Weather Radar Echoes Research Articles

TrajPredRNN+: A new approach for precipitation nowcasting with weather radar echo images based on deep learning

：Short-term rainfall prediction is a crucial and practical research area, with the accuracy of rainfall prediction, particularly for heavy rainfall, significantly impacting people's lives, property, and even their safety. Existing models, such as ConvLSTM, TrajGRU, and PredRNN, exhibit limitations in capturing fine-grained appearances due to insufficient memory units or addressing positional misalignment issues, thereby compromising the accuracy of model predictions. In this study, we propose trajPredRNN+, an innovative approach that integrates the trajectory segmentation model and the PredRNN deep learning model to address both limitations in nowcasting precipitation using weather radar echo images. By incorporating attention mechanisms, the model demonstrates an enhanced focus on short-term and imminent heavy rainfall events. To ensure improved stability during training, a residual network is introduced. Lastly, a more rational and effective training loss function is proposed, encompassing weight mechanism, SSIM index, and GAN loss. To validate the proposed model, we conducted a comparative experiment and an ablation experiment using the radar echo map dataset obtained from the Shenzhen Meteorological Bureau. The results of these experiments demonstrate that our model has achieved significant improvements across multiple key performance indicators.

Classifying Microwave Radiometer Observations over the Netherlands into Dry, Shallow, and Nonshallow Precipitation Using a Random Forest Model

Abstract Spaceborne microwave radiometers represent an important component of the Global Precipitation Measurement (GPM) mission due to their frequent sampling of rain systems. Microwave radiometers measure microwave radiation (brightness temperatures Tb), which can be converted into precipitation estimates with appropriate assumptions. However, detecting shallow precipitation systems using spaceborne radiometers is challenging, especially over land, as their weak signals are hard to differentiate from those associated with dry conditions. This study uses a random forest (RF) model to classify microwave radiometer observations as dry, shallow, or nonshallow over the Netherlands—a region with varying surface conditions and frequent occurrence of shallow precipitation. The RF model is trained on five years of data (2016–20) and tested with two independent years (2015 and 2021). The observations are classified using ground-based weather radar echo top heights. Various RF models are assessed, such as using only GPM Microwave Imager (GMI) Tb values as input features or including spatially aligned ERA5 2-m temperature and freezing level reanalysis and/or Dual-Frequency Precipitation Radar (DPR) observations. Independent of the input features, the model performs best in summer and worst in winter. The model classifies observations from high-frequency channels (≥85 GHz) with lower Tb values as nonshallow, higher values as dry, and those in between as shallow. Misclassified footprints exhibit radiometric characteristics corresponding to their assigned class. Case studies reveal dry observations misclassified as shallow are associated with lower Tb values, likely resulting from the presence of ice particles in nonprecipitating clouds. Shallow footprints misclassified as dry are likely related to the absence of ice particles. Significance Statement Published research concerning rainfall retrieval algorithms from microwave radiometers is often focused on the accuracy of these algorithms. While shallow precipitation over land is often characterized as problematic in these studies, little progress has been made with these systems. In particular, precipitation formed by shallow clouds, where shallow refers to the clouds being close to Earth’s surface, is often missed. This study is focused on detecting shallow precipitation and its physical characteristics to further improve its detection from spaceborne sensors. As such, it contributes to understanding which shallow precipitation scenes are challenging to detect from microwave radiometers, suggesting possible ways for algorithm improvement.

