Extreme Weather Forecasting

Extreme Weather Forecasting reviews current knowledge about extreme weather events, including key elements and less well-known variables to accurately forecast them. The book covers multiple temporal scales as well as components of current weather forecasting systems. Sections cover case studies on successful forecasting as well as the impacts of extreme weather predictability, presenting a comprehensive and model agnostic review of best practices for atmospheric scientists and others who utilize extreme weather forecasts.

• Reviews recent developments in numerical prediction for better forecasting of extreme weather events
• Covers causes and mechanisms of high impact extreme events and how to account for these variables when forecasting
• Includes numerous case studies on successful forecasting, outlining why they worked

• Language: English
• Publisher: Elsevier Science
• Release date: Oct 11, 2022
• ISBN: 9780128202432

Book preview

Front Cover for Extreme Weather Forecasting - State of the Science, Uncertainty and Impacts - 1st edition - by Marina Astitha, Efthymios Nikolopoulos

Extreme Weather Forecasting

State of the Science, Uncertainty and Impacts

Edited by

Marina Astitha

Department of Civil and Environmental Engineering, School of Engineering, University of Connecticut, Storrs, CT, United States

Efthymios Nikolopoulos

Civil and Environmental Engineering, Rutgers University, New Brunswick, NJ, United States

Table of Contents

Cover image

Title page

Copyright

List of contributors

Foreword

Preface

Chapter 1. Overview of extreme weather events, impacts and forecasting techniques

Abstract

Subchapter 1.1. Definition of extreme weather events

1.1.1 Extreme heat

1.1.2 Extreme cold—severe winter storms

1.1.3 Tropical and extratropical storms

1.1.4 Severe convective storms

1.1.5 Extreme rainfall

Subchapter 1.2. Weather forecasting

Subchapter 1.3. Extreme weather forecasting in urban areas

1.3.1 Introduction

1.3.2 Urban heat island

1.3.3 Heat wave forecasting

1.3.4 Air quality modeling and prediction

1.3.5 Forecasting urban precipitation

1.3.6 Forecasting coastal urban flooding

Subchapter 1.4. Wildfires and weather

1.4.1 Introduction: wildfires and weather—a coupled system

1.4.2 Wildfire prediction and risk assessment

1.4.3 Data requirements and data quality

1.4.4 Wildfire prediction sensitivities and uncertainties

1.4.5 Improved wildfire modeling for improved wildfire preparedness

Chapter 2. Operational multiscale predictions of hazardous events

Abstract

2.1 Introduction

2.2 Example case: 2015 European heatwave

2.3 Key factors of predictability

2.4 Hazard forecasting

2.5 Evaluation of hazardous events

2.6 Conclusion

2.7 Summary

References

Chapter 3. Forecasting extreme weather events and associated impacts: case studies

Subchapter 3.1. Extreme heat

3.1.1 Introduction

3.1.2 Data

3.1.3 Methodology

3.1.4 Results

3.1.5 Conclusions

Acronyms

Subchapter 3.2. Atmospheric rivers

3.2.1 Introduction

3.2.2 Atmospheric river evolution

3.2.3 Forecasting atmospheric rivers

3.2.4 Regional models

3.2.5 Ensemble forecast systems

3.2.6 Verification

3.2.7 Decision support

3.2.8 Summary

Subchapter 3.3. The hydrological Hillslope-Link Model for space-time prediction of streamflow: insights and applications at the Iowa Flood Center

3.3.1 Introduction

3.3.2 A generic set of ordinary differential equations to model water flows in the landscape and the river network

3.3.3 Domain decomposition and model inputs for the implementation of Hillslope-Link Model

3.3.4 Example of model performance using different configurations of vertical and horizonal fluxes at the hillslope scale

3.3.5 Insights and real-time applications of the Hillslope-Link Model at the Iowa Flood Center

3.3.6 Summary and conclusions

3.3.7 Future work and upcoming challenges

Acknowledgments

Subchapter 3.4. Social impacts: integrating dynamic social vulnerability in impact-based weather forecasting

3.4.1 Drivers of social impacts from extreme weather events

3.4.2 The need for integrated forecasting tools to anticipate social impacts

3.4.3 Insights of methodological advances in modeling the coupled sociohydrometeorological system in high-impact weather events

3.4.4 Toward operational decision-making in high-impact weather events: insights from a participatory role-playing experiment

