Radar Echo Images Research Articles

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.

SCENT: A new precipitation nowcasting method based on sparse correspondence and deep neural network

Precipitation nowcasting is an important research topic in meteorology, which relates to many aspects of people’s life and social development. Under the combined influence of resolution and corresponding timestamp, a certain nonlinear relationship is satisfied between the echo intensity and precipitation. Therefore, the short-term precipitation prediction scheme based on radar echo extrapolation has become the main research method. By analyzing the spatiotemporal characteristics of the radar echo images, we found that precipitation results are related not only to currently observed radar echo images but also to some non-image features such as wind speed and shape of cloud clusters. Inspired by optical flow method, combined with the characteristics of radar reflectivity, we propose a new method SCENT to achieve precipitation prediction. Firstly, the sparse correspondence method based on Fast feature detection and SIFT matching is used to radar echo extrapolate and complete the extraction of non-image influence features. Afterwards, an improved neural network is utilized for regression calculation to obtain the total precipitation. By comparing with existing prediction models based on deep neural network, our new method can make precipitation nowcasting more accurate.

Application of PFI-4DVar Data Assimilation Technique to Nowcasting of Numerical Model

Nowcasting is mainly based on radar echo or satellite image extrapolation method. However, the prediction ability of extrapolation method decreases with time, because this method cannot describe the physical mechanism during the occurrence, development and extinction of severe convective weather systems. Considering that the prediction ability of numerical model improves with time, the nowcasting system should be based on numerical forecast model. And appropriate data assimilation technology can be used to produce a more accurate initial field, making the integral forecast results closer to the reality. The PFI-4DVar assimilation method (four-dimensional variational technology under physical filter initialization) can filter in the process of assimilation rather than model integration, thus shortening the model spin-up time and getting a more dynamic and physically coordinated analysis field. Therefore, PFI-4DVar assimilation method not only improves model prediction results, but also makes initial field closer to observations, which is very suitable for nowcasting.Using WRF model and WRFDA assimilation system, effects of PFI-4DVar on prediction ability of numerical nowcasting are explored. Through the precipitation case in North China on 11 August 2018, prediction results in control and assimilation tests are discussed. According to ETS scores, the precipitation prediction of assimilation test is closer to the observation compared with control test. The water vapor in assimilated ground and sounding data, the dynamic field in assimilated radar radial wind data and the appropriate cumulus parameterization scheme make the amplitude of divergence in high-level and convergence in low-level in analysis field of assimilation test much stronger than those in background field, thus creating vertical motion. Moreover, the precipitation of assimilation test is mainly caused by process of cumulus.A batch test is carried out on 17 precipitation cases of North China in August 2018. It shows that PFI-4DVar can significantly improve the prediction ability for short precipitation (especially large order precipitation) and timely predict the fall area of heavy rain or rainstorm. After assimilation, ETS scores of 6-hour accumulated precipitation (greater than 25.0 mm) in batch test increase from 0.125 to 0.190, and ETS scores of 6-hour accumulated precipitation (greater than 60.0 mm) increase from 0.016 to 0.081. PFI-4DVar significantly improves the precipitation nowcasting. Calculations are reduced by selecting 12-minute assimilation time window, which greatly saves computational resources. And the time of assimilation test is shortened, ensuring the time efficiency of 6-hour forecast. Therefore, PFI-4DVar can improve and enhance the prediction ability of precipitation nowcasting.

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.

Strong Spatiotemporal Radar Echo Nowcasting Combining 3DCNN and Bi-Directional Convolutional LSTM

In order to solve the existing problems of easy spatiotemporal information loss and low forecast accuracy in traditional radar echo nowcasting, this paper proposes an encoding-forecasting model (3DCNN-BCLSTM) combining 3DCNN and bi-directional convolutional long short-term memory. The model first constructs dimensions of input data and gets 3D tensor data with spatiotemporal features, extracts local short-term spatiotemporal features of radar echoes through 3D convolution networks, then utilizes constructed bi-directional convolutional LSTM to learn global long-term spatiotemporal feature dependencies, and finally realizes the forecast of echo image changes by forecasting network. This structure can capture the spatiotemporal correlation of radar echoes in continuous motion fully and realize more accurate forecast of moving trend of short-term radar echoes within a region. The samples of radar echo images recorded by Shenzhen and Hong Kong meteorological stations are used for experiments, the results show that the critical success index (CSI) of this proposed model for eight predicted echoes reaches 0.578 when the echo threshold is 10 dBZ, the false alarm ratio (FAR) is 20% lower than convolutional LSTM network (ConvLSTM), and the mean square error (MSE) is 16% lower than the real-time optical flow by variational method (ROVER), which outperforms the current state-of-the-art radar echo nowcasting methods.