DEUCE v1.0: a neural network for probabilistic precipitation nowcasting with aleatoric and epistemic uncertainties

Abstract. Precipitation nowcasting (forecasting locally for 0–6 h) serves both public security and industries, facilitating the mitigation of losses incurred due to, e.g., flash floods and is usually done by predicting weather radar echoes, which provide better performance than numerical weather prediction (NWP) at that scale. Probabilistic nowcasts are especially useful as they provide a desirable framework for operational decision-making. Many extrapolation-based statistical nowcasting methods exist, but they all suffer from a limited ability to capture the nonlinear growth and decay of precipitation, leading to a recent paradigm shift towards deep-learning methods which are more capable of representing these patterns. Despite its potential advantages, the application of deep learning in probabilistic nowcasting has only recently started to be explored. Here we develop a novel probabilistic precipitation nowcasting method, based on Bayesian neural networks with variational inference and the U-Net architecture, named DEUCE. The method estimates the total predictive uncertainty in the precipitation by combining estimates of the epistemic (knowledge-related and reducible) and heteroscedastic aleatoric (data-dependent and irreducible) uncertainties, using them to produce an ensemble of development scenarios for the following 60 min. DEUCE is trained and verified using Finnish Meteorological Institute radar composites compared to established classical models. Our model is found to produce both skillful and reliable probabilistic nowcasts based on various evaluation criteria. It improves the receiver operating characteristic (ROC) area under the curve scores 1 %–5 % over STEPS and LINDA-P baselines and comes close to the best-performer STEPS on a continuous ranked probability score (CRPS) metric. The reliability of DEUCE is demonstrated with, e.g., having the lowest expected calibration error at 20 and 25 dBZ reflectivity thresholds and coming second at 35 dBZ. On the other hand, the deterministic performance of ensemble means is found to be worse than that of extrapolation and LINDA-D baselines. Last, the composition of the predictive uncertainty is analyzed and described, with the conclusion that aleatoric uncertainty is more significant and informative than epistemic uncertainty in the DEUCE model.

Quantification of Migration Birds Based on Polarimetric Weather Radar

Weather radar plays an important role in monitoring aerial animal migration, providing a stable data source for biological studies with large-scale coverage and consecutive-time samples. The accurate estimation of bird density from weather radar echoes is fundamental for quantitative biological studies. We analyzed the bird observation model in weather radar, and proposed a method to build the bird quantification model by jointly utilizing dual-polarization Doppler weather radar and scanning bird radar. We designed a detailed process to remove tracks or echoes from non-bird targets, ensuring the effectiveness of bird observations. The field experiments validated the quantification method, showing that the average radar cross section of birds in Jinan was 19.09 dBscm (i.e., 81.19 cm2; 95% confidence interval, CI: 18.92–19.27 dBscm) for the S-band weather radar, with an R2 of 0.79. In addition, through the correlation analysis, we found that the ground terrain may affect the distribution pattern of aerial bird density.

Multi-Scale Feature Residual Feedback Network for Super-Resolution Reconstruction of the Vertical Structure of the Radar Echo

The vertical structure of radar echo is crucial for understanding complex microphysical processes of clouds and precipitation, and for providing essential data support for the study of low-level wind shear and turbulence formation, evolution, and dissipation. Therefore, finding methods to improve the vertical data resolution of the existing radar network is crucial. Existing algorithms for improving image resolution usually focus on increasing the width and height of images. However, improving the vertical data resolution of weather radar requires a focus on improving the elevation angle resolution while maintaining distance resolution. To address this challenge, we propose a network for super-resolution reconstruction of weather radar echo vertical structures. The network is based on a multi-scale residual feedback network (MR-FBN) and uses new multi-scale feature residual blocks (MSRB) to effectively extract and utilize data features at different scales. The feedback network gradually generates the final high-resolution vertical structure data. In addition, we propose an elevation upsampling layer (EUL) specifically for this task, replacing the traditional image subpixel convolution layer. Experimental results show that the proposed method can effectively improve the elevation angle resolution of weather radar echo vertical structure data, providing valuable help for atmospheric detection.