3.4.5 Conclusion

Subchapter 3.5. Landslides and debris flows

3.5.1 Introduction

3.5.2 Data and methodology

3.5.3 Results

3.5.4 Discussion

3.5.5 Conclusions

Acknowledgments

Subchapter 3.6. Weather-induced power outages

3.6.1 Power grid outages and severe weather

3.6.2 Modeling weather impact on the electric grid

Afterword

Index

Copyright

Elsevier

Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom

50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States

Copyright © 2023 Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

ISBN: 978-0-12-820124-4

For Information on all Elsevier publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Candice Janco

Acquisitions Editor: Amy Shapiro

Editorial Project Manager: Andrae Akeh

Production Project Manager: R. Vijay Bharath

Cover Designer: Christian J. Bilbow

Typeset by MPS Limited, Chennai, India

List of contributors

Emmanouil Anagnostou

Department of Civil and Environmental Engineering, University of Connecticut, CT, United States

Eversource Energy Center, University of Connecticut, CT, United States

Sandrine Anquetin,     Université Grenoble Alpes, CNRS, IRD, Grenoble INP, IGE, Grenoble, France

Marina Astitha,     Department of Civil and Environmental Engineering, School of Engineering, University of Connecticut, Storrs, CT, United States

Tibebu Ayalew,     Iowa Flood Center, The University of Iowa, Iowa City, IA, United States

Martina Calovi,     Department of Geography, Norwegian University of Science and Technology, Trondheim, Norway

Forest Cannon,     Center for Western Weather and Water Extremes, Scripps Institution of Oceanography, University of California, San Diego, CA, United States

Amy DeCastro,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

Diego Cerrai

Department of Civil and Environmental Engineering, University of Connecticut, CT, United States

Eversource Energy Center, University of Connecticut, CT, United States

Guido Cervone,     Department of Geography and Institute for Computational and Data Sciences, The Pennsylvania State University, University Park, PA, United States

Dan Chen,     Institute of Urban Meteorology, CMA, Beijing, China

Quang-Van Doan,     University of Tsukuba, Tokyo, Japan

Laura Clemente,     Geospatial Research Laboratory, Engineer Research and Development Center, Alexandria, VA, United States

Masih Eghdami,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

F. Di Giuseppe,     European Centre for Medium-Range Weather Forecasts-ECMWF, Reading, United Kingdom

Timothy W. Juliano,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

Domingo Muñoz-Esparza,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

C. Di Napoli

European Centre for Medium-Range Weather Forecasts-ECMWF, Reading, United Kingdom

University of Reading, Reading, United Kingdom

Morgan Fonley,     Alma College, Alma, MI, United States

Maria Frediani,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

Weiming Hu,     Center for Western Weather and Water Extremes, Scripps Institution of Oceanography, University of California, San Diego, CA, United States

Navid Jadidoleslam,     School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA, United States

Pedro Jimenez,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

Jason C. Knievel,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

Sana Khan,     Earth System Science Interdisciplinary Center (ESSIC), University of Maryland, College Park, MD, United States

Dalia B. Kirschbaum,     Hydrological Sciences Laboratory, NASA Goddard Space Flight Center, Greenbelt, MD, United States

Branko Kosovi,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

Witold F. Krajewski

Department of Civil and Environmental Engineering, The University of Iowa, Iowa City, IA, United States

Iowa Flood Center, The University of Iowa, Iowa City, IA, United States

Hiroyuki Kusaka,     University of Tsukuba, Tokyo, Japan

Linus Magnusson,     European Centre for Medium-Range Weather Forecasts-ECMWF, Reading, United Kingdom

Ricardo Mantilla,     Department of Civil Engineering, Price Faculty of Engineering, University of Manitoba, MB, Canada

Luca Delle Monache,     Center for Western Weather and Water Extremes, Scripps Institution of Oceanography, University of California, San Diego, CA, United States

Efthymios Nikolopoulos,     Civil and Environmental Engineering, Rutgers University, New Brunswick, NJ, United States

F. Pappenberger,     European Centre for Medium-Range Weather Forecasts-ECMWF, Reading, United Kingdom

C. Prudhomme,     European Centre for Medium-Range Weather Forecasts-ECMWF, Reading, United Kingdom

Felipe Quintero,     Iowa Flood Center, The University of Iowa, Iowa City, IA, United States