Complex sequential understanding through the awareness of spatial and temporal concepts

Understanding sequential information is a fundamental task for artificial intelligence. Current neural networks attempt to learn spatial and temporal information as a whole, limited their abilities to represent large scale spatial representations over long-range sequences. Here, we introduce a new modeling strategy called Semi-Coupled Structure (SCS), which consists of deep neural networks that decouple the complex spatial and temporal concepts learning. Semi-Coupled Structure can learn to implicitly separate input information into independent parts and process these parts respectively. Experiments demonstrate that a Semi-Coupled Structure can successfully annotate the outline of an object in images sequentially and perform video action recognition. For sequence-to-sequence problems, a Semi-Coupled Structure can predict future meteorological radar echo images based on observed images. Taken together, our results demonstrate that a Semi-Coupled Structure has the capacity to improve the performance of LSTM-like models on large scale sequential tasks.

Feature Extraction of Radar Echo Image Based on Improved Convolutional Neural Network

Feature extraction of radar echo image is an important part of identification and recognition of air targets. In recent years, with the rapid development of deep learning, new solutions are provided for it. In this paper, the convolution neural network (CNN) is applied to feature extraction. Based on the classic CNN model, Adam is used to update the model parameters, dropout is used to prevent over fitting, and an improved CNN model is constructed. Then, the radar echo image data set is used to train the model, so as to extract target features and classify them. Simulation results show that the accuracy of the improved model is 99%, and the training speed is greatly improved. Different from the traditional extraction method which relies on manual experience, the improved CNN can improve the efficiency of feature extraction of radar echo image and lay a solid foundation for further research and identification of air targets.

Generative Adversarial Networks Capabilities for Super-Resolution Reconstruction of Weather Radar Echo Images

Improving the resolution of degraded radar echo images of weather radar systems can aid severe weather forecasting and disaster prevention. Previous approaches to this problem include classical super-resolution (SR) algorithms such as iterative back-projection (IBP) and a recent nonlocal self-similarity sparse representation (NSSR) that exploits the data redundancy of radar echo data, etc. However, since radar echoes tend to have rich edge information and contour textures, the textural detail in the reconstructed echoes of traditional approaches is typically absent. Inspired by the recent advances of faster and deeper neural networks, especially the generative adversarial networks (GAN), which are capable of pushing SR solutions to the natural image manifold, we propose using GAN to tackle the problem of weather radar echo super-resolution to achieve better reconstruction performance (measured in peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM)). Using authentic weather radar echo data, we present the experimental results and compare its reconstruction performance with the above-mentioned methods. The experimental results showed that the GAN-based method is capable of generating perceptually superior solutions while achieving higher PSNR/SSIM results.

Integrate Weather Radar and Monitoring Devices for Urban Flooding Surveillance.

With the increase of extreme weather events, the frequency and severity of urban flood events in the world are increasing drastically. Therefore, this study develops ARMT (automatic combined ground weather radar and CCTV (Closed Circuit Television System) images for real-time flood monitoring), which integrates real-time ground radar echo images and automatically estimates a rainfall hotspot according to the cloud intensity. Furthermore, ARMT combines CCTV image capturing, analysis, and Fourier processing, identification, water level estimation, and data transmission to provide real-time warning information. Furthermore, the hydrograph data can serve as references for relevant disaster prevention, and response personnel may take advantage of them and make judgements based on them. The ARMT was tested through historical data input, which showed its reliability to be between 83% to 92%. In addition, when applied to real-time monitoring and analysis (e.g., typhoon), it had a reliability of 79% to 93%. With the technology providing information about both images and quantified water levels in flood monitoring, decision makers can quickly better understand the on-site situation so as to make an evacuation decision before the flood disaster occurs as well as discuss appropriate mitigation measures after the disaster to reduce the adverse effects that flooding poses on urban areas.