Nowcasting Multiparameter Phased-Array Weather Radar (MP-PAWR) Echoes of Localized Heavy Precipitation Using a 3D Recurrent Neural Network Trained with an Adversarial Technique

Abstract We present nowcasts of sudden heavy rains on meso-γ scales (2–20 km) using the high spatiotemporal resolution of a multiparameter phased-array weather radar (MP-PAWR) sensitive to rain droplets. The onset of typical storms is successfully predicted with 10-min lead time, i.e., the current predictability limit of rainfall caused by individual convective cores. A supervised recurrent neural network based on long short-term memory with 3D spatial convolutions (RN3D) is used to account for the horizontal and vertical changes of the convective cells with a time resolution of 30 s. The model uses radar reflectivity at horizontal polarization ZH and the differential reflectivity. The input parameters are defined in a volume of 64 × 64 × 8 km3 with the lowest level at 1.9 km and a resolution of 0.4 × 0.4 × 0.25 km3. The prediction is a 10-min sequence of ZH at the lowest grid level. The model is trained with a large number of observations of summer 2020 and an adversarial technique. RN3D is tested with different types of rapidly evolving localized heavy rainfalls of summers 2018 and 2019. The model performance is compared to that of an advection model for 3D extrapolation of PAWR echoes (A3DM). RN3D better predicts the formation and dissipation of precipitation. However, RN3D tends to underestimate heavy rainfall especially when the storm is well developed. In this phase of the storm, A3DM nowcast scores are found slightly higher. The high skill of RN3D to predict the onset of sudden localized rainfall is illustrated with an example for which RN3D outperforms the operational precipitation nowcasting system of Japan Meteorological Agency (JMA). Significance Statement Temporal extrapolation of radar observations is a means of nowcasting sudden heavy rains, i.e., forecasts with a few tens of minutes and a high spatial resolution better than 500 m. They are necessary to set up warning systems to anticipate damage to infrastructure and reduce the fatalities these storms cause. It is a difficult task due to the storm suddenness, restricted area, and nonlinear behavior that are not well captured by current operational observation and numerical systems. In this study, we use a new high-resolution weather radar with polarimetric information and a 3D recurrent neural network to improve 10-min nowcasts, the current limit of operational systems. This is a first and essential step before applying such a method for increasing the prediction lead time.

Weather Radar Echo Extrapolation with Dynamic Weight Loss

Precipitation nowcasting is an important tool for economic and social services, especially for forecasting severe weather. The crucial and challenging part of radar echo image prediction is the focus of radar-based precipitation nowcasting. Recently, a number of deep learning models have been designed to solve the problem of extrapolating radar images. Although these methods can generate better results than traditional extrapolation methods, the issue of error accumulation in precipitation forecasting is exacerbated by using only the mean square error (MSE) and mean absolute error (MAE) as loss functions. In this paper, we approach the problem from the perspective of the loss function and propose dynamic weight loss (DWL), a simple but effective loss function for radar echo extrapolation. The method adds model self-adjusted dynamic weights to the weighted loss function and structural similarity index measures. Radar echo extrapolation experiments are performed on four models, ConvLSTM, ConvGRU, PredRNN, and PredRNN++. Radar reflectivity is predicted using Nanjing University C-band Polarimetric (NJU-CPOL) weather radar data. The quantitative statistics show that using the DWL method reduces the MAE of the four models by up to 10.61%, 5.31%, 14.8%, and 13.63%, respectively, over a 1 h prediction period. The results show that the DWL approach is effective in reducing the accumulation of errors over time, improving the predictive performance of currently popular deep learning models.

A Nonlinear Grid Transformation Method for Extrapolating and Predicting the Convective Echo of Weather Radar

A nonlinear grid transformation (NGT) method is proposed for weather radar convective echo extrapolation prediction. The change in continuous echo images is regarded as a nonlinear transformation process of the grid. This process can be reproduced by defining and solving a 2 × 6 transformation matrix, and this approach can be applied to image prediction. In ideal experiments with numerical and path changes of the target, NGT produces a prediction result closer to the target than does a conventional optical flow (OF) method. In the presence of convection lines in real cases, NGT is superior to OF: the critical success index (CSI) for 40 dBZ of the echo prediction at 60 min is approximately 0.2 higher. This is due to the better estimation of the movement of the whole cloud system in the NGT results since it reflects the continuous change in the historical images. For the case with a mesoscale convective complex, the NGT results are better than the OF results, and a deep learning result is cited from a previous study for the same case for 20 and 30 dBZ. However, the result is the opposite for 40 dBZ, where the deep learning method may produce an overestimation of the stronger echo.

