Search

Tsunami simulations require powerful computational resources to be performed efficiently. Although the modern graphics processing units (GPUs) allow the acceleration of this kind of simulations, they can still last many minutes or even hours for simulations which have to deal with very high spatial resolutions or simulation times. In this paper, we propose a method to predict the alert or inundation level of a tsunami generated by an earthquake using deep learning methods. In particular, we train multilayer perceptron (MLP) neural networks for predicting the alert level due to a tsunami at given coastal locations. Ensemble methods are used to improve the predictions of the neural networks. Tsunamis caused by ruptures of several fault segments at different time instants, application to real events, probabilistic forecasting and comparison with other machine learning algorithms are also addressed. Results on realistic scenarios confirm that good accuracies are obtained. The inference times of the trained networks and ensembles are also very low, lasting less than one second to predict the results of thousands of simulations. The proposed method could be used in a tsunami early warning system along with the application of scaling laws. Plain Language Summary: In tsunami early warning systems (TEWS) it is crucial to perform fast predictions of the risk of a tsunami in coastal areas in order to take appropriate measures before the arrival of the tsunami. Currently, tsunami simulations may take many minutes or even hours despite using modern computational resources. In this work we propose a method based on deep learning techniques to predict the alert level of a tsunami generated by an earthquake at given coastal locations in a fraction of a second. This method could be used in TEWS along with techniques to estimate the values of some parameters of the earthquake. Application to real events and probabilistic predictions are also addressed. Key Points: A deep learning method to predict different tsunami alert levels is proposedTsunamis caused by the rupture of multiple fault segments at different time instants are supportedApplication to real events with probabilistic predictions are addressed [ABSTRACT FROM AUTHOR]

Satellite infrared detectors cannot penetrate clouds, especially precipitating clouds. Improving precipitation estimation accuracy based on infrared brightness temperature has always been important but challenging. In this paper, based on the infrared brightness temperature of the Advanced Geosynchronous Radiation Imager (AGRI) onboard China's Feng‐Yun 4A satellite, we develop and evaluate a new precipitation estimation method. First, using static data, physical characteristics of clouds, cloud image texture features, temporal motion features, and AGRI infrared channel brightness temperature, we construct features for a machine learning model. Then, we develop precipitation estimation methods. Precipitation is estimated in two steps: classification and regression. We employ a random forest classification model to identify whether there is precipitation in a given field of view. If there is precipitation, a multi‐model ensemble regression learning method is used to estimate the areas with this precipitation. The ensemble learning method uses convex optimization to integrate prediction results based on the optimization of hyperparameters of five basic models (i.e., those of random forest, XGBoost, LightGBM, decision tree, and extra tree models). Furthermore, two regression stacking ensemble models—the Least Absolute Shrinkage and Selection Operator (herein referred to as Stacking1‐LASSO) and K‐nearest neighbor (herein referred to as Stacking2‐KNN)—are used to predict the results of the aforementioned basic models. The results of basic models are used as inputs of these two stacking models. Finally, based on the Integrated Multi‐satellitE Retrievals for GPM (IMERG) precipitation product and rain gauge precipitation data, we conduct precipitation estimation experiments and evaluate our methods. The results show that ensemble learning models have greater accuracy in estimating precipitation than the basic models. When using IMERG precipitation as the target precipitation, ensemble learning models can estimate the central area of heavy precipitation during typhoons Ampil and Maria. The ensemble learning estimation effect is better than that of Stacking2‐KNN. Moreover, when rain gauge data is used as the target precipitation, ensemble learning can also estimate the center of heavy precipitation and with good consistency with recorded satellite brightness temperature data. Key Points: A convex optimization weighted and stacked ensemble learning method for satellite data estimation of precipitation is developedNear real‐time precipitation estimates can be obtained throughout the day using only geographic information and Advanced Geosynchronous Radiation Imager infrared brightness temperature dataThe proposed ensemble learning method can estimate the central areas of typhoons and short‐term heavy rainfall [ABSTRACT FROM AUTHOR]

