About

Alex Anderson-Frey is a Professor in the Department of Atmospheric and Climate Science at the University of Washington. His areas of expertise include mesoscale meteorology, thunderstorms and tornadoes, synoptic meteorology, and statistical applications and machine learning. He teaches courses such as Hurricanes and Thunderstorms: Their Science and Impact, Methods of Atmospheric Data Analysis, and Instruments and Observations. His research focuses on understanding severe weather phenomena, particularly tornadoes, and improving warning skill through environmental analysis. Anderson-Frey has contributed to the field through numerous publications on tornado environments, warning systems, and environmental baselines for tornado warning skill. He has received several awards and fellowships, including the John C. Mather Nobel Scholarship from NASA and the Al and Betty Blackadar Graduate Scholarship in Meteorology. He is actively involved in professional service, including peer review for scientific journals, editorial roles, and participation in committees related to meteorology and climate science.

Research topics

• Geography
• Environmental science
• Meteorology
• Computer Science
• Engineering
• Geology
• Physics
• Atmospheric sciences

Selected publications

• Performance of the Pangu‐Weather Deep Learning Model in Forecasting Tornadic Environments

  Geophysical Research Letters · 2025-04-09 · 3 citations

  articleOpen accessSenior author

  Abstract The development of deep learning (DL) weather forecasting models has made rapid progress and achieved comparable or better skill than traditional Numerical Weather prediction (NWP) models, which are generally computationally intensive. However, applications of these DL models have yet to be fully explored, including for severe convective events. We evaluate the DL model Pangu‐Weather in forecasting tornadic environments with one‐day lead times using convective available potential energy (CAPE), 0–6 bulk wind difference (BWD6), and 0–3 km storm‐relative helicity (SRH3). We also compare its performance to the National Centers for Environmental Prediction (NCEP)'s Global Forecast System (GFS), a traditional NWP model. Pangu‐Weather generally outperforms GFS in predicting BWD6 and SRH3 at the closest grid point and hour of the storm report. However, Pangu‐Weather tends to underpredict the maximum values of all convective parameters in the 1–2 hr before the storm across the surrounding grid points compared to the GFS.

  PublisherOA PDFDOI

• Analyzing Tornado Warning Performance during Individual Storm Life Cycles

  Weather and Forecasting · 2023-03-22 · 2 citations

  articleSenior author

  Abstract The National Weather Service plays a critical role in alerting the public when dangerous weather occurs. Tornado warnings are one of the most publicly visible products the NWS issues given the large societal impacts tornadoes can have. Understanding the performance of these warnings is crucial for providing adequate warning during tornadic events and improving overall warning performance. This study aims to understand warning performance during the lifetimes of individual storms (specifically in terms of probability of detection and lead time). For example, does probability of detection vary based on if the tornado was the first produced by the storm, or the last? We use tornado outbreak data from 2008 to 2014, archived NEXRAD radar data, and the NWS verification database to associate each tornado report with a storm object. This approach allows for an analysis of warning performance based on the chronological order of tornado occurrence within each storm. Results show that the probability of detection and lead time increase with later tornadoes in the storm; the first tornadoes of each storm are less likely to be warned and on average have less lead time. Probability of detection also decreases overnight, especially for first tornadoes and storms that only produce one tornado. These results are important for understanding how tornado warning performance varies during individual storm life cycles and how upstream forecast products (e.g., Storm Prediction Center tornado watches, mesoscale discussions, etc.) may increase warning confidence for the first tornado produced by each storm. Significance Statement In this study, we focus on better understanding real-time tornado warning performance on a storm-by-storm basis. This approach allows us to examine how warning performance can change based on the order of each tornado within its parent storm. Using tornado reports, warning products, and radar data during tornado outbreaks from 2008 to 2014, we find that probability of detection and lead time increase with later tornadoes produced by the same storm. In other words, for storms that produce multiple tornadoes, the first tornado is generally the least likely to be warned in advance; when it is warned in advance, it generally contains less lead time than subsequent tornadoes. These findings provide important new analyses of tornado warning performance, particularly for the first tornado of each storm, and will help inform strategies for improving warning performance.

  PublisherDOI

• How Are Tornadic Supercell Soundings Significantly Different From Nearby Baseline Environments?

