Explore the vast range of titles available.

Category 4 landfalling hurricane Harvey poured more than a metre of rainfall across the heavily populated Houston area, leading to unprecedented flooding and damage. Although studies have focused on the contribution of anthropogenic climate change to this extreme rainfall event 1 – 3 , limited attention has been paid to the potential effects of urbanization on the hydrometeorology associated with hurricane Harvey. Here we find that urbanization exacerbated not only the flood response but also the storm total rainfall. Using the Weather Research and Forecast model—a numerical model for simulating weather and climate at regional scales—and statistical models, we quantify the contribution of urbanization to rainfall and flooding. Overall, we find that the probability of such extreme flood events across the studied basins increased on average by about 21 times in the period 25–30 August 2017 because of urbanization. The effect of urbanization on storm-induced extreme precipitation and flooding should be more explicitly included in global climate models, and this study highlights its importance when assessing the future risk of such extreme events in highly urbanized coastal areas. Modelling the contribution of urbanization to the impacts associated with hurricane Harvey in August 2017 shows that urbanization worsens rainfall and flooding.

Tropical cyclones have always been a concern of meteorologists, and there are many studies regarding the axisymmetric structures, dynamic mechanisms, and forecasting techniques from the past 100 years. This research demonstrates the ongoing progress as well as the many remaining problems. Machine learning, as a means of artificial intelligence, has been certified by many researchers as being able to provide a new way to solve the bottlenecks of tropical cyclone forecasts, whether using a pure data-driven model or improving numerical models by incorporating machine learning. Through summarizing and analyzing the challenges of tropical cyclone forecasts in recent years and successful cases of machine learning methods in these aspects, this review introduces progress based on machine learning in genesis forecasts, track forecasts, intensity forecasts, extreme weather forecasts associated with tropical cyclones (such as strong winds and rainstorms, and their disastrous impacts), and storm surge forecasts, as well as in improving numerical forecast models. All of these can be regarded as both an opportunity and a challenge. The opportunity is that at present, the potential of machine learning has not been completely exploited, and a large amount of multi-source data have also not been fully utilized to improve the accuracy of tropical cyclone forecasting. The challenge is that the predictable period and stability of tropical cyclone prediction can be difficult to guarantee, because tropical cyclones are different from normal weather phenomena and oceanographic processes and they have complex dynamic mechanisms and are easily influenced by many factors.

Current climate models often have significant wet biases in the Tibetan Plateau and encounter particular difficulties in representing the climatic effect of the Central Himalaya Mountain (CHM), where the gradient of elevation is extremely steep and the terrain is complex. Yet, there were few studies dealing with the issue in the high altitudes of this region. In order to improve climate modeling in this region, a network consisting of 14 rain gauges was set up at elevations > 2800 m above sea level along a CHM valley. Numerical experiments with Weather Research and Forecasting model were conducted to investigate the effects of meso- and micro-scale terrain on water vapor transport and precipitation. The control case uses a high horizontal resolution (0.03°) and a Turbulent Orographic Form Drag (TOFD) scheme to resolve the mesoscale terrain and to represent sub-grid microscale terrain effect. The effects of the horizontal resolution and the TOFD scheme were then analyzed through comparisons with sensitivity cases that either use a low horizontal resolution (0.09°) or switch off the TOFD scheme. The results show that the simulations with high horizontal resolution, even without the TOFD scheme, can not only increase the spatial consistency (correlation coefficient 0.84–0.92) between the observed and simulated precipitation, but also considerably reduce the wet bias by more than 250%. Adding the TOFD scheme further reduces the precipitation bias by 50% or so at almost all stations in the CHM. The TOFD scheme reduces precipitation intensity, especially heavy precipitation (> 10 mm h−1) over high altitudes of the CHM. Both high horizontal resolution and TOFD enhance the orographic drag to slow down wind; as a result, less water vapor is transported from lowland to the high altitudes of CHM, causing more precipitation at lowland area of the CHM and less at high altitudes of CHM. Therefore, in this highly terrain-complex region, it is crucial to use a high horizontal resolution to depict mesoscale complex terrain and a TOFD scheme to parameterize the drag caused by microscale complex terrain.