Prathap Ramamurthy,     City College of New York, NY, United States

Pallav Ray,     Florida Institute of Technology, Melbourne, FL, United States

Isabelle Ruin,     Université Grenoble Alpes, CNRS, IRD, Grenoble INP, IGE, Grenoble, France

Scott Small,     Iowa Flood Center, The University of Iowa, Iowa City, IA, United States

Amanda Siems-Anderson,     Research Applications Laboratory, National Center for Atmospheric Research, Boulder, CO, United States

Galateia Terti,     Université Grenoble Alpes, CNRS, IRD, Grenoble INP, IGE, Grenoble, France

Mukul Tewari,     IBM, Thomas J. Watson Research Center, Yorktown Heights, NY, United States

Nicolas Velasquez,     Iowa Flood Center, The University of Iowa, Iowa City, IA, United States

Zhihua Wang,     Arizona State University, Phoenix, AZ, United States

Foreword

Walter A. Robinson, Department of Marine, Earth, & Atmospheric Sciences, North Carolina State University, Raleigh, NC, United States

Before dawn on August 24, 1979, Ms. Debra Stevens drove her van into flood waters in Fort Smith Arkansas (USA). Ms. Stevens was on her newspaper delivery route, which she had run daily for 21 years. During her 911 call, she said she had never encountered flooding this severe. (Ms. Stevens later drowned.) The US National Weather Service recorded 4.04 inches (102.6 mm) of rain at Fort Smith that day, on top of an additional 4.46 inches (113.3 mm) that fell over the previous 2 days.

This incident, and countless others like it, drive home key points about extreme weather:

• Accurate, timely, and effectively disseminated forecasts of extreme weather are a matter of life and death.

• Climate change—the never seen it before effect—is increasing the severity of extreme weather and our vulnerability to it.

• Loss of life and property most often comes not from a weather extreme itself but from an impact of that extreme weather (an unseen flood on a road). Thus improved understanding and predictions of impacts are fully as important as better weather forecasts.

Better forecasts of extreme weather events and their impacts are central to any adaptive response to climate change. This volume, therefore, is exceptionally timely in laying out the nature of extreme weather and its impacts, how they are predicted, how predictions are being improved, and the prospects for better forecasts in the future. Anyone tasked with protecting assets or public safety will benefit from learning how extreme weather forecasts are made, as well as what can and cannot be predicted and with what confidence.

The weather extremes and impacts discussed in this book—extreme heat, wildfire, floods, landslides, and power outages—are salient from recent news reports. Much like Tolstoy’s unhappy families, each is unique in its geographic, meteorological, and social context. That said, overarching themes emerge when they are juxtaposed, as they are in this book.

Most evident, perhaps, is that weather extremes and their impacts emerge across wide ranges of space and time scales. A wildfire, for example, may start during a high-wind event, but the risk of fire can grow during a season of drought, and it may be enhanced by a fuel load that accumulated over years of fire suppression. Each scale or system component has its own intrinsic predictability, realized with different modeling tools. Thus a figure like 2.6, which refers to wildfire, could be drawn for any weather extreme. The shared challenge is to stitch models together into a seamless forecast system.

Human impacts of extremes are mediated through complex systems. Predicting these impacts demands that all components of the system, well beyond the conventional meteorological variables, be observed. Thus wildfire forecasts require estimates of fuel loads, flood forecasts depend on knowledge of antecedent soil moisture and stream flow, and predictions of air pollution episodes need emissions data. These data must be assimilated into forecast systems, as has long been done for weather data but is being extended to impact relevant quantities, such as stream flows and fire perimeters.

Furthermore, coping with this system, complexity calls for the application of new tools, such as model, ensembles that provide probabilistic forecasts and expose uncertainty, and machine learning models, especially for those parts of systems, including engineered components, such as the electric power grid, that are not readily described by well-known physical laws. Finally, impacts are, in part, socially determined, even within a given physical environment, and forecasts and warnings must communicate impacts in a manner salient to their audiences (e.g., warmings of roads flooded rather than inches of rainfall).