Correction: Wang et al. Characteristic Analysis of Dual-Polarization Weather Radar Echoes of Convective Precipitation and Snowfall in the Mount Everest Region. Atmosphere 2021, 12, 1671

In the published manuscript [...]

Weather Radar Echo Extrapolation Method Based on Deep Learning

In order to forecast some high intensity and rapidly changing phenomena, such as thunderstorms, heavy rain, and hail within 2 h, and reduce the influence brought by destructive weathers, this paper proposes a weather radar echo extrapolation method based on deep learning. The proposed method includes the design and combination of the data preprocessing, convolutional long short-term memory (Conv-LSTM) neuron and encoder–decoder model. We collect eleven thousand weather radar echo data in high spatiotemporal resolution, these data are then preprocessed before they enter the neural network for training to improve the data’s quality and make the training better. Next, the neuron integrates the structure and the advantages of convolutional neural network (CNN) and long short-term memory (LSTM), called Conv-LSTM, is applied to solve the problem that the full-connection LSTM (FC-LSTM) cannot extract the spatial information of input data. This operation replaced the full-connection structure in the input-to-state and state-to-state parts so that the Conv-LSTM can extract the information from other dimensions. Meanwhile, the encoder–decoder model is adopted due to the size difference of the input and output data to combine with the Conv-LSTM neuron. In the neural network training, mean square error (MSE) loss function weighted according to the rate of rainfall is added. Finally, the matrix “point-to-point” test method, including the probability of detection (POD), critical success index (CSI), false alarm ratio (FAR) and spatial test method contiguous rain areas (CRA), is used to examine the radar echo extrapolation’s results. Under the threshold of 30 dBZ, at the time of 1 h, we achieved 0.60 (POD), 0.42 (CSI) and 0.51 (FAR), compared with 0.42, 0.28 and 0.58 for the CTREC algorithm, and 0.30, 0.24 and 0.71 for the TITAN algorithm. Meanwhile, at the time of 1 h, we achieved 1.35 (total MSE ) compared with 3.26 for the CTREC algorithm and 3.05 for the TITAN algorithm. The results demonstrate that the radar echo extrapolation method based on deep learning is obviously more accurate and stable than traditional radar echo extrapolation methods in near weather forecasting.

Signal Simulation of Dual-Polarization Weather Radar and Its Application in Range Ambiguity Mitigation

In this paper, a dual-polarization weather radar echo signal simulation method is proposed for the evaluation of the performance enhancement of dual-polarization weather radar systems, the optimization of signal processing algorithms and the improvement of scanning strategies. The actual weather radar base data are used in the simulation as the reference weather scene, which avoids using a complex algorithm for weather modeling. Moreover, based on radar weather equations, the radar system parameters are added into the radar echo signal modeling to establish the relationship between the simulated echo signal and radar system. As a result, the final simulated echo signal not only shows both the time and frequency domain characteristics of the weather target, but also includes the effects of the important performance of the dual-polarization weather radar system. In addition, to evaluate the performance of range ambiguity mitigation using phase coding and batch working modes, two different simulation methods for the radar signal are established on the method above; one is for batch working mode with long-PRT (pulse repetition time) and short-PRT transmission and receiving, and the other is for phase-coded mode with phase-coded transmission and phase-uncoded receiving. Under the same weather scene, the observation of these two different methods of range ambiguity mitigation are simulated and compared. Results show that the performance of the phase coding mode for mitigating range ambiguity is better than that of the batch mode. Obviously, the simulation method can be used to directly show the observation of different algorithms for mitigation range ambiguity under the same weather process, and quickly compare and evaluate the algorithm’s performance, which is not possible for real radars.