Economic growth, air pollution, and forest fires in some states in the United States have increased the concentration of particulate matter with a diameter less than or equal to 2.5 μm (PM2.5). Although previous studies have tried to observe PM2.5 both spatially and temporally using aerosol remote sensing and geostatistical estimation, they were limited in accuracy by coarse resolution. In this paper, the performance of machine learning models on predicting PM2.5 is assessed with linear regression (LR), decision tree (DT), gradient boosting regression (GBR), AdaBoost regression (ABR), XGBoost (XGB), k‐nearest neighbors (K‐NN), long short‐term memory (LSTM), random forest (RF), and support vector machine (SVM) using PM2.5 station data from 2017 to 2021. To compare the accuracy of all the nine machine learning models, the coefficient of determination (R2), root mean square error (RMSE), Nash‐Sutcliffe efficiency (NSE), root mean square error ratio (RSR), and percent bias (PBIAS) were evaluated. Among all nine models, the RF (100 decision trees with a max depth of 20) and support vector regression (SVR; nonlinear kernel, degree 3 polynomial) models were the best for predicting PM2.5 concentrations. Additionally, comparison of the PM2.5 performance metrics displayed that the models had better predictive behavior in the western United States than that in the eastern United States. Key Points: We present the prediction of PM2.5 concentrations over the United States using various machine learning (ML) algorithmsWe show ML as a new approach for analyzing large data sets due to the computational speed and easy implementation for massive amounts of dataThe study is important for improving our understanding of the differences among ML algorithms for Earth Science research [ABSTRACT FROM AUTHOR]

Nowcasting is a term originating from economics, finance, and meteorology. It refers to the process of determining the uncertain state of the economy, markets or the weather at the current time by indirect means. In this paper, we describe a simple two‐parameter data analysis that reveals hidden order in otherwise seemingly chaotic earthquake seismicity. One of these parameters relates to a mechanism of seismic quiescence arising from the physics of strain‐hardening of the crust prior to major events. We observe an earthquake cycle associated with major earthquakes in California, similar to what has long been postulated. An estimate of the earthquake hazard revealed by this state variable time series can be optimized by the use of machine learning in the form of the Receiver Operating Characteristic skill score. The ROC skill is used here as a loss function in a supervised learning mode. Our analysis is conducted in the region of 5° × 5° in latitude‐longitude centered on Los Angeles, a region which we used in previous papers to build similar time series using more involved methods (Rundle & Donnellan, 2020, https://doi.org/10.1029/2020EA001097; Rundle, Donnellan et al., 2021, https://doi.org/10.1029/2021EA001757; Rundle, Stein et al., 2021, https://doi.org/10.1088/1361-6633/abf893). Here we show that not only does the state variable time series have forecast skill, the associated spatial probability densities have skill as well. In addition, use of the standard ROC and Precision (PPV) metrics allow probabilities of current earthquake hazard to be defined in a simple, straightforward, and rigorous way. Plain Language Summary: Earthquake nowcasting refers to the determination of hazard for major earthquakes at the present time, the recent past, and the near future. Nowcasting is an idea borrowed from economics, markets, and meteorology, where it has been frequently used. In this paper, we show that there is order hidden within chaotic earthquake seismicity using a very simple transformation of the data. Small earthquakes appear to transition from unstable stick‐slip events that produce seismic waves, to stable sliding where no seismic waves are produced. Our hypothesis is that this transition is due to a material phenomenon called strain‐hardening, that is frequently observed in laboratory rock mechanics experiments. The result is a state variable time series, computed over the last 51 years in California, that strongly resembles the long‐anticipated cycle of stress accumulation and release. Using supervised machine learning techniques, we can optimize the two‐parameter model. From that optimized model, we can rigorously calculate the probability of current hazard from major earthquakes. Extending these methods, we can also compute spatial hazard as well. The result is a new method for assessing earthquake hazard that may be useful for a variety of applications. Key Points: "Chaotic" seismicity contains hidden structure in the form of state variable time seriesStandard data science methods can be used to convert the time series to probabilitiesBoth temporal and spatial probabilities can be computed [ABSTRACT FROM AUTHOR]