Progress is being made across the activities that yield better forecasts of extreme weather hazards and impacts. First and foremost, for a meteorologist, are the gains in forecast skill that result from larger ensembles of models with better physical parameterizations being run at ever higher spatial resolutions. As the examples in this volume show, the coupling of weather models to hazard and impact models is becoming more sophisticated, and impact models are themselves improving, often through applications of machine learning. Given these advances in computing, model development, observations, and systems-level understanding of how extremes affect people, readers should come away with some optimism that improved forecasts of extreme weather and its impacts are forthcoming. Moreover, they may be persuaded that relatively modest investments in such improvements will yield significant gains in human health and safety, even, or especially, in the face of a changing climate.

https://www.cnn.com/2019/08/31/us/arkansas-woman-drowns-911-dispatcher/index.html

https://floodlist.com/america/usa/flash-floods-arkansas-august-2019

Preface

Marina Astitha

The impacts of extreme weather are multidimensional and multifaceted and are recognized by world organizations and individual nations as most influential for health, economic, and social development worldwide. Results of the 2018 Global Risks Report published by the World Economic Forum revealed severe weather as the most likely threat to the world over a 10-year period, topping weapons of mass destruction (WEF, 2018). The 2021 Atlas of mortality and economic losses from weather, climate, and water extremes (1970–2019) from the World Meteorological Organization states that "the number of disasters has increased by a factor of five over the 50-year period, driven by climate change, more extreme weather and improved reporting. But, thanks to improved early warnings and disaster management, the number of deaths decreased almost three-fold" (WMO, 2021). This not only constitutes a significant success for human advances in understanding the laws of nature that govern such events, advances in technology, and international collaboration but also serves as an unnerving reminder that the future holds more challenges for the human race to overcome.

We should note here that the majority of impacts from extreme weather are felt by developing nations. The WMO’s 2021 report states that both the United Nations’ and the World Bank’s economic classification methodologies reveal that "the majority of reported deaths from weather, climate and water extremes occurred in developing countries, while countries with developed economies incurred the majority of economic losses". Approximately 80%–90% of all deaths from weather extremes occurred in developing economies. This is essential information that highlights the need for consistent technology transfer and substantial knowledge exchange between developed and developing nations. In the heart of preparedness, risk assessment, and mitigation of impacts from extreme weather rests the capability and efficacy of forecasting such events reliably, accurately, and in a timely manner.

Forecasting extreme weather events, such as extreme heat/cold, extreme rainfall, tropical and extratropical storms, severe winter and convective storms, flooding and wildfires, is equally important with forecasting their impacts. Humans have long been curious and interested in extreme weather, as most have encountered at least one actual event (most probably multiple) that showed its wrath and catastrophic consequences. Hurricane Katrina in 2005 was one of the most devastating hurricanes on record, with more than US$ 163 billion in losses and over 10,000 deaths in the United States. Central America and the Caribbean have witnessed significant economic losses and deaths from storms, flooding, and landslides. Canada’s drought in 1977 was the costliest event for the country. The heat wave of 2003 in Europe and the Russian Federation in 2010 caused significant loss of life compared to the past 50 years. Africa has been devastated by droughts, with four drought events to account for 89% of total deaths in Africa from weather, climate, and water extremes in the last 50 years. Storms and flooding are record-breaking impactful events in Asia; flooding, landslides, drought, and extreme temperatures for South America; storms, droughts, and wildfires for South-West Pacific. This is just a quick reminder of the global consequences from severe weather that do not stop at country borders. With this book, we intend to provide a close look at the current state-of-the-science techniques to forecast such events and their associated impacts, focusing on specific case studies. The authors of each chapter provide context towards the challenges and successes of efficient, reliable, and timely weather forecast and the necessary steps to enhance our forecasting capabilities in the future.

Extreme Weather Forecasting is organized into three chapters that offer a review of current techniques to forecast extreme weather in multiple scales (from global to urban, hourly scale to seasonal) and a curated selection of extreme weather case studies and its impacts. The case studies include extreme heat, atmospheric rivers, flood forecasting, debris flows and landslides, social impacts, and weather-induced power outages. A special section is devoted to wildfires, a hazard that devastates societies around the globe. Each chapter can be read independently from the rest, and the reader does not need to follow a particular order.

Chapter 1, "Overview of extreme weather events, impacts and forecasting technique," provides definitions of extreme weather events, summary of weather forecasting techniques, and reviews of extreme weather in urban areas and prediction of wildfires. Chapter 1.1 covers definitions for extreme heat, extreme cold and severe winter storms, extreme rainfall, severe convective storms, tropical and extratropical storms. Chapter 1.2 gives a brief overview of weather forecasting techniques that includes numerical prediction, statistical forecasting, and newly adopted machine learning and artificial intelligence approaches. Chapter 1.3 summarizes extreme weather forecasting from an urban area perspective, with specific descriptions of the urban heat island phenomenon and related heat waves, air quality forecasting, urban precipitation, and urban flooding. Chapter 1.4 examines the coupling of wildfires and weather, offers detailed description of wildfire ignition processes and impacts, discusses the connection between meteorological conditions and extreme wildfires, and addresses current advances and challenges of prediction and risk assessment of wildfires.