RainPredRNN: A New Approach for Precipitation Nowcasting with Weather Radar Echo Images Based on Deep Learning

Precipitation nowcasting is one of the main tasks of weather forecasting that aims to predict rainfall events accurately, even in low-rainfall regions. It has been observed that few studies have been devoted to predicting future radar echo images in a reasonable time using the deep learning approach. In this paper, we propose a novel approach, RainPredRNN, which is the combination of the UNet segmentation model and the PredRNN_v2 deep learning model for precipitation nowcasting with weather radar echo images. By leveraging the abilities of the contracting-expansive path of the UNet model, the number of calculated operations of the RainPredRNN model is significantly reduced. This result consequently offers the benefit of reducing the processing time of the overall model while maintaining reasonable errors in the predicted images. In order to validate the proposed model, we performed experiments on real reflectivity fields collected from the Phadin weather radar station, located at Dien Bien province in Vietnam. Some credible quality metrics, such as the mean absolute error (MAE), the structural similarity index measure (SSIM), and the critical success index (CSI), were used for analyzing the performance of the model. It has been certified that the proposed model has produced improved performance, about 0.43, 0.95, and 0.94 of MAE, SSIM, and CSI, respectively, with only 30% of training time compared to the other methods.

Adaptive Sparse Domain Selection for Weather Radar Superresolution

Accurate and high-resolution weather radar data reflecting detailed structure information of radar echo plays an important role in analysis and forecast of extreme weather. Typically, this is done using interpolation schemes, which only use several neighboring data values for computational approximation to get the estimated, resulting the loss of intense echo information. Focus on this limitation, a superresolution reconstruction algorithm of weather radar data based on adaptive sparse domain selection (ASDS) is proposed in this article. First, the ASDS algorithm gets a compact dictionary by learning the precollected data of model weather radar echo patches. Second, the most relevant subdictionaries are adaptively select for each low-resolution echo patches during the spare coding. Third, two adaptive regularization terms are introduced to further improve the reconstruction effect of the edge and intense echo information of the radar echo. Experimental results show that the ASDS algorithm substantially outperforms interpolation methods for ×2 and ×4 reconstruction in terms of both visual quality and quantitative evaluation metrics.

Weather Radar Image Superresolution Using a Nonlocal Residual Network

Accurate and high-resolution weather radar images reflecting detailed structure information of radar echo are vital for analysis and forecast of extreme weather. Typically, this is performed by using interpolation schemes, which only use several neighboring data values for computational approximation to get the estimated value regardless of the large-scale context feature of weather radar images. Inspired by the striking performance of the convolutional neural network (CNN) applied in feature extraction and nonlocal self-similarity of weather radar images, we proposed a nonlocal residual network (NLRN) on the basis of CNN. The proposed network mainly consists of several nonlocal residual blocks (NLRB), which combine short skip connection (SSC) and nonlocal operation to train the deep network and capture large-scale context information. In addition, long skip connection (LSC) added in the network avoids learning low-frequency information, making the network focus on high-level features. Extensive experiments of ×2 and ×4 super-resolution reconstruction demonstrate that NLRN achieves superior performance in terms of both quantitative evaluation metrics and visual quality, especially for the reconstruction of the edge and detailed information of the weather radar echo.