We propose a new machine learning‐based method for nowcasting earthquakes to image the time‐dependent earthquake cycle. The result is a timeseries that may correspond to the process of stress accumulation and release. The timeseries are constructed by using principal component analysis of regional seismicity. The patterns are found as eigenvectors of the cross‐correlation matrix of a collection of seismicity timeseries in a coarse grained regional spatial grid (pattern recognition via unsupervised machine learning). The eigenvalues of this matrix represent the relative importance of the various eigenpatterns. Using the eigenvectors and eigenvalues, we compute the weighted correlation timeseries of the regional seismicity. This timeseries has the property that the weighted correlation generally decreases prior to major earthquakes in the region, and increases suddenly just after a major earthquake occurs. As in a previous paper (Rundle & Donnellan, 2020, https://doi.org/10.1029/2020ea001097), we find that this method produces a nowcasting timeseries that resembles the hypothesized regional stress accumulation and release process characterizing the earthquake cycle. We then address the problem of whether the timeseries contain information regarding future large earthquakes. For this, we compute a receiver operating characteristic and determine the decision thresholds for several future time periods of interest (optimization via supervised machine learning). We find that signals can be detected that can be used to characterize the information content of the timeseries. These signals may be useful in assessing present and near‐future seismic hazards. Plain Language Summary: Major earthquakes on fault systems in a tectonically active region are thought to occur in approximately repetitive cycles as a result of the buildup and release of tectonic forces (stress). Nowcasting is a technique adopted from weather, finance, and other fields that use readily observable proxy data to represent the unobservable stress accumulation process of interest. This paper presents a method that computes a timeseries representing the weighted correlation of small earthquake activity in the California region from 1950 to 2020. Prior to major magnitude M > 7 earthquakes, the timeseries trends toward lower values. Just after the earthquake occurs, the timeseries increases suddenly in association with the earthquake, before resuming its gradual trend toward lower values. Plotting the timeseries on an inverted scale, one sees a cyclic behavior that strongly resembles the hypothesized earthquake cycle. In principle, we can therefore use this timeseries for nowcasting, as a proxy for stress accumulation and release. Using methods of signal detection first developed for radar by the British in the 1940's, we find that the timeseries contain information about future large earthquakes that can be used for hazard assessment. Key Points: The current state of the earthquake cycle of tectonic stress accumulation and release is unobservable with existing methodsWe show that readily observable small earthquake correlations can be used to nowcast the current state of the earthquake cycleMachine learning techniques indicate that signals corresponding to future large earthquakes can be detected in a correlation time series [ABSTRACT FROM AUTHOR]

Characterizing the location of the outer boundary of the outer radiation belt is a key aspect of improving radiation belt models and helps to constrain our understanding of the mechanisms by which the source and seed electron populations are transported into the radiation belts. In this paper, we hypothesize that there are statistical differences in the electron distribution function across the radiation belt outer boundary, and thus analyze electron flux data from the THEMIS (Time History of Events and Macroscale Interactions during Substorms) satellites to identify this location. We validate our hypothesis by using modeled electron L* values to approximately characterize the differences between electron distribution functions inside and outside of the radiation belts. Initially, we perform a simple statistical analysis by studying the radial evolution of the electron distribution functions. This approach does not yield a clear discontinuity, thus highlighting the need for more complex statistical treatment of the data. Subsequently, we employ machine learning (with no dependence on radial position or L*) to test a range of candidate outer boundary locations. By analyzing the performance of the models at each candidate location, we identify a statistical boundary at ≈8 RE, with results suggesting some variability. This statistical boundary is typically further out than those used in current radiation belt models. Plain Language Summary: Earth's magnetic field traps highly energetic particles in a donut shaped region, referred to as "the radiation belts". Our work focuses on the outer belt, comprised of electrons. Many spacecraft orbit within this region, exposing them to potential damage. To mitigate this, the radiation belts must be understood and modeled. The outer boundary is crucial to modeling, driving changes in radiation belt activity. The boundary is also important because its location helps us to understand which processes form the radiation belts. In this paper, we analyze electron data measured by satellites to identify the location of the radiation belt's outer boundary by using simple statistical methods and machine learning. Our results show that simple statistical methods cannot be used to deduce an outer boundary. Using machine learning, we test many candidate boundary locations and by quantifying the model performances at each of these locations, we are able to identify a statistical boundary location. This boundary is located at approximately eight Earth radii away from the planet, which is typically further out than the boundaries currently used by radiation belt models, although our analysis suggests the boundary location may be variable. Key Points: Simple statistical analysis of Time History of Events and Macroscale Interactions during Substorms data does not yield a boundary locationMachine learning is used to identify a boundary location using electron distribution functionsThe boundary location is found to be significantly beyond geosynchronous orbit [ABSTRACT FROM AUTHOR]