Chapter 2, "Operational multiscale predictions of hazardous events," offers a detailed description of real-time operational weather forecasting on various temporal scales from the European Centre for Medium-Range Weather Forecasts. The case of the 2015 European heat wave illustrates how extreme weather is predicted at different timescales and how to evaluate the outcome of those predictions (Chapter 2.2). Key factors for predictability are analyzed with focus on specific extreme weather events (cold spells, heat waves, wind storms, extreme precipitation, severe convection) (Chapter 2.3). Chapter 2.4 discusses how weather forecast is used for predicting hazards, such as floods, drought, heat stress, and wildfires. The final section, Chapter 2.5, highlights how we evaluate forecasts based on the purpose of the evaluation; inform about the forecast skill; and/or guide the development of the forecasting system.

Chapter 3, "Forecasting extreme weather events and associated impacts: case studies," details forecasting techniques, challenges and successes for various types of extreme weather and its impacts. The first case study focuses on forecasting heat wave events in New York City at high spatiotemporal resolution implementing the analog ensemble approach (Chapter 3.1). The second case study discusses how advancements in numerical weather prediction have benefitted atmospheric river forecast skill in the 21st century and addresses the potential of machine learning and effective communication strategies to augment decision support (Chapter 3.2). The third case study presents the flood forecasting system developed and implemented by the Iowa Flood Center to predict stream flow fluctuations in the state of Iowa and provide real-time flood warnings (Chapter 3.3). Chapter 3.4 addresses the crucial topic of social impacts related to extreme weather events and details the need to integrate modern forecasting-warning chains with dynamic human vulnerability models describing fluctuations in the activities, behaviors, and personal and situational constraints of the people exposed to the weather extremes. Chapter 3.5 describes in detail the opportunities and challenges of using forecasted precipitation within landslide models, with the goal of advancing landslide early warning at global scales. In the last section of Chapter 3, we learn about a nuanced area of research that connects extreme weather forecasting with prediction of power outages using artificial intelligence and machine learning (Chapter 3.6).

References

WEF Global Risks Report, 2018 WEF Global Risks Report, 2018. World Economic Forum, 13th ed. Geneva, Switzerland. www.weforum.org. ISBN: 978-1-944835-15-6.

WMO, 2021 WMO, 2021. The Atlas of Mortality and Economic Losses from Weather, Climate and Water Extremes (1970–2019). ISBN: 978-92-63-11267-5.

Chapter 1

Overview of extreme weather events, impacts and forecasting techniques

Abstract

This chapter is dedicated to the overview of extreme weather events and their impacts, as well as the introduction of forecasting concepts and techniques. There are many types of weather extremes that are accompanied by a continuously increasing research literature that contemplates interconnections between human and natural influences, especially with regard to climate change (Robinson, 2021). The events we describe in this chapter are not a complete list. We provide definitions of extreme weather events (Section 1.1) that pose significant societal and economic impacts and are represented by case studies in the chapters that follow. Most of the event types and techniques described here are analyzed further and more extensively in Chapter 2. An overview of current weather forecasting practices is presented in Section 1.2. Furthermore, in this chapter, we provide comprehensive reviews for extreme weather forecasting in urban areas (Section 1.3) and prediction of wildfires (Section 1.4). Urban areas represent scales and complex environments that affect a large part of the population. Wildfires, despite being primarily of natural origin, have a complex relationship with extreme weather conditions and increasingly devastating effects on the built environment and the society. Thus, we need to be able to better understand and predict them accurately in space and time.

