Webinars

GOES Radar Estimation via Machine Learning to Inform NWP

The objective of this research is to develop techniques for assimilating GOES-R Series observations in precipitating scenes for the purpose of improving short-term convective-scale forecasts of high-impact weather hazards. A convolutional neural network (CNN) is developed to transform GOES-R radiances and lightning into synthetic radar reflectivity fields to make use of existing radar assimilation techniques. The ability of CNNs to utilize spatial context is essential for this application and offers breakthrough improvement in accuracy compared to traditional pixel-by-pixel approaches. This presentation will describe the development of GREMLIN, present results using GREMLIN in numerical weather prediction, and discuss current work to make GREMLIN available over the Full Disk.

University at Albany - United States

Severe storm reports of tornadoes, hail and wind damage, as well as flash flooding, are used by many weather services to verify severe weather forecasts and warnings. These databases are also used by researchers to better understand the frequency, regional variability and intensity of severe storms, and by governments and decision makers to better prepare for the multiple hazards associated with severe storms. Long severe storm reports databases exist in the United States and Europe, but many places around the world where severe storms are frequent, such as South America, have no coordinated documentation of storm reports. 

In 2018, a group of volunteer meteorologists decided to start documenting reports of severe weather associated with convective storms in Brazil. This talk will cover the existing datasets in the world and their uses, with focus on the Brazilian dataset and parallel efforts in other countries of South America. The reporting methodology and quality control will be discussed. The frequency of severe storms in Brazil derived from the database will be investigated. Finally, a case study will be shown to illustrate the applications of the database in nowcasting operations.

The Northern Tornadoes Project – Toward a Better Understanding of Severe Local Storms in Canada

The US National Weather Service (NWS) Weather Surveillance Radar 1988 Doppler (WSR-88D) network was upgraded with dual-polarization capability from 2011 to 2014. At the core of this upgrade is a fundamental change in the manner in which the radar transmits and receives signal. Dual-polarization radars are able to transmit and receive both horizontally and vertically polarized electromagnetic waves, yielding more information about a target's size, shape, orientation and composition, in addition to the traditional returned power and velocity. Thus, forecasters and researchers can utilize these data to improve understanding and prediction of many hazards, including hail identification and sizing, tornado identification and real-time rating estimation, updraft analysis, and heavy rain / flood potential via improved quantitative precipitation estimation. Over the last decade, the US weather enterprise (NOAA / NWS, universities, private industry, media) has realized a wealth of benefits from this upgrade. This presentation will introduce the basic principles behind dual-polarization radar and examples of these benefits during the warm season, including cases of severe convective storms and heavy rainfall events, to illustrate the operational advantages of this radar technology.

ProbSevere -- using AI to nowcast convective weather

Forecasting the timing, location, and intensity of thunderstorm hazards is a primary responsibility of NOAA's National Weather Service and remains a challenge, particularly in situations with widespread, fast-evolving storms. Machine-learning methods have shown promise in quickly distilling meaningful signals from the vast array of remotely sensed data available to forecasters. NOAA's Probsevere portfolio of nowcasting tools use remotely sensed meteorological data and machine learning/artificial intelligence methods to diagnose and predict convective hazards in the short-term. This talk will give an overview of the ProbSevere models and their use in the United States, but will focus on the satellite-only models of ProbSevere, which can be employed anywhere in the GOES-East and GOES-West geostationary rings. 

Civil Defense, protection, and risk management of Santa Catarina state

The state of Santa Catarina is located in the center of the Southern Region of Brazil, which is also composed of the states of Paraná and Rio Grande do Sul. As it lies between the parallels 26º00’09”S 29º21’03”S and the meridians 48º21’30”W and 53º50’09”W it belongs to the global transition region between the tropics and mid-latitudes.

Since 2014, the Civil Defense of Santa Catarina State (DC/SC) has been working intensively on prevention of natural disasters.

There are specific strategies for forecasting severe weather in Brazil that still need to be discussed. Better knowledge of the atmospheric environments that leads a development of storms is essential. A fundamental importance is to extract the most qualified information to be able to quantify it in a way that contributes to the best decision-making in the meteorological operation. This qualification and quantification can be carried out in such a way that the strategy used involves the calculation of meteorological parameters, that objectively and concisely highlight the conducive conditions to the development of systems that may cause damage and disasters.

The Civil Defense Regional Coordinations were structured in 20 regions, 3 (4) dual-polarization weather radars and also the GOES-16 satellite data reception antenna were acquired, working on a network of hydro-meteorological stations expansion, high resolution meteorological and hydrological modeling service for nowcasting, flood situations and operation of state dams. These are some of the strategies for civil defense, protection and risk  management of Santa Catarina state.