Characteristic Analysis of Dual-Polarization Weather Radar Echoes of Convective Precipitation and Snowfall in the Mount Everest Region

This paper introduces the X-band weather radar dual-polarization parameters of isolated convective cell precipitation and meso/microscale snowfall on Mount Everest and presents the first precipitation observations based on dual-polarization weather radar in this area. Compared with the Chengdu Plain, Mount Everest experienced convective precipitation on smaller horizontal and vertical scales with a narrower Zdr probability density spectrum (uniformly distributed around approximately 0). The Zh profile on Mount Everest displayed two peaks, unlike that over the plains, and the precipitation at the strong convective core was denser. Furthermore, during winter snowfall on the northern slope of Mount Everest, when the boundary layer exhibited sufficient water vapor and dynamic uplift, due to the low boundary layer temperature (<0 °C), water vapor produced stratiform clouds in the middle and lower layers (approximately 1.5 km above ground level (AGL)). Water vapor condensation at 1.5–2.5 km AGL led to latent heat release, which increased the temperature of regional stratiform clouds with increasing height. Consequently, the temperature at the stratiform cloud top height (2.5 km AGL) unexpectedly exceeded 0 °C. Additionally, the −20 °C isotherm was at approximately 4 km AGL, indicating that the middle- and upper-layer atmospheric temperatures remained low. Therefore, thermal instability occurred between the stratiform cloud top height and the middle/upper atmosphere, forming convective motion. These findings confirm the occurrence of elevated winter snowfall convection above Mount Everest and may have certain reference value for retrieving raindrop size distributions, quantitatively estimating precipitation, and parameterizing cloud microphysical processes in numerical prediction models for the Qinghai-Tibetan Plateau.

Characteristic analysis of dual-polarization weather radar echoes of clear air and precipitation in the Mount Everest region

X-band dual-polarization weather radar parameters characterized by clear-air turbulence and precipitation over Mount Everest are introduced. As the X-band radar wavelength is short, data quality control is exercised over the obtained differential phase (Φdp), reflectivity (Zh), differential reflectivity (Zdr), and other polarization physical quantities. Based on the X-band dual-polarization weather radar, FY4 satellite, and EC reanalysis data, the 3D structural characteristics of clear-air turbulence, weak precipitation, and meso-microscale strong convective precipitation on the northern slope of Mount Everest from June 2019 to July 2019 are analyzed. Our results reveal that clear-air turbulence on Mount Everest is mainly wet turbulence, and Zh and Zdr are significantly greater than those in the plain area. The ground clutter in some locations exhibits low Zdr and high ρhv characteristics, opposite to those in the plain area. Weak precipitation on Mount Everest has a typical three-layer structure similar to that in the plain area, and when Zh > 25 dBZ, Zh and Zdr display a positive linear correlation. Alternatively, Mount Everest’s atmospheric environment is conducive to triggering strong convective weather, consisting of common isolated convection cells at the microscale. The echo top height and maximum precipitation intensity of strong convection weather on Mount Everest are greater than those in the plain area. However, the whole layer convective thickness is much smaller than that in the plain area. Our preliminary results reveal the raindrop size distribution, quantitative precipitation estimation, and microphysical parameterization scheme of cloud precipitation in the Mount Everest region.

The 2018 phreatic eruption at Mt. Motoshirane of Kusatsu\u2013Shirane volcano, Japan: eruption and intrusion of hydrothermal fluid observed by a borehole tiltmeter network

We estimate the mass and energy budgets for the 2018 phreatic eruption of Mt. Motoshirane on Kusatsu–Shirane volcano, Japan, based on data obtained from a network of eight tiltmeters and weather radar echoes. The tilt records can be explained by a subvertical crack model. Small craters that were formed by previous eruptions are aligned WNW–ESE, which is consistent with the strike of the crack modeled in this study. The direction of maximum compressive stress in this region is horizontal and oriented WNW–ESE, allowing fluid to intrude from depth through a crack with this orientation. Based on the crack model, hypocenter distribution, and MT resistivity structure, we infer that fluid from a hydrothermal reservoir at a depth of 2 km below Kusatsu–Shirane volcano has repeatedly ascended through a pre-existing subvertical crack. The inflation and deflation volumes during the 2018 eruption are estimated to have been 5.1 × 105 and 3.6 × 105 m3, respectively, meaning that 1.5 × 105 m3 of expanded volume formed underground. The total heat associated with the expanded volume is estimated to have been ≥ 1014 J, similar to or exceeding the annual heat released from Yugama Crater Lake of Mt. Shirane and that from the largest eruption during the past 130 year. Although the ejecta mass of the 2018 phreatic eruption was small, the eruption at Mt. Motoshirane was not negligible in terms of the energy budget of Kusatsu–Shirane volcano. A water mass of 0.1–2.0 × 107 kg was discharged as a volcanic cloud, based on weather radar echoes, which is smaller than the mass associated with the deflation. We suggest that underground water acted as a buffer against the sudden intrusion of hydrothermal fluids, absorbing some of the fluid that ascended through the crack.