Convolutional Neural Networks Applied on Weather Radar Echo Extrapolation

Extrapolation technique of weather radar echo possesses a widely application prospects in short-term nowcast. The traditional methods of radar echo extrapolation are difficult to obtain long limitation period and lacking in utilization rate of radar data. To solve this problem, this paper proposes a method of weather radar echo extrapolation based on convolutional neural networks (CNNs). In order to adapt the strong correlation between weather radar echo images of contiguous time, on the basis of traditional convolutional neural networks, this method present a new CNNs model, namely, Recurrent Dynamic Convolutional Neural Networks (RDCNN). RDCNN consists of recurrent dynamic sub-network and probability prediction layer, and constructs a cyclic structure in the convolution layer, which improve the ability of RDCNN to process time-related images. In the experiments of radar data from Nanjing, Hangzhuo and Xiamen, compared with traditional methods, this method has achieved higher accurate of extrapolation and extended the limitation period effectively, which meets the requirement for application.

Ground penetrating radar numerical simulation with interpolating wavelet scales method and research on fourth-order Runge-Kutta auxiliary differential equation perfectly matched layer

Ground penetrating radar (GPR) forward is one of the geophysical research directions.Through the forward of geological model,the database of radar model can be enriched and the characteristics of typical geological radar echo images can be understood,which in turn can guide the data interpretation of GPR measured profile,thereby improving the GPR data interpretation level.In this article,the interpolating wavelet scale function by using iterative interpolation method is presented,and the derivative of scale function is used in spatial differentiation of discrete Maxwell equations. The forward modeling formula of GPR based on the interpolation wavelet scale method is derived by using fourth-order Runge-Kutta method (RK4) for calculating the higher time derivative.Compared with the conventional finite difference time domain (FDTD) algorithm based on the central difference method,the interpolation wavelet scale algorithm improves the accuracy of GPR wave equation in both space and time discretization.Firstly,the FDTD algorithm and the interpolation wavelet scale method are applied to the forward modeling of a layered model with analytic solution. Single channel radar data and analytical solution fitting indicate that the interpolation wavelet scale method has higher accuracy than FDTD,with the same mesh generation used.Therefore,auxiliary differential equation perfectly matching layer (ADE-PML) boundary condition is used on an interpolation wavelet scale,and the comparisons between reflection errors obtained using CPML (FDTD),RK4ADE-PML (FDTD),and RK4ADE-PML (interpolating wavelet scales) in a homogeneous medium model show that the absorption effect of RK4ADE-PML (interpolating wavelet scales) is better than the other two absorbing boundaries.Finally,interpolation wavelet scale method,with both UPML,FDTD and RK4ADE-PML loaded,is used for two-dimensional GPR forward modeling,showing good absorption effect for evanescent wave.From all the experimental results,the following conclusions are obtained.1) Using the derivative of the interpolating wavelet scale function instead of central difference schemes for the spatial derivative discretization of Maxwell equations and time derivative calculated using the fourth-order Runge Kutta algorithm,the interpolating wavelet scale algorithm has higher accuracy than regular FDTD algorithm due to the improvement in the spatial and time accuracy of GPR wave equation.2) The best absorption layer parameters of interpolating wavelet scale RK4ADE-PML are selected, when the maximum value of the reflection error is the minimum.The maximum reflection error can reach-93 dB,which increases 20 dB compared with that of UMPL boundary in FDTD algorithm.And the higher simulation accuracy of interpolating wavelet scale algorithm than FDTD algorithm is confirmed after calculating single channel radar data.3) Comparing wave field snapshots of GPR forward modeling,radar pictures from wide-angle method and section method indicates that interpolating wavelet scale RK4ADE-PML reduces reflection error of absorption boundary,improves both spatial and time accuracy,is more effective than UPML boundary in eliminating false reflection of large angle incidence, and has better absorption effect for evanescent wave and low-frequency wave.

Preliminary Analysis on the Wind Retrieval Method in Nowcasting

Extrapolation nowcasting of radar echoes is implemented based on reflectivity data and radial velocity data of Doppler weather radars. VAD algorithm is used to retrieve wind fields, which are used as motion vectors to implement the extrapolation for the reflectivity data within 60 mins. Extrapolation nowcasting results of a weather case by the method agree well with the fact according to radar echo images and statistics. However, nowcasting accuracy decreases as the extrapolation time increases.

Study on the Radar Echo Dynamic Image Simulation

The characteristics of Dalian Maritime University, Transas and Furuno radar simulators are described. The current radar target echo simulation technologies are researched, such as radio echo modeling, radar target echo image simulation and rendering. Finally, to put forward a radar 2D simulation method, including terrene rendering, rain/snow rendering, sea rendering and target ship rendering.