Temperature forecasting plays an important role in human production and operational activities. Traditional temperature forecasting mainly relies on numerical forecasting models to operate, which takes a long time and has higher requirements for the computing power and storage capacity of computers. In order to reduce computation time and improve forecast accuracy, deep learning-based temperature forecasting has received more and more attention. Based on the atmospheric temperature, dew point temperature, relative humidity, air pressure, and cumulative wind speed data of five cities in China from 2010 to 2015 in the UCI database, multivariate time series atmospheric temperature forecast models based on recurrent neural networks (RNN) are established. Firstly, the temperature forecast modeling of five cities in China is established by RNN for five different model configurations; secondly, the neural network training process is controlled by using the Ridge Regularizer (L2) to avoid overfitting and underfitting; and finally, the Bayesian optimization method is used to adjust the hyper-parameters such as network nodes, regularization parameters, and batch size to obtain better model performance. The experimental results show that the atmospheric temperature prediction error based on LSTM RNN obtained a minimum error compared to using the base models, and these five models obtained are the best models for atmospheric temperature prediction in the corresponding cities. In addition, the feature selection method is applied to the established models, resulting in simplified models with higher prediction accuracy.

Over the course of his career, Fuqing Zhang drew vital new insights into the dynamics of meteorologically significant mesoscale gravity waves (MGWs), including their generation by unbalanced jet streaks, their interaction with fronts and organized precipitation, and their importance in midlatitude weather and predictability. Zhang was the first to deeply examine “spontaneous balance adjustment”—the process by which MGWs are continuously emitted as baroclinic growth drives the upper-level flow out of balance. Through his pioneering numerical model investigation of the large-amplitude MGW event of 4 January 1994, he additionally demonstrated the critical role of MGW–moist convection interaction in wave amplification. Zhang’s curiosity-turned-passion in atmospheric science covered a vast range of topics and led to the birth of new branches of research in mesoscale meteorology and numerical weather prediction. Yet, it was his earliest studies into midlatitude MGWs and their significant impacts on hazardous weather that first inspired him. Such MGWs serve as the focus of this review, wherein we seek to pay tribute to his groundbreaking contributions, review our current understanding, and highlight critical open science issues. Chief among such issues is the nature of MGW amplification through feedback with moist convection, which continues to elude a complete understanding. The pressing nature of this subject is underscored by the continued failure of operational numerical forecast models to adequately predict most large-amplitude MGW events. Further research into such issues therefore presents a valuable opportunity to improve the understanding and forecasting of this high-impact weather phenomenon, and in turn, to preserve the spirit of Zhang’s dedication to this subject.

Data assimilation (DA) is integrated with machine learning in order to perform entirely data‐driven online state estimation. To achieve this, recurrent neural networks (RNNs) are implemented as pretrained surrogate models to replace key components of the DA cycle in numerical weather prediction (NWP), including the conventional numerical forecast model, the forecast error covariance matrix, and the tangent linear and adjoint models. It is shown how these RNNs can be initialized using DA methods to directly update the hidden/reservoir state with observations of the target system. The results indicate that these techniques can be applied to estimate the state of a system for the repeated initialization of short‐term forecasts, even in the absence of a traditional numerical forecast model. Further, it is demonstrated how these integrated RNN‐DA methods can scale to higher dimensions by applying domain localization and parallelization, providing a path for practical applications in NWP. Plain Language Summary Weather forecast models derived from fundamental equations of physics continue to increase in detail and complexity. While this evolution leads to consistently improving daily weather forecasts, it also leads to associated increases in computational costs. In order to make a forecast at any given moment, these models must be initialized with our best guess of the current state of the atmosphere, which typically includes information from a limited set of observations as well as forecasts from the recent past. Modern methods for initializing these computer forecasts typically require running many copies of the model, either simultaneously or in sequence, to compare with observations over the recent past and ensure that our best guess estimate of the current state of the atmosphere agrees closely with those observations before making a new forecast. This repeated execution of the computer forecast model is often a time‐consuming and costly bottleneck in the initialization process. Here, it is shown that techniques from the fields of artificial intelligence and machine learning (AI/ML) can be used to produce simple surrogate models that provide sufficiently accurate approximations to replace the original costly model in the initialization phase. The resulting process is self‐contained, and does not require any further utilization of the original computer model when making daily forecasts. Key Points Recurrent neural networks (RNNs) can replace conventional forecast models, producing accurate ensemble forecast statistics and linearized dynamics Data assimilation (DA) is compatible with RNNs by applying state estimation in the hidden state space using a modified observation operator The integrated RNN‐DA methods can be scaled to higher dimensions by applying domain localization techniques