Weather radar echo prediction method based on convolution neural network and Long Short-Term memory networks for sustainable e-agriculture

As one of the main reference data of weather forecast, weather radar echo image is very important to the stability of agricultural production. Different radar echo patterns represent different disastrous weather, such as hail, severe convection, and so on. Weather radar echo shape prediction can help meteorologists to judge the future changes of disastrous weather, help to avoid the harm of extreme weather to agriculture, and minimize agricultural economic losses. With the application of deep learning in the meteorological field, the deep learning method shows great potential in radar echo prediction. However, there are few research methods to predict the change of weather radar echo shape. This paper presents a radar echo prediction method based on c Convolutional Neural Networks and Long Short-Term Memory networks, which can effectively predict the shape of weather radar echo. The actual data is used to train and test our model. Experiments show that the model can accurately predict the change of echo shape. The quantitative evaluation of the model uses detection probability, false alarm rate, critical success index, and heidke skill score. The average scores predicted by this model in 1.5 h were 0.8223,0.2012,0.6812 and 0.7564, respectively, which were better than those predicted by ConvLSTM and TrajGRU models. The qualitative and quantitative results verify the effectiveness of the model, which shows that the model can be effectively applied to the actual weather forecast and improve the stability of agricultural production.

Precipitation Nowcasting with Weather Radar Images and Deep Learning in São Paulo, Brasil

Precipitation nowcasting can predict and alert for any possibility of abrupt weather changes which may cause both human and material risks. Most of the conventional nowcasting methods extrapolate weather radar echoes, but precipitation nowcasting is still a challenge, mainly due to rapid changes in meteorological systems and time required for numerical simulations. Recently video prediction deep learning (VPDL) algorithms have been applied in precipitation nowcasting. In this study, we use the VPDL PredRNN++ and sequences of radar reflectivity images to predict the future sequence of reflectivity images for up to 1-h lead time for São Paulo, Brazil. We also verify the feasibility for the continuous use of the VPDL model, providing the meteorologist with trends and forecasts in precipitation edges regardless of the weather event occurring. The results obtained confirm the potential of the VPDL model as an additional tool to assist nowcasting. Even though meteorological systems that trigger natural disasters vary by location, a general solution can contribute as a tool to assist decision-makers and consequently issue efficient alerts.

A new method based on stacked auto-encoders to identify abnormal weather radar echo images

It is not denied that real-time monitoring of radar products is an important part in actual meteorological operations. But the weather radar often brings out abnormal radar echoes due to various factors, such as climate and hardware failure. So it is of great practical significance and research value to realize automatic identification of radar anomaly products. However, the traditional algorithms to identify anomalies of weather radar echo images are not the most accurate and efficient. In order to improve the efficiency of the anomaly identification, a novel method combining the theory of classical image processing and deep learning was proposed. The proposed method mainly includes three parts: coordinate transformation, integral projection, and classification using deep learning. Furthermore, extensive experiments have been done to validate the performance of the new algorithm. The results show that the recognition rate of the proposed method can reach up to more than 95%, which can successfully achieve the goal of screening abnormal radar echo images; also, the computation speed of it is fairly satisfactory.