In most data assimilation and numerical weather prediction systems, the Gaussian assumption is prevalent for the behavior of the random variables/errors that are involved. At the Cooperative Institute for Research in the Atmosphere theory has been developed for different forms of variational data assimilation schemes that enables the Gaussian assumption to be relaxed. For certain variable types, a lognormally distributed random variable can be combined in a mixed Gaussian‐lognormal distribution to better capture the interactions of the errors of different distributions. However, assuming that a distribution can change in time, then developing techniques to know when to switch between different versions of the data assimilation schemes becomes very important. By dynamically changing the formulation of the data assimilation system we are able to assimilate observations in a way that reflects the flow‐dependent variability of their distribution. In this paper, we present results with a machine learning technique (the support vector machine) to switch between data assimilation methods based on the detection of a change in the Lorenz 1963 model's z component's probability distribution. Given the machine learning technique's detection/prediction of a change in the distribution, then either a Gaussian or a mixed Gaussian‐lognormal 3DVar based cost function is used to minimize the errors in this period of time. It is shown that switching from a Gaussian 3DVar to a lognormal 3DVar at lognormally distributed parts of the attractor improves the data assimilation analysis error compared to using one distribution type for the entire attractor. Key Points: Machine learning can be used to detect non‐Gaussian distributionsGiven the machine learning detection, we can switch between optimal data assimilation schemesData assimilation can be improved using machine learning techniques [ABSTRACT FROM AUTHOR]

Nowadays, S‐band weather radars are designed to observe weather conditions within large area. However, S‐band radars suffer from large blind zone at low elevation and fixed scan strategy so that they are not efficient for surveillance of convective weathers such as tornado. Networked X‐band weather radars are thereby proposed to overcome this issue. One key problem of networked radar is to automatically determine the type and the precise position of convective cells that are worthwhile to detect. In this paper, a tailor‐made framework is proposed to automatically find the convective cells and retrieve information of them from reflectivity product of networked X‐band weather radars. The framework consists of three substeps: convection pixel retrieval by Back Propagation Neural Network (BPNN), convection cell construction by Density‐Based Spatial Clustering of Applications with Noise (DBSCAN), and convection cell classification by Convolutional Neural Network (CNN). Evaluation results show that the proposed algorithm is capable to identify isolated single‐cell and multicell convective storms from the reflectivity image accurately. Therefore, the proposed framework is capable to provide information for networked X‐band weather radars so that they can track harmful convective weathers. The proposed framework has been embedded in the meteorological command and control (MCC) center of networked X‐band radar system in Chengdu. Plain Language Summary: Weather radar is significant in modern meteorological observation system. Recently, networked X‐band weather radars are attractive because it provides an efficient way to detect and track harmful convective weather such as tornadoes. One of the most difficult tasks of networked X‐band weather radars is to automatically determine clouds that are possible to endanger our regular live. It is not an easy mission because clouds are notoriously changeable, and the time resolution of our weather radars is relative low. In this paper, we propose an algorithm to challenge this problem by utilizing machine learning methods. According to the evaluation results, our algorithm is accurate and computational efficient, and it has been embedded into the networked X‐band radar system in Chengdu, China. Key Points: A framework of region‐of‐interest detection and classification is proposed for networked X‐band weather radar systemReal‐time machine learning‐based algorithms are introduced to achieve accurate identification with simple inputThe framework is subjectively and objectively evaluated in a networked X‐band radar system located in Chengdu, China [ABSTRACT FROM AUTHOR]

