A new radar‐based storm identification and warning technique

Li, Nan; Wei, Ming; Niu, Ben; Mu, Xingmin

doi:10.1002/met.249

Cited by 8 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We are interested in predicting thunderstorm impacts at the hourly scale, with a resolution of a few kilometres: a thunderstorm will be deemed to occur if a lightning strike is observed within 10 km and 30 min of the observation, or the maximum radar reflectivity in this neighbourhood is greater than 35dBZ (this threshold is commonly used for radar thunderstorm detection; see Li et al (2012) and references therein). This criterion is applied at all full hours and on each point of a regular latitude-longitude grid of 20 km mesh, in order to generate a set of pseudo-observations.…”

Section: Thunderstorm Observationsmentioning

confidence: 99%

Probabilistic thunderstorm forecasting by blending multiple ensembles

Bouttier

Marchal

2020

Tellus A: Dynamic Meteorology and Oceanography

View full text Add to dashboard Cite

In numerical weather prediction models, point thunderstorm forecasts tend to have little predictive value beyond a few hours. Thunderstorms are difficult to predict due largely to their typically small size and correspondingly limited intrinsic predictability. We present an algorithm that predicts the probability of thunderstorm occurrence by blending multiple ensemble predictions. It combines several post-processing steps: spatial neighbourhood smoothing, dressing of probability density functions, adjusting sensitivity to model output, ensemble weighting, and calibration of the output probabilities. These operators are tuned using a machine learning technique that optimizes forecast value measured by event detection and false alarm rates. An evaluation during summer 2018 over western Europe demonstrates that the method can be deployed using about a month of historical data. Post-processed thunderstorm probabilities are substantially better than raw ensemble output. Forecast ranges from 9 hours to 4 days are studied using four ensembles: a three-member lagged ensemble, a 12-member non-lagged limited area ensemble, and two global ensembles including the recently implemented ECMWF thunderstorm diagnostic. The ensembles are combined in order to produce forecasts at all ranges. In most tested configurations, the combination of two ensembles outperforms single-ensemble output. The performance of the combination is degraded if one of the ensembles used is much worse than the other. These results provide measures of thunderstorm predictability in terms of effective resolution, diurnal variability and maximum forecast horizon.

show abstract

Section: Thunderstorm Observationsmentioning

confidence: 99%

Probabilistic thunderstorm forecasting by blending multiple ensembles

Bouttier

Marchal

2020

Tellus A: Dynamic Meteorology and Oceanography

View full text Add to dashboard Cite

show abstract

“…For example, VIL was not considered in this study, although many severity algorithms apply it (e.g. Winston and Ruthi, ; Joe et al , ; Hering et al , ; Li et al , ). VIL is not commonly used in the operational radar production in Finland because the melting layer bright band often interferes with VIL estimation in this climate.…”

Section: Data Sources and Observationsmentioning

confidence: 99%

“…A straightforward radar‐based severity ranking method is also included in the Thunderstorm Radar Tracking system, which is used operationally in Switzerland (Hering et al , ; Rotach et al , ). In a recent work, Li et al use a trained support vector machine with radar‐derived VIL, echo‐top altitude, maximum vertical reflectivity and storm centroid height to classify ‘strong’ and ‘general’ storms in Nanjing, China. Many of the above‐mentioned methods have proven their ability to identify significant convective weather in real time, which is crucial, for example, in operational forecasting applications.…”

Section: Introductionmentioning

confidence: 99%

Analysis of a statistically initialized fuzzy logic scheme for classifying the severity of convective storms in Finland

et al. 2013

View full text Add to dashboard Cite

This paper proposes a method for classifying the severity of individual convective storms with real-time weather radar and lightning location data. The algorithm is based on a statistically initialized fuzzy logic model with human-oriented linguistic inference rules. When combined with an object-oriented convective storm tracking algorithm, the severity classification uses the past severity values in addition to the current state of the storm. Furthermore, the proposed method can be customized to correspond to the user-specific needs of different end user groups.The membership functions of the fuzzy logic model are initialized using the statistical analysis of various storm attributes derived from radar and lightning data, which potentially allows the adaptation of the model to different climates. The statistically initialized severity classification is also stable with respect to systematic errors in the measurements. To adjust the model to the Finnish climate, approximately 40 000 storms were tracked with a weather radar-based object-oriented convective storm tracking algorithm. The tracking was performed over a relatively large study area, with the composite of eight C-band weather radars, enabling also the analysis of relatively long-lived intense storms, such as mesoscale convective systems. In addition, the presented work illustrates the statistical characteristics of convective storms in Finland. The study also discusses the importance of careful data quality control when conducting a statistical analysis with an object-oriented storm tracking algorithm.

show abstract

“…Another approach to forecasting longevity comes from algorithms such as the Thunderstorm Identification, Tracking, Analysis, and Nowcasting system (TITAN; Dixon and Wiener 1993;Li et al 2012;Wolfson et al 1994). Both Li et al (2012) and Wolfson et al (1994) use machine learning (ML) to automate part of the tracking and identification process, but neither uses ML to predict storm longevity. MacKeen et al (1999) is most related to our work, as they use linear regression to predict storm longevity.…”

Section: Introductionmentioning

confidence: 99%

Quasi-Operational Testing of Real-Time Storm-Longevity Prediction via Machine Learning

McGovern

Karstens

Smith

et al. 2019

Weather and Forecasting

View full text Add to dashboard Cite

Real-time prediction of storm longevity is a critical challenge for National Weather Service (NWS) forecasters. These predictions can guide forecasters when they issue warnings and implicitly inform them about the potential severity of a storm. This paper presents a machine-learning (ML) system that was used for real-time prediction of storm longevity in the Probabilistic Hazard Information (PHI) tool, making it a Research-to-Operations (R2O) project. Currently, PHI provides forecasters with real-time storm variables and severity predictions from the ProbSevere system, but these predictions do not include storm longevity. We specifically designed our system to be tested in PHI during the 2016 and 2017 Hazardous Weather Testbed (HWT) experiments, which are a quasi-operational naturalistic environment. We considered three ML methods that have proven in prior work to be strong predictors for many weather prediction tasks: elastic nets, random forests, and gradient-boosted regression trees. We present experiments comparing the three ML methods with different types of input data, discuss trade-offs between forecast quality and requirements for real-time deployment, and present both subjective (human-based) and objective evaluation of real-time deployment in the HWT. Results demonstrate that the ML system has lower error than human forecasters, which suggests that it could be used to guide future storm-based warnings, enabling forecasters to focus on other aspects of the warning system.

show abstract

A new radar‐based storm identification and warning technique

Cited by 8 publications

References 18 publications

Probabilistic thunderstorm forecasting by blending multiple ensembles

Probabilistic thunderstorm forecasting by blending multiple ensembles

Analysis of a statistically initialized fuzzy logic scheme for classifying the severity of convective storms in Finland

Quasi-Operational Testing of Real-Time Storm-Longevity Prediction via Machine Learning

Contact Info

Product

Resources

About