A Method of Building Vertical Wind Shear Feature and Convective Storms Recognition Based on This Feature

There is a very close relation between severe convective storms and vertical wind shear. Therefore, it is particularly important to extract the vertical wind shear feature used to recognize storms and distinguish the type of storms such as rainstorm, hail with small or large scale. In meteorology, Doppler radar echo images are used to detect storms, which are reflectivity images(RI) and radial velocity images(RVI). This paper delineates probable storms firstly in RI with low elevation angles, then raises gradually the elevation angle to calculate radial wind shear in RVI point by point and generate a feature called the Vertical Shear Factor(VSF). A lot of testing show that the VSF's difference between rainstorm and large hail is very marked and only by this feature the radio of distinguishing rainstorms from large hails can be increased closely 5∼10 times than reflectivity strength usually used by weathermen.

Analysing and Processing of Rainfall Observation Datain Wave Monitoring System by Nautical Radar

In process of observing ocean wave with nautical radar,the roughness of offing will be changed for emergence of rainfall.So the intensity of echo signal which received from radar will be influenced.In this paper an analyzing and processing method is proposed for rainfall observation data in wave monitoring system by nautical radar.Firstly the statistic characteristic-mean value and variation coefficient of radar echo image are used as classifying criterion to distinguish rainfall and non-rainfall.Then the median filtering algorithm is approached to dealing with rainfall observation data in order to reduce the influence from rainfall.The experimental results indicate that the algorithm proposed in this paper has a good robustness and general performance.

Feature extraction and recognition based on banana-radar echo images

It is difficult for the high-way green channel inspection department to detect prohibited goods in the vehicle by artificial methods.In order to solve this problem,this paper put forward a discernment and classification method based on Back-Propagation(BP) neural network,making use of gray level co-occurrence matrix to extract feature from collected radar echo image with banana.Furthermore,the software for discernment and classification of radar echo image with banana was programmed.Practical application in the high-way green channel inspection department in Henan province shows that the software has good discernment and classification performance.

Analysis of Lightning Characteristics in a Thunderstorm with Gust

One severe thunderstorm with damaging straight-line wind occurred on 21 June 2005 in north Shandong Province. Based on the analysis of lightning activities, radar echo and cloud image, it was found that the cloud-to-ground (CG) flash rates increased rapidly at the initial stage, and kept in a high level (about above 20 fl/5min) during the whole mature stage. At the dissipating stage, positive CG flashes gradually became dominant. For the advent of the peak value, the hourly flash rate lagged behind the minimum brightness temperature, and the area of cold cloud shield with temperature <−50°C lagged behind the hourly flash rate. The pulse of CG flash rate was slightly ahead of the occurrence of severe surface wind. Downburst caused the damaging wind with the maximum measured value of wind speed approximately the same to the maximum potential speed calculated by WINDEX.

Wave Run-Up Analyses Under Dissipative Condition Using X-Band Radar

An X-band nautical radar system was employed to examine wave run-up during a typhoon around the research pier HORS in Hasaki, Japan. Analyses on radar echo images were done to estimate the spatio-temporal variation of water fronts by manually digitizing cross-shore time stack images. Digitized instantaneous water fronts were validated with wave gauge measurements with an acceptable agreement. Longshore distribution of mean shoreline positions and inter-tidal foreshore slopes were then estimated using time-averaged images. Run-up, the height of discrete water level maxima, was estimated from the digitized water fronts with the help of foreshore profile. Run-up variations under dissipative condition were parameterized with surf similarity parameter. Low frequency variances in the run-up motion were observed, which travelled in the longshore direction.

Longshore Characteristics of Wave Run-up under Stormy Condition

An X-band nautical radar system was employed to examine wave run-up during a typhoon around the research pier HORS in Hasaki, Japan. Analyses on radar echo images were done to estimate temporal and spatial variations of water fronts by manual digitizing of cross-shore time stack images. Digitized instantaneous water fronts were validated with wave gauge measurements with an acceptable agreement. Long-shore distribution of mean shoreline positions and intertidal foreshore slopes were then estimated using time averaged images. Run-up, the height of discrete waterlevel maxima, was estimated from the digitized water fronts with the help of foreshore profile. Run-up variationswere parameterized with surf similarity parameter. Low frequency variances in run-up motion were observed, whichwere traveling in the longshore.

Automatic fixing of ship position by simulation-and-matching

In this paper, we propose an automatic method, named simulation-and-matching, for fixing ship position both from a radar echo image and from a synthetic echo image which is synthesized from three-dimensional topographic data, and we show that simulation-and-matching is a more accurate and robust method than those methods that are based on coastline matching. For bulky three-dimensional data, data compression by spline approximation is also considered.