Massive unstructured geoscience data are buried in geological reports. Geological text classification provides opportunities to leverage this wealth of data for geology and mineralization research. Existing studies of massive geoscience documents/reports have not provided effective classification results for further knowledge discovery and data mining and often lack adequate domain‐specific knowledge. In this paper, we present a novel and unified framework (namely, Dic‐Att‐BiLSTM) that combines domain‐specific knowledge and bidirectional long short‐term memory (BiLSTM) for effective geological text classification. Dic‐Att‐BiLSTM benefits from a matching strategy by incorporating domain‐specific knowledge developed based on geoscience ontology to grasp the linguistic geoscience clues. Furthermore, Dic‐Att‐BiLSTM brings together the capacity of a geoscience dictionary matching approach and an attention mechanism to construct a dictionary attention layer. Finally, the network framework of Dic‐Att‐BiLSTM can utilize domain‐specific knowledge and classify geological text automatically. Experimental verifications are conducted on two constructed data sets, and the results clearly indicate that Dic‐Att‐BiLSTM outperforms other state‐of‐the‐art text classification models. Plain Language Summary: Several existing research efforts use technical methods/models to improve the performance of their text classification (TC), but the performance is limited by the nature of the TC categories. In this paper, a dictionary‐guided bidirectional long short‐term memory (BiLSTM) neural network algorithm that incorporates both a geoscience dictionary and a document‐level attention mechanism into BiLSTM for automatic TC from geoscience reports is proposed. We hope that our approach will serve as an alternative method that deserves further study. Key Points: An effective text filter method is proposed for reducing unnecessary computational processing of irrelevant geological textA novel and unified framework that combines domain‐specific knowledge BiLSTM for effective geological text classification is proposedA training data set is constructed for evaluating the experimental performance [ABSTRACT FROM AUTHOR]

Classifying images using supervised machine learning (ML) relies on labeled training data—classes or text descriptions, for example, associated with each image. Data‐driven models are only as good as the data used for training, and this points to the importance of high‐quality labeled data for developing a ML model that has predictive skill. Labeling data is typically a time‐consuming, manual process. Here, we investigate the process of labeling data, with a specific focus on coastal aerial imagery captured in the wake of hurricanes that affected the Atlantic and Gulf Coasts of the United States. The imagery data set is a rich observational record of storm impacts and coastal change, but the imagery requires labeling to render that information accessible. We created an online interface that served labelers a stream of images and a fixed set of questions. A total of 1,600 images were labeled by at least two or as many as seven coastal scientists. We used the resulting data set to investigate interrater agreement: the extent to which labelers labeled each image similarly. Interrater agreement scores, assessed with percent agreement and Krippendorff's alpha, are higher when the questions posed to labelers are relatively simple, when the labelers are provided with a user manual, and when images are smaller. Experiments in interrater agreement point toward the benefit of multiple labelers for understanding the uncertainty in labeling data for machine learning research. Plain Language Summary: After hurricanes and storms, pictures taken from a plane can be used to observe how the coast was impacted. A single flight might take thousands of pictures. If a computer could automatically analyze the pictures, then a person would not need to look at them one‐by‐one. To teach a computer to analyze images, we need many pictures and many labels that describe what is visible in each picture. But where do we get those labels? Typically, a coastal scientist labels the pictures by sorting them into folders or typing codes into a spreadsheet. But does every coastal scientist label pictures the same way? Some labeling questions are easy to answer, and scientists mostly agree ("Is this image all water?"). Other labeling questions are harder to answer, and cause disagreement ("Was there damage to buildings?"). This paper is about how well scientists agree when labeling the same pictures, and how we can improve agreement among scientists. We try some experiments and offer a few ideas on how to improve agreement. We suggest writing very clear questions, using smaller images, and having a comprehensive manual. It turns out that having a manual with examples—and reading the manual!—really helps. Key Points: We measure agreement among coastal scientists labeling the same sets of poststorm imagesCoastal scientists agree more when rating landforms, less when labeling inferred processesIterating on questions, providing documentation, and using smaller image sizes all increase agreement [ABSTRACT FROM AUTHOR]