Keywords

Extreme weather; definition; weather forecasting

This chapter is dedicated to the overview of extreme weather events and their impacts, as well as the introduction of forecasting concepts and techniques. There are many types of weather extremes that are accompanied by a continuously increasing research literature that contemplates interconnections between human and natural influences, especially with regard to climate change (Robinson, 2021; AghaKouchak et al., 2020; Ahrens and Samson, 2011; Chen et al., 2018; Easterling et al., 2000; Ebi et al., 2021; Schleussner et al., 2018, among others). The events we describe in this chapter are not a complete list. We provide definitions of extreme weather events (see Section 1.1) that pose significant societal and economic impacts and are represented by case studies in the chapters that follow. Most of the event types and techniques described here are analyzed further and more extensively in Chapter 2. An overview of current weather forecasting practices is presented in Section 1.2. Furthermore, in this chapter we provide comprehensive reviews for extreme weather forecasting in urban areas (see Section 1.3) and prediction of wildfires (see Section 1.4). Urban areas represent scales and complex environments that affect a large part of the population. Wildfires, despite being primarily of natural origin, have a complex relationship with extreme weather conditions and increasingly devastating effects on the built environment and the society. Thus we need to be able to better understand and predict them accurately in space and time.

Subchapter 1.1

Definition of extreme weather events

Marina Astitha¹ and Efthymios Nikolopoulos²,    ¹Department of Civil and Environmental Engineering, School of Engineering, University of Connecticut, Storrs, CT, United States,    ²Civil and Environmental Engineering, Rutgers University, New Brunswick, NJ, United States

An extreme weather event is defined as a natural occurrence over a certain time period and space with rare characteristics in terms of magnitude, location, duration, and/or extent. The premise under which we discuss forecasting of extreme weather events is primarily the impact those events have on our society and their potential to exacerbate in a future climate. An indication of that impact, worldwide, is illustrated in Fig. 1.1.1, adopted from the 2021 World Meteorological Organization (WMO) report Atlas of Mortality and Economic Losses (WMO, 2021). The numbers clearly show that the number of disasters has increased by a factor of five over the 50 years’ period (Fig. 1.1.1A); deaths decreased almost threefold (Fig. 1.1.1B), a likely result of advances in early warning systems worldwide (IPCC, 2012). Economic losses have increased sevenfold from the 1970s to the 2010s (Fig. 1.1.1C), with storms being the most significant cause of damage and with an attributed percentage that increases over the decades. In the following sections, we offer definitions of extreme heat, extreme cold and severe winter storms, extreme rainfall, severe convective storms, tropical and extratropical storms.

Figure 1.1.1 Distribution of (A) number of disasters, (B) number of deaths, and (C) economic losses by hazard type by decade globally. Graph originally from WMO, 2021. Atlas of mortality and economic losses from weather, climate and water extremes (1970–2019). WMO No. 1267, ISBN: 978-92-63-11267-5.

1.1.1 Extreme heat

Extreme heat (also referred to as excessive heat or heat wave) is defined as a prolonged period of time with abnormally high day and night atmospheric temperatures, usually above the average expected temperature for that specific region and time of the year and usually accompanied by high humidity (Robinson, 2001; Ramis and Amengual, 2018). Even though a formalized, globally accepted definition with specific temperature and humidity limits does not exist worldwide, a common criterion is that society is susceptible to or unable to cope with these events (WMO, 2004). The WMO defines extreme heat as five or more consecutive days during which the daily maximum temperature exceeds the average maximum temperature by 5°C (9°F) or more (WMO, 2015). According to Federal Emergency Management Agency, in the United States, extreme heat is a period of 2–3 days of high heat and humidity with temperatures above 90°F. The Glossary of Meteorology of the American Meteorological Society defines a heat wave as a period of abnormally and uncomfortably hot and usually humid weather. Such a period lasts at least one day, but conventionally lasts from several days to several weeks. The United States (US) National Weather Service (NWS) has established heat watches and warnings based on the calculation of the heat index (a measure of how hot it feels, taking into account temperature and humidity), to inform and protect the public (NWS, 1994). The frequency, duration, length, and intensity of extreme heat waves have increased in the United States, from an average of two heat waves per year during the 1960s to six per year during the 2010s (see Fig. 1.1.2; data source: US NOAA and EPA; adapted from an analysis by Habeeb et al., 2015).

Figure 1.1.2 Heat wave characteristics in the United States from 1961 to 2019. The graphs show averages across 50 metropolitan areas by decade ( NOAA, 2021). EPA’s Climate Change Indicators in the United States: http://www.epa.gov/climate-indicators. Data from NOAA (National Oceanic and Atmospheric Administration). 2021. Heat stress datasets and documentation. Accessed February 2021. http://www.ncdc.noaa.gov/societal-impacts/heat-stress/data.