  Geophysical Research Letters · 2023-04-15 · 6 citations

  articleOpen accessSenior author

  Abstract This study explores how tornadic supercell soundings significantly differ from the same‐location and same‐hour baseline environment soundings, sampled from the days prior to or following the event. Permutation testing is used to identify whether sounding‐derived parameters mixed‐layer convective available potential energy and 0–1 km storm‐relative helicity are significantly different between the tornadic and baseline environment. Typically, in an environment with marginal values of certain key environmental parameters, anomalous values of those environmental parameters are more strongly associated with supercell tornadoes. Furthermore, many tornadic events already exhibit environmental conditions favorable for tornadic supercells a day prior to the event itself. Generally, supercell tornadoes that occur during typical peak tornadic activity time frames are easier to distinguish from baseline (non‐tornadic) environments compared to those occurring in other time frames. Spatiotemporal variations of distinguishability between tornadic and baseline environmental parameters add complexity to traditional parameter‐based fixed threshold forecasting.

  PublisherOA PDFDOI

• Comparing Performance: What Is (and Is Not) Tornado Warning Skill?

  Bulletin of the American Meteorological Society · 2022-12-01

  article1st authorCorresponding
  PublisherDOI

• Self-Organizing Maps for the Classification of Spatial and Temporal Variability of Tornado-Favorable Parameters

  Monthly Weather Review · 2021-12-07 · 19 citations

  articleSenior author

  Abstract A nuanced analysis of the spatial and temporal distribution of supercell tornadoes and the characteristics of the near-storm environments associated with those tornadoes is critical to improving our understanding of the range of environments that can be considered tornado favorable. This work classifies both supercell tornado probabilities and their associated environmental parameters on hourly and daily time scales based on geographical regions: regional probability of tornado events and the probability of deviation above or below the median tornadic near-storm environmental parameter values are estimated by kernel density estimation and classified by self-organizing maps (SOMs). The SOM classification for tornado probability allows for further examination of the deviation of the environmental parameters from the median for each probability cluster. Regions that have similar tornado probabilities but differ in the deviation of the environmental parameters (“parameter anomalies”) are also highlighted using SOMs. The anomaly patterns for different regions and parameters generally evolve along either seasonal or diurnal scales, but rarely both, highlighting the need for flexible models of tornado potential based on the near-storm environment. The spatial and temporal variability of parameter anomalies add complexity to traditional forecasting approaches that depend upon a fixed set of environmental parameter thresholds. This work highlights the need to develop region-specific and potentially time-specific environmental baseline evaluation to improve forecast and warning skill.

  PublisherDOI

• WSR-88D Tornado Intensity Estimates. Part I: Real-Time Probabilities of Peak Tornado Wind Speeds

  Weather and Forecasting · 2020 · 18 citations

  Senior authorCorresponding
  • Computer Science
  • Meteorology
  • Environmental science

  Abstract The Storm Prediction Center (SPC) has developed a database of damage-surveyed tornadoes in the contiguous United States (2009–17) that relates environmental and radar-derived storm attributes to damage ratings that change during a tornado life cycle. Damage indicators (DIs), and the associated wind speed estimates from tornado damage surveys compiled in the Damage Assessment Toolkit (DAT) dataset, were linked to the nearest manual calculations of 0.5° tilt angle maximum rotational velocity V rot from single-site WSR-88D data. For each radar scan, the maximum wind speed from the highest-rated DI, V rot , and the significant tornado parameter (STP) from the SPC hourly objective mesoscale analysis archive were recorded and analyzed. Results from examining V rot and STP data indicate an increasing conditional probability for higher-rated DIs (i.e., EF-scale wind speed estimate) as both STP and V rot increase. This work suggests that tornadic wind speed exceedance probabilities can be estimated in real time, on a scan-by-scan basis, via V rot and STP for ongoing tornadoes.

  PublisherOA PDFDOI

• WSR-88D Tornado Intensity Estimates. Part II: Real-Time Applications to Tornado Warning Time Scales

  Weather and Forecasting · 2020 · 12 citations

  Senior authorCorresponding
  • Environmental science
  • Meteorology
  • Atmospheric sciences