Airborne dust has broad adverse effects on human activity, including aviation, human health, and agriculture. Remote sensing observations are used to detect dust and aerosols in the atmosphere using long established techniques. False color Red‐Green‐Blue (RGB) imagery using band differences sensitive to dust absorption (Dust RGB) is currently used operationally to assist forecasters and decision‐makers in identifying dust at night, but there are still limitations, subjectivity, and nuances to image interpretation making night‐time dust identification difficult even for experts. This study applies machine learning to the problem of night‐time dust detection with a simple random forest (RF) model using Geostationary Operational Environmental Satellite‐16 (GOES‐16) Advanced Baseline Imager (ABI) infrared imagery, band differences sensitive to dust absorption, and Dust RGB color components as inputs to the model. The RF model achieves an Area‐Under‐Curve (AUC) of 0.97 with a standard deviation of 0.04 for dust cases. For images with dust present, the model correctly labels 85% of dust pixels and 99.96% of no‐dust pixels for all dust images in the validation data set. The addition of a single null case to the training data set drastically reduces error in labeling no‐dust pixels as dust from 45% to 14.5%. Application of the machine learning model to the April 13–14, 2019 dust event demonstrates the ability of the model to identify dust during night‐time hours when visual dust detection is limited by the cooling ground surface characteristics. Plain Language Summary: Airborne dust adversely affects many facets of human activity, including aviation, human health, and agriculture. While many observational systems can detect airborne dust during the day, nighttime observations are limited, making airborne dust difficult to identify and track at night. This paper combines satellite infrared imagery with machine learning to enable improved dust detection at night when visual dust detection is hindered. The relatively simple model correctly labels 85% of dust pixels in the satellite imagery, supporting operational forecasters in airborne dust identification and hazard detection at night. Key Points: Detection of hazardous dust events in nighttime satellite imagery is limited as dust is difficult to distinguish from the cooling groundA physically based approach to developing a machine learning model with satellite imagery inputs correctly labels 85% of dust pixelsApplication of the machine learning model to a dust event enhances identification of dust, distinguishing the boundaries of the plume [ABSTRACT FROM AUTHOR]

A variety of detailed data about geological topics and geoscience knowledge are buried in the geoscience literature and rarely used. Named entity recognition (NER) provides both opportunities and challenges to leverage this wealth of data in the geoscience literature for data analysis and further information extraction. Existing NER models and techniques are mainly based on rule‐based and supervised approaches, and developing such systems requires a costly manual effort. In this paper, we first design a generic stepwise framework for domain‐specific NER. Following this framework, domain‐specific entities and domain‐general words are collected and selected as seed terms. Normalization and grouping processes are then applied to these seed terms for further analysis. A random extraction algorithm based on a unigram language model is used to generate a large‐scale training data set consisting of probabilistically labeled pseudosentences. Each generated sentence is then used as input to the self‐training and learning algorithm. Experimental results on two constructed data sets demonstrate that the proposed model effectively recognizes and identifies geological named entities. Plain Language Summary: Existing entity recognition and classification methods are less functional for automatic geological name entity recognition. In this paper, we propose a stepwise unsupervised approach to geological name entity recognition in geological reports that requires no labeled data. This approach dynamically adapts to extract and identify unseen instances. We hope that our approach will serve as an alternative method that deserves further study. Key Points: Geological named entities are extracted from unstructured Chinese geoscience reports using deep learningThe proposed framework can be easily extended to other subject domains through fine‐tuningA training data set is constructed based on both domain‐specific entities and domain‐general words [ABSTRACT FROM AUTHOR]

