Large hail detection using radar-based VIL calibrated with isotherms from the ERA5 reanalysis

Pilorz, Wojciech; Zieba, M; Szturc, Jan; Łupikasza, Ewa

doi:10.1016/j.atmosres.2022.106185

Cited by 9 publications

(6 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When a hailstorm occurs, it is essential to assess the causes for initiating safety measures to minimize the loss of livestock and crops and ensure a better life for the farmers (Malik, Kumar, Salih, & Yaseen, 2021). Generally, the Severe Hail Index (SHI) is utilized for computing the probability of severe hail (POSH) and the maximum estimated size of hail (MESH) (Pilorz et al, 2022; Wilson et al, 2009). Further, various types of agricultural impacts occur due to storms.…”

Section: Introductionmentioning

confidence: 99%

Severity of natural calamities and crop yield prediction using hybrid deep learning model in Uttar Pradesh

Kumar,

Mahapatra

2024

World Water Policy

View full text Add to dashboard Cite

Crop yield prediction has gained major potential for global food production. Predicting crop yields based on specific parameters like soil, environment, crop, and water has been an interesting research topic in recent decades. To accurately predict crop yields, measuring the severities of natural calamities including water level is mainly required. However, the existing studies failed to predict crop yields accurately because of various issues like overfitting problems, difficulty in training, inability to handle large data, and reduced learning capability. Thus, the proposed study develops an efficient mechanism for accurately predicting crop yields by analyzing several natural calamities. Here, the input samples are initially pre‐processed to remove unwanted noises using data normalization and standardization. To enhance the performance of crop yield prediction, natural calamities are computed by using an Extreme Gradient Boosting (XGBoost) model based on parameters like the Palmer Drought Severity Index (PDSI), Severe Hail Index (SHI), and Storm Severity Index (SSI). Also, the hyperparameters of XGBoost model are tuned by utilizing Sheep Flock Optimization Algorithm (SFOA). Finally, the crop yield is predicted by proposing a new one‐dimensional convolutional gated recurrent unit neural network (1D‐CGRU). The proposed classifier predicts the crop yields with reduced error rates like mean square error (MSE) of 0.4363, root mean square error (RMSE) of 0.1904, normalized root mean squared error (NRMSE) of 0.00101, mean absolute error (MAE) of 0.2437, and R‐squared (R2) of .2756. Also, the significant findings of the proposed study positively indicate that this study can be applicable to real‐time agricultural practices and is highly suitable for water quality predictions. Also, it can assist farmers and farming businesses in predicting the yield of crops in a specific season when to harvest and crop a plant for attaining improved crop yields.

show abstract

Section: Introductionmentioning

confidence: 99%

Severity of natural calamities and crop yield prediction using hybrid deep learning model in Uttar Pradesh

Kumar,

Mahapatra

2024

World Water Policy

View full text Add to dashboard Cite

show abstract

“…This criterion is still used in several European countries as a proxy for hail occurrence (Delobbe and Holleman, 2006;Foote et al, 2005;Trefalt et al, 2023). In an effort to utilise this vertical information in storms, studies have sought to produce proxies that integrate reflectivity over the vertical, such as the Vertically Integrated Liquid (VIL, Greene and Clark, 1972;Pilorz et al, 2022) and the VIL density (VILd, Amburn and Wolf, 1997).…”

Section: Introductionmentioning

confidence: 99%

“…These aforementioned methods using Z H as a main variable are still widely used operationally in weather services, either for real-time applications (Smith et al, 2016) or for the production of hail climatologies (US: Wang et al, 2018, Australia: Soderholm et al, 2017Brook et al, 2024, Switzerland: Nisi et al, 2020. While providing a high probability of detection depending on the validation methodology, these techniques are known to suffer from false alarms (Holleman, 2001;Ortega, 2021;Pilorz et al, 2022).…”

Section: Introductionmentioning

confidence: 99%

Comment on egusphere-2024-1336

2024

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Severe hail detection with C-band dual-polarisation radars using convolutional neural networks

Forcadell,

Augros,

Caumont

et al. 2024

Preprint

View full text Add to dashboard Cite

Abstract. Radar has consistently proven to be the most reliable source of information for the remote detection of hail within storms in real-time. Currently, existing hail detection techniques have limited ability to clearly distinguish storms that produce severe hail from those that do not. This often results in a prohibitive number of false alarms that hamper real-time decision-making. This study utilises convolutional neural network (CNN) models trained on dual-polarisation radar data to detect severe hail occurrence on the ground. The morphology of the storms is studied by leveraging the capabilities of a CNN. A database of images of 60 km x 60 km containing 19 different radar-derived features is built above severe hail reports (above 2 cm) and above rain or small hail reports (rain or hail below 2 cm) created for the occasion with the help of a cell-identification algorithm. After a tuning phase on the CNN architecture and its input size, the CNN is trained to output one probability of severe hail on the ground per image of 30 km x 30 km. A test set of 1396 images between 2018 and 2023 demonstrates that the CNN method outperforms state-of-the-art methods according to various metrics. A feature importance study indicates that existing hail proxies as input features are beneficial to the CNN, particularly the maximum estimated size of hail (MESH). The study demonstrates that many of the existing radar hail proxies can be adjusted using a threshold value and a threshold area to achieve similar performance to that of the CNN for severe hail detection. Finally, the output of ten fitted CNN models in inference mode on a hail event is shown.

show abstract

Large hail detection using radar-based VIL calibrated with isotherms from the ERA5 reanalysis

Cited by 9 publications

References 61 publications

Severity of natural calamities and crop yield prediction using hybrid deep learning model in Uttar Pradesh

Severity of natural calamities and crop yield prediction using hybrid deep learning model in Uttar Pradesh

Comment on egusphere-2024-1336

Severe hail detection with C-band dual-polarisation radars using convolutional neural networks

Contact Info

Product

Resources

About