Outcomes of the HIGHWAY Lake Victoria Basin project in East Africa and implications for a sustainable Nowcasting and Early Warning System

Lake Victoria in East Africa is one of the deadliest bodies of water in the world due to the high number (>1000) of fatalities that occur annually over the lake.   The Lake Victoria Basin (LVB) supports a population of 5.4 million people.   Approximately 217,000 fishermen and travelers who depend on the lake for their livelihoods, are often caught unawares by strong outflows from intense, nocturnal thunderstorms that result in increased wave heights that capsize their boats. The frequency of LVB thunderstorms are known for producing one of the lightning hotspots on the earth.  Fatalities also occur during the daytime, in the absence of thunderstorms, due to strong mesoscale winds over the lake that cause overloaded boats to capsize.   The East African National Meteorological and Hydrological Services submitted a direct request to the World Meteorological Organization (WMO) for assistance in improving forecasts for LVB.  The WMO  ran a 3 year project in the region called HIGHWAY that was directed at collecting enhanced observations over the lake to better understand the evolution of severe weather and importantly, to improve forecasts and convective outlooks of stormy weather for fishermen and the lake communities. The ultimate goal of HIGHWAY is to develop a sustainable Early Warning System for the region.   In this talk I will discuss the major outcomes from the HIGHWAY project, emphasizing the positive results but also the challenges that still remain in nowcasting and warning of severe weather in East Africa; challenges that also exist for many regions of the world. 

The European Severe Storms Laboratory: A dedicated centre for research and forecaster training in Europe

In 2006, ESSL was founded as an association with members by a group of scientists from different European countries to provide a framework for studying severe storms internationally within Europe.  Since then, the ESSL has slowly grown and employs 12 employees, half of them part-time.

In the meantime, 13 European weather services as well as EUMETSAT and ECMWF have become members of ESSL. ESSL's core tasks include the management of the European Severe Weather Database, carrying out various research projects on forecasting of severe weather and climate change, and on evaluating new forecast tools at the ESSL Testbed.

Moreover, ESSL organizes courses for weather forecasters on predicting severe weather at its premises in Wiener Neustadt, Austria. In my talk I will discuss ESSL's research work, the collection of severe weather data, and the Testbed, as well as our recent work on an internationally usable rating scale for tornado and wind damage to be implemented in the near future.

Research at the Interface of Operations: Exploring Severe Convective Storm Forecasting and High-Resolution Models in NOAA's Hazardous Weather Testbed

Research in NOAA’s Hazardous Weather Testbed focuses on the prediction of severe convective storms at lead times from minutes to days, using cutting-edge tools developed by researchers and informed by operational forecaster needs. Collaborative efforts in the Hazardous Weather Testbed by the Storm Prediction Center (SPC) and the National Severe Storms Laboratory (NSSL) currently center on annual Spring Forecasting Experiments (SFEs). Annual SFEs provide a five-week proving ground for new numerical weather prediction and forecasting techniques, allowing for researchers and forecasters to work together to analyze and improve the tools being developed for forecasters prior to operational implementation. This collaboration accelerates the transition from research to operations, as well as providing researchers with insight into problems that operational forecasters face and where more research is needed.

During the SFE, participants complete a variety of forecasting and evaluation activities, supplying subjective verification of modeling aspects that are not so easily captured by objective verification. Participants also use the experimental guidance to create forecasts, testing the guidance in a quasi-operational setting. The focus of the SFE is at scales most relevant to operational products issued by the SPC (i.e., from a few hours to a few days in advance). However, SFE participants come from across the National Weather Service, national laboratories, universities, and international weather communities to ensure that we solicit ideas from people with a variety of backgrounds and roles in the meteorological community. With dozens of models contributed to the SFE each year, datasets created during the spring are further analyzed and examined by facilitators and collaborators throughout the off-season to explore ensemble configuration strategies, physical parameterization schemes, and novel post-processing and calibrated guidance methodologies. Each year, results from the experiment are used to develop the next year’s activities, with an aim to improve the SFE experience every year. In addition to discussing the mechanics of creating and executing the SFE, recent findings from the 2019–2022 SFEs will be discussed herein. This talk will conclude with potential future directions for hybrid testbed activities, recognizing how virtual and in-person components both can be used to improve collaboration and innovation across the weather enterprise.

Nowcasting operations at the National Meteorological Service of Argentina

The National Meteorological Service of Argentina has been issuing public short term warnings for 20 years. During this period, the amount of operational weather radars increased from 1 to 16, with 12 of these radars installed in the previous 8 years. This presentation will focus on the characteristics of the short term warning, a few examples, warning verification and some of the key operational challenges detected.

The Utility of Dual-Polarization Weather Radar in Analyzing and Predicting Warm-Season Hazards

The US National Weather Service (NWS) Weather Surveillance Radar 1988 Doppler (WSR-88D) network was upgraded with dual-polarization capability from 2011 to 2014. At the core of this upgrade is a fundamental change in the manner in which the radar transmits and receives signal. Dual-polarization radars are able to transmit and receive both horizontally and vertically polarized electromagnetic waves, yielding more information about a target's size, shape, orientation and composition, in addition to the traditional returned power and velocity. Thus, forecasters and researchers can utilize these data to improve understanding and prediction of many hazards, including hail identification and sizing, tornado identification and real-time rating estimation, updraft analysis, and heavy rain / flood potential via improved quantitative precipitation estimation. Over the last decade, the US weather enterprise (NOAA / NWS, universities, private industry, media) has realized a wealth of benefits from this upgrade. This presentation will introduce the basic principles behind dual-polarization radar and examples of these benefits during the warm season, including cases of severe convective storms and heavy rainfall events, to illustrate the operational advantages of this radar technology.