1.1.2 Extreme cold—severe winter storms

Extreme cold (also referred to as a cold wave or severe cold) is the drop of atmospheric temperature near or below freezing point, caused by movement of arctic air over the region of interest. As with extreme heat, extreme cold definitions vary by region and depend on the atmospheric conditions determined as normal for that region and time of the year. The premise of the extreme cold definition is closely related to the impacts it has on the human body and physiology causing frostbites or hypothermia that have the potential to become life-threatening. The US NWS has established the Wind Chill Temperature (WCT) index to inform and protect the public: Wind Chill Advisory is issued when the wind chill index is between –15°F and –24°F for at least 3 hours and Wind Chill Warning is issued when the Wind chill index is below –25°F for at least 3 hours. Severe cold outbreaks might sound counterintuitive to the global warming scenarios, but the shifts in atmospheric circulation due to climate change have the potential to increase cold wave occurrences in the middle latitudes (Walsh et al., 2001).

Winter storms are weather events where precipitation manifests as freezing rain or frozen hydrometeors such as snow and sleet. Types of winter storms are snowstorms, blizzards, ice storms, and lake effect storms (where the snow is formed due to moisture abundance in the Great Lakes). The combination of heavy snow and wind chills, mostly associated with gusty winds, turns winter storms into severe events that can be life threatening (NOAA NSSL, accessed November 2021). An ice storm is a storm, which results in the accumulation of at least 0.25 inches of ice on exposed surfaces. A blizzard is a dangerous snowstorm that combines heavy snowfall and winds over 35 mph that result in low visibility for at least 3 hours. In the United States, winter storm watches and warnings are issued by the NWS to inform and protect the public.

1.1.3 Tropical and extratropical storms

A tropical cyclone is a warm-core nonfrontal synoptic-scale cyclone that develops over tropical waters and has organized deep convection with a closed wind circulation around a well-defined center (US NOAA/NWS and the National Hurricane Center-NHC). A tropical storm is a tropical cyclone that has maximum sustained surface winds (using the US 1-minute average) ranging from 39 to 73 mph (34 to 63 knots). A hurricane is a tropical cyclone that has maximum sustained surface winds of 74 mph or greater (64 knots or greater). According to the 2021 WMO report, tropical cyclones represent 17% of weather-, climate- and water related disasters and are responsible for one third of both deaths and economic losses over a 50-year period (1970–2019) (WMO, 2021; Fig. 1.1.3).

Figure 1.1.3 Distributions of (A) number of disasters, (B) number of deaths, and (C) economic losses by hazard globally (1970–2019). Graph originally from WMO, 2021. Atlas of mortality and economic losses from weather, climate and water extremes (1970–2019). WMO No. 1267, ISBN: 978-92-63-11267-5.

An extratropical storm is a cold-core cyclone that derives its energy from the release of potential energy when cold and warm air masses interact. These storms have one or more fronts associated with them, and can occur over land or ocean. Winds vary and can be as weak as in a tropical depression, or as strong as a hurricane. Blizzards, Nor'easters, and ordinary low pressure systems are examples of extratropical storms (US NWS Glossary, accessed Nov 2021).

1.1.4 Severe convective storms

In this category, there are events that result from organized convection (e.g., supercells, squall lines, and multicell thunderstorm complexes), capable of causing extensive damages and injuries from tornadoes, damaging winds, lightning and thunder, heavy rain or large hail. Convection occurs when rising motions are driven by buoyancy, which happens when rising air is warmer than its surroundings and accelerates upward (Robinson, 2021). These events are local in nature and usually develop within few kilometers and in a short period of time, making their accurate forecast very challenging.

