the QMA and TMA were coherently coupled with precipitation, which suggested the feedback between increasing evaporation and precipitation with the recently warming climate. The precipitation increase was larger than that of evaporation, indicating a net wetting trend in the QMA and TMA. The increase in zonal horizontal and vertical moisture advection terms contributed more to the increased precipitation in the QMA. The increase in meridional moisture advection contributed more to the increased precipitation in the TMA. We concluded comprehensive frameworks of the water vapor transport in climate change in mountain areas in NWC which aimed to contribute to the understanding of arid region hydrology.

Predicting rapid intensification (RI) of tropical cyclones (TCs) is critical in operational forecasting. Statistical schemes rely on human-driven feature extraction and predictor correlation to predict TC intensities. Deep learning provides an opportunity to further improve the prediction if data, including satellite images of TC convection and conventional environmental predictors, can be properly integrated by deep neural networks. This study shows that deep learning yields enhanced intensity and RI prediction performance by simultaneously handling the human-defined environmental/TC-related parameters and information extracted from satellite images. From operational and practical perspectives, we use an ensemble of 20 deep-learning models with different neural network designs and input combinations to predict intensity distributions at +24 h. With the intensity distribution based on the ensemble forecast, forecasters can easily predict a deterministic intensity value demanded in operations and be aware of the chance of RI and the prediction uncertainty. Compared with the operational forecasts provided for western Pacific TCs, the results of the deep learning ensemble achieve higher RI detection probabilities and lower false-alarm rates.

This study introduces a framework to extract the high-dimensional nonlinear relationships among state variables for aggregated convection. The prototype of such a framework is developed that applies the convolutional neural network models (CNN models) to retrieve the cloud characteristics from cloud-resolving model (CRM) simulations. CNN model prediction factors are hidden in the high dimensional weighted parameters in each neural network layer. Therefore, we can dig out relevant physics processes by iterating the CNN models' training process and eliminating the features with the physics explanation we can provide at a given stage. Within a few iterations, explainable nonlinear relationships among variables can be provided. We identified that the averaged cloud water path (CWP), the maximum value of CWP in each cloud, and the cloud coverage rate are essential for identifying aggregation. Furthermore, by analyzing the encoded channels of the CNN model, we found a strong relationship between aggregation, cloud peripherals, and fractal dimensions. The results suggest that the important nonlinear cloud characteristics for identifying the aggregation can be captured with the proper adjustment and limitation of the input data to the CNN models. Our framework provides a possibility that we can explore the high dimensional relationship between the physics process with the assistance of the CNN model.

Accurate knowledge on spatiotemporal characteristics of historical precipitation extremes could provide great potential guidance for preventing hydrological-related disasters caused by precipitation extremes in the future. On the basis of the fifth generation of atmospheric reanalysis precipitation data by the European Centre for Medium Range Weather Forecasts (ERA5, 0.25°, 1 hourly, 1950–2020) with high spatiotemporal resolutions, continuity and quality, this study analyzed the spatiotemporal characteristics of precipitation extremes over the Ganjiang River Basin and its surroundings during 1950–2020. The main conclusions include, but are not limited to, the following: (1) In general, precipitation extremes present increasing trends over most areas of the basin and its surroundings. For instance, areas showing upward trends of R10, SDII and PRCPTOT account for ~93.45%, ~66.36%, and ~88.18%, respectively. (2) The spatiotemporal variations of precipitation extremes over the Ganjiang River Basin and its surroundings show obvious northwest–southeast differences. For instance, precipitation extremes are increasing in the southeastern parts, but they are decreasing in the northwestern parts. (3) High-value clusters are also identified in the southeast (e.g., R10, SDII, R95P and PRCPTOT, accounting for ~20.71%, ~20.72%, ~25.88%, and ~22.56%, respectively) and low-value clusters in the northwest (e.g., Rx5day, SDII and R95P, accounting for ~18.05%, ~27.03%, and ~21.18%, respectively). (4) The spatiotemporal variations of precipitation extremes in both the southeast and northwest are quite stable. For example, regions with less than five abrupt change points of R10, SDII, and PRCPTOT account for 77.49%, 54.84%, and 81.74%, respectively.

Non-stationarity in heavy rainfall time series is often apparent in the form of trends because of long-term climate changes. We have built non-stationary (NS) models for annual maximum daily (AMP1) and 2-day precipitation (AMP2) data observed between 1984 and 2020 years by 71 stations and between 1960 and 2020 by eight stations over Thailand. The generalized extreme value (GEV) models are used. Totally, 16 time-dependent functions of the location and scale parameters of the GEV model are considered. On each station, a model is selected by using Bayesian and Akaike information criteria among these candidates. The return levels corresponding to some years are calculated and predicted for the future. The stations with the highest return levels are Trad, Samui, and Narathiwat, for both AMP1 and AMP2 data. We found some evidence of increasing (decreasing) trends in maximum precipitation for 22 (10) stations in Thailand, based on NS GEV models.

Ozonesonde measurements from Hong Kong and Naha are combined with satellite observations to investigate the vertical structure of ozone and water vapor around the tropopause during typhoon Soudelor over the northwestern Pacific in 2015. The results show that the tropopause height is decreased compared to the mean tropopause during 2000–2017 in the western Pacific mode of the Asian summer monsoon anticyclone (ASMA). The vertical transport associated with Soudelor decreases the ozone concentration by 60% in the upper troposphere. Further the equatorward transport from high latitudes around the western Pacific mode of the ASMA increases ozone concentration by 40% in the lower stratosphere. Cross-tropopause transport of water is observed above typhoon Soudelor, and water vapor to be enhanced at 80–100 hPa compared to nontyphoon regions. Dehydration is observed below the tropopause around the eye of Soudelor.

To handle the complexity of the atmospheric boundary layer (ABL) and make accurate feature detection (top height, low-level jets, inversions, etc.), a prior necessary step is to identify the type of boundary layer. This study proposes a new method to identify the boundary layer type through unsupervised classification and the synergistic use of ground-based remote sensing. Unsupervised classification is used to lighten the human supervision. The new classification was applied to a 1-day case study collected during wintertime in the Arve River valley near Chamonix–Mont-Blanc during the Passy-2015 field experiment. The ABL classification obtained from microwave radiometer and ceilometer observations (ground-based remote sensors [GBReS]) combination is compared with high-frequency radiosoundings (RS) data and the French convective scale AROME model outputs. Classifications from RS and GBReS broadly agree, demonstrating the good behavior of the method, AROME leading to different results at night. The difference of AROME is likely due to the different nature of the data (model fields are smoother and include forecasting errors). The results show the ability of unsupervised classification to segment relevant objects in the boundary layer and the benefit to use a combination of GBReS.