The Bureau of Meteorology Atmospheric high-resolution Regional Reanalysis for Australia (BARRA) is the first atmospheric regional reanalysis over a large region covering Australia, New Zealand, and Southeast Asia. The production of the reanalysis with approximately 12 km horizontal resolution – BARRA-R – is well underway with completion expected in 2019. This paper describes the numerical weather forecast model, the data assimilation methods, the forcing and observational data used to produce BARRA-R, and analyses results from the 2003–2016 reanalysis. BARRA-R provides a realistic depiction of the meteorology at and near the surface over land as diagnosed by temperature, wind speed, surface pressure, and precipitation. Comparing against the global reanalyses ERA-Interim and MERRA-2, BARRA-R scores lower root mean square errors when evaluated against (point-scale) 2 m temperature, 10 m wind speed, and surface pressure observations. It also shows reduced biases in daily 2 m temperature maximum and minimum at 5 km resolution and a higher frequency of very heavy precipitation days at 5 and 25 km resolution when compared to gridded satellite and gauge analyses. Some issues with BARRA-R are also identified: biases in 10 m wind, lower precipitation than observed over the tropical oceans, and higher precipitation over regions with higher elevations in south Asia and New Zealand. Some of these issues could be improved through dynamical downscaling of BARRA-R fields using convective-scale (<2 km) models.

A key component of the 4D-Var data assimilation method used widely for numerical weather prediction is the linear forecast model, which is approximately tangent linear to the forecast model. Traditionally this has been based on differentiating the forecast model, though recently some authors have experimented with an ensemble regression technique, the localized ensemble tangent linear model (LETLM). We propose a hybrid of the two, in which a simplified conventional tangent-linear model (e.g., just the dynamical core) is used together with an LETLM-like adjustment every time step to account for the remaining processes (in this example, the parameterized physics). This is much cheaper than the LETLM, and in tests using the Met Office’s linear model performs considerably better than either a pure LETLM (with a very large ensemble) or the existing linear model.

Understanding the physical processes that affect the turbulent structure of the nocturnal urban boundary layer (UBL) is essential for improving forecasts of air quality and the air temperature in urban areas. Low-level jets (LLJs) have been shown to affect turbulence in the nocturnal UBL. We investigate the interaction of a mesoscale LLJ with the UBL during a 60-h case study. We use observations from two Doppler lidars and results from two high-resolution numerical-weather-prediction models (Weather Research and Forecasting model, and the Met Office Unified Model for limited-area forecasts for the U.K.) to study differences in the occurrence frequency, height, wind speed, and fall-off of LLJs between an urban (London, U.K.) and a rural (Chilbolton, U.K.) site. The LLJs are elevated (≈ 70 m) over London, due to the deeper UBL, while the wind speed and fall-off are slightly reduced with respect to the rural LLJ. Utilizing two idealized experiments in the WRF model, we find that topography strongly affects LLJ characteristics, but there is still a substantial urban influence. Finally, we find that the increase in wind shear under the LLJ enhances the shear production of turbulent kinetic energy and helps to maintain the vertical mixing in the nocturnal UBL.

The sources of predictability for the February 2021 cold air outbreak (CAO) over the central United States, which led to power grid failures and water delivery shortages in Texas, are diagnosed using a machine learning–based prediction model called a linear inverse model (LIM). The flexibility and low computational cost of the LIM allows its forecasts to be used for identifying and assessing the predictability of key physical processes. The LIM may also be run as a climate model for sensitivity and risk analysis for the same reasons. The February 2021 CAO was a subseasonal forecast of opportunity, as the LIM confidently predicted the CAO’s onset and duration four weeks in advance, up to two weeks earlier than other initialized numerical forecast models. The LIM shows that the February 2021 CAO was principally caused by unpredictable internal atmospheric variability and predictable La Niña teleconnections, with nominally predictable contributions from the previous month’s sudden stratospheric warming and the Madden–Julian oscillation. When run as a climate model, the LIM estimates that the February 2021 CAO was in the top 1% of CAO severity and suggests that similarly extreme CAOs could be expected to occur approximately every 20–30 years.