1.1.5 Extreme rainfall

Extreme (or heavy) rainfall is a result of extreme weather events, such as the ones described above. The specific mention of extreme rainfall here is to highlight its devastating impacts posed by rainfall-induced hazards such as floods and landslides, that threaten human lives and disrupt several functions of the society. Extreme rainfall, primarily expressed as intensity or accumulation over a period of time, is defined with respect to local rainfall climatology and therefore identifying a single global threshold for the definition of extremes is neither possible nor meaningful. The term extreme bares the notion of rarity, which is why extreme rainfall events are commonly defined with respect of their frequency. A practical way to describe rainfall extremes is using the Intensity-Duration-Frequency (IDF) curves, which, at a given location, directly relate the precipitation intensity averaged over a given duration to its probability of exceedance (Wurbs and James, 2002) but many other relevant metrics have been used in the literature. For example, extreme rainfall metrics used by the Climate Science Special Report of the Fourth National Climate Assessment (USGCRP, 2017) for a long-term assessment are: maximum daily precipitation in consecutive 5-year blocks; amount of precipitation falling in daily events that exceed the 99th percentile of all nonzero precipitation days; number of 2-day events with a precipitation total exceeding the largest 2-day amount that is expected to occur, on average, only once every 5 years, as calculated over a historic time period (i.e., 1901–2016). In the same report, observed increases in precipitation extremes over the United States were found by region from the 20-year return values of seasonal daily precipitation over the period 1948–2015 (Easterling et al., 2017). When forecasting extreme rainfall occurrences in the United States, named Excessive Rainfall Outlooks, the National Weather Service forecasts the probability that rainfall will exceed flash flood guidance within 40 km (25 miles) of a point. The European Environment Agency (EEA) defines an indicator of heavy precipitation as the maximum annual 5-day consecutive precipitation (EEA, 2021). As with other extreme weather metrics, extreme rainfall definition differs depending on the geographic location and time of the year. The WMO defines heavy rainfall as rainfall greater than or equal to 50 mm in the past 24 hours (https://severeweather.wmo.int/).

Subchapter 1.2

Weather forecasting

Marina Astitha¹, Linus Magnusson² and Efthymios Nikolopoulos³,    ¹Department of Civil and Environmental Engineering, School of Engineering, University of Connecticut, Storrs, CT, United States,    ²European Centre for Medium-Range Weather Forecasts-ECMWF, Reading, United Kingdom,    ³Civil and Environmental Engineering, Rutgers University, New Brunswick, NJ, United States

In this section, we provide a brief overview of weather forecasting techniques that include numerical prediction, statistical forecasting, and newly adopted machine learning and artificial intelligence approaches. For a detailed description of mathematical formulations related to weather prediction, we refer the reader to specialized books in the literature such as Pielke (1984), Kalnay (2002), Warner (2010), Coiffier (2012), Olafsson and Bao (2021), Robertson and Vitart (2018), NRC (2010), Inness and Dorling (2013), Palmer and Hagerdorn (2006), reprints of the originals such as Richardson (2007), among others.

The first part of this section aims to give a short introduction to numerical weather prediction (NWP) and to highlight the challenges that multiscale predictions of severe weather pose for an operational weather forecasting system. NWP solves an initial/boundary value problem posed by Navier–Stokes (motion) equations numerically, by iterating forward in time (Richardson, 2007). Modelers divide the Earth (for global models) or the region of interest (for limited-area models) in a three dimensional grid, apply the basic equations at each grid point, and calculate the state of the atmosphere for each time-step (i.e., temperature, pressure, wind, humidity, heat transfer, radiation, among others; Fig. 1.2.1). For this calculation, the initial and boundary conditions as well as some model parameters need to be determined beforehand and the equations need to be approximated into a numerical solver. The creation of an atmospheric forecast with an NWP model consists of the following steps: collection of observations, data assimilation of observations into the model grid, model integration and model output. Other considerations are ensemble forecasting, reforecasts and additional earth system components that can be coupled interactively to the atmospheric model (for example air-sea interactions; Fig. 1.2.2). Numerical prediction techniques, especially those in operational forecasting (short or long term forecast of weather conditions for the next few days to a week), are accompanied by statistical and mathematical methods to limit the prediction error, by assimilating observed data.

Figure 1.2.1 Illustration of Earth divided in a three-dimensional grid and physical processes that are represented in an atmospheric model (representation of a global model). Courtesy National Oceanic and Atmospheric Administration (NOAA).

Figure 1.2.2 Illustration of the components considered in an Earth System Model. Courtesy ECMWF.

Observations are necessary to create realistic initial conditions. Observations are collected from a range of sources and are often grouped into conventional and remote-sensing observations. Conventional observations are direct measurements of the state of the atmosphere from ground-based stations, soundings of the atmosphere from ascending balloons (radiosondes) or descending with a parachute (dropsondes), and measurements from aircrafts. The advantage of the conventional observations is that they are measuring quantities that are similar to the modeled parameters (temperature, winds, and humidity). The disadvantage is the heterogeneous distribution of the observations over the globe, with a good coverage over land but poor