Seismic bursts in Southern California are sequences of small earthquakes strongly clustered in space and time and include seismic swarms and aftershock sequences. A readily observable property of these events, the radius of gyration (RG), allows us to connect the bursts to the temporal occurrence of the largest M ≥ 7 earthquakes in California since 1984. In the Southern California earthquake catalog, we identify hundreds of these potentially coherent space‐time structures in a region defined by a circle of radius 600 km around Los Angeles. We compute RG for each cluster then filter them to identify those bursts with large numbers of events closely clustered in space, which we call "compact" bursts. Our basic assumption is that these compact bursts reflect the dynamics associated with large earthquakes. Once we have filtered the burst catalog, we apply an exponential moving average to construct a time series for the Southern California region. We observe that the RG of these bursts systematically decreases prior to large earthquakes, in a process that we might term "radial localization." The RG then rapidly increases during an aftershock sequence, and a new cycle of "radial localization" then begins. These time series display cycles of recharge and discharge reminiscent of seismic stress accumulation and release in the elastic rebound process. The complex burst dynamics we observe are evidently a property of the region as a whole, rather than being associated with individual faults. This new method allows us to improve earthquake nowcasting, which is a technique to evaluate the current state of hazard in a seismically active region. Plain Language Summary: Earthquake nowcasting is a method to evaluate the current state of seismic hazard from large earthquakes. In this paper, we connect the temporal occurrence of the largest and most potentially destructive earthquakes in California since 1984 with a readily observable property of small earthquake seismicity in the region. Our method involves the calculation of the time history of the average radius (horizontal size or extent) of "bursts" of small earthquakes, in the time leading up to and following major earthquakes in the region. We observe that the radius systematically and gradually decreases leading up to major earthquakes, increasing suddenly and discontinuously following the event. This observable pattern resembles the long‐hypothesized cycle of regional tectonic stress buildup and release, or elastic rebound, associated with large destructive earthquakes. We propose that the radius of these bursts might be considered to be a proxy variable for the changing state of regional stress in Southern California. Key Points: A new nowcasting method using earthquake swarms and aftershocks reveals the hazard from large M ≥ 7 earthquakes in Southern CaliforniaMachine learning defines a time series of bursts using an exponential moving average filtered and optimized to reveal temporal structureCycles of recharge and discharge associated with M ≥ 7 earthquakes are revealed, reminiscent of stress accumulation and release [ABSTRACT FROM AUTHOR]

Automatic waveform classification and arrival picking methods are widely studied to reduce or replace the manual works. Machine learning based methods, especially neural networks, and clustering based methods have shown great potentials in previous studies. However, most of the existing methods are sensitive to noise. The convolution neural networks (CNNs), developed from the traditional neural networks, have been successfully applied in many different fields, but are rarely studied in seismic waveform classification. In this paper, we propose a novel antinoise CNN architecture for waveform classification and also propose to combine k‐means clustering (KC) with CNN classification to pick arrivals (CNN‐KC). Seismic data are sampled to 1‐D vectors using a specific time window. Using the trained CNN classifier, these 1‐D vectors are classified into two categories: waveform and nonwaveform. With the constraint of the first waveform, CNN‐KC can pick the arrival more accurately. We also apply the proposed methods to the synthetic microseismic data with different noise levels and the actual field microseismic data to test their robustness. CNNs perform much better than the traditional multilayer perceptron on the waveform classification of the noisy microseismic data. Based on the analysis of the CNN internal architecture, we also conclude that the main reason that CNN is insensitive to noise is the convolution and pooling layers which behave like smooth operation in some ways. The final results show that the CNN and CNN‐KC are effective and robust methods for waveform classification and arrival picking. Key Points: We introduce the concept of waveform classification, by which we can automatically pick the arrivals and potentially reject noiseWe propose a novel CNN architecture for waveform classification and also combine k‐means clustering with CNN to pick arrivalsCNNs perform much better than the traditional multilayer perceptron (MLP) on the waveform classification of the noisy microseismic data [ABSTRACT FROM AUTHOR]