  Abstract A sample of damage-surveyed tornadoes in the contiguous United States (2009–17), containing specific wind speed estimates from damage indicators (DIs) within the Damage Assessment Toolkit dataset, were linked to radar-observed circulations using the nearest WSR-88D data in Part I of this work. The maximum wind speed associated with the highest-rated DI for each radar scan, corresponding 0.5° tilt angle rotational velocity V rot , significant tornado parameter (STP), and National Weather Service (NWS) convective impact-based warning (IBW) type, are analyzed herein for the sample of cases in Part I and an independent case sample from parts of 2019–20. As V rot and STP both increase, peak DI-estimated wind speeds and IBW warning type also tend to increase. Different combinations of V rot , STP, and population density—related to ranges of peak DI wind speed—exhibited a strong ability to discriminate across the tornado damage intensity spectrum. Furthermore, longer duration of high V rot (i.e., ≥70 kt) in significant tornado environments (i.e., STP ≥ 6) corresponds to increasing chances that DIs will reveal the occurrence of an intense tornado (i.e., EF3+). These findings were corroborated via the independent sample from parts of 2019–20, and can be applied in a real-time operational setting to assist in determining a potential range of wind speeds. This work provides evidence-based support for creating an objective and consistent, real-time framework for assessing and differentiating tornadoes across the tornado intensity spectrum.

  PublisherOA PDFDOI

• AMS Early Career Leadership Academy

  100th American Meteorological Society Annual Meeting · 2020-01-14

  article
  Publisher

• Compared to What? Establishing Environmental Baselines for Tornado Warning Skill

  Bulletin of the American Meteorological Society · 2020-12-02 · 12 citations

  articleOpen access1st authorCorresponding

  Abstract In any discussion of forecast evaluation, it is tempting to fall back on statements reflecting unverified assumptions: “this tornado warning had lower skill because the underlying meteorology reflected a complicated or atypical scenario,” or “that forecast performed worse than we would have expected given the straightforward setup.” These statements of what is and is not a reasonable expectation for warning skill are particularly relevant as the meteorological community’s focus has begun to emphasize non-classic storm environments (e.g., tornadoes spawned by quasi-linear convective systems). In this paper, we build a proof-of-concept methodology to quantify the effect of the near-storm environment on tornado warning skill, and we then test these methods on a 15-yr dataset composed of tens of thousands of tornado events and warnings over the contiguous United States. Our findings include that significant tornadoes rated (E)F2+ have a higher probability of detection (POD) than expected based on their near-storm environments, that nocturnal tornadoes have both worse POD and false alarm ratio (FAR) than even their marginal near-storm environments would suggest, and that tornadoes occurring during the summer months also show worse POD and FAR than their environment-based expectation. Quantifying these shifts in performance in an environmental skill score framework allows us to target the situations in which the greatest improvements may be possible, in terms of forecaster training and/or conceptual models. This work also highlights the essential question that should always be asked in the context of forecast verification: what, exactly, is the baseline standard to which we are comparing forecast performance?

  PublisherOA PDFDOI

• Tornado Fatalities: An Environmental Perspective

  Weather and Forecasting · 2019-10-22 · 31 citations

  articleOpen access1st authorCorresponding

  Abstract Deadly tornadoes are rare events, but that level of rarity varies with many factors. In this work, we summarize and update past research on tornado fatalities, and also discuss the environments of deadly tornadoes both from the perspective of proximity soundings (i.e., point-based) and self-organizing maps (i.e., two-dimensional). In our study of 16 232 tornado events from 2003 to 2017, we find that deadly tornadoes are disproportionately likely to have high (E)F-scale ratings, to have right-moving supercell parent storm modes (deadly QLCS tornadoes are exceptionally rare and tend to result in only one death when they do occur), and to occur during the winter and spring. Warning skill is generally higher for deadly tornadoes than for nondeadly tornadoes: 87% of deadly tornadoes were warned in advance, and nearly 95% of tornado deaths occurred within an active warning. The same environments are warned well for both deadly and nondeadly tornadoes, but the deadly tornadoes tend to occur in environments that are less conducive to weaker (E)F0–1 tornadoes. We identify four prototypical deadly tornado scenarios using self-organizing maps, ranging from marginal environments resulting in relatively few fatalities to major deadly outbreak events. Overall results indicate that the most dangerous tornadoes (i.e., those with high numbers of deaths per deadly tornado) also generally occur in environments and under conditions in which warning skill is high. While, generally speaking, the correct storms are being warned, we include some recommendations for additional research and further improvement.

  PublisherOA PDFDOI

Recent grants

• Collaborative Research: Combining Self-organized Maps and Idealized Storm-scale Simulations to Investigate the Effect of Future Climate Change on Severe Convective Storms

  NSF · $423k · 2022–2026

Frequent coauthors

• Andrew R. Dean

  NOAA Storm Prediction Center

  35 shared

• Richard L. Thompson

  NOAA Storm Prediction Center

  35 shared

• Bryan T. Smith

  NOAA Storm Prediction Center

  35 shared

• Yvette Richardson

  Pennsylvania State University

  31 shared

• Harold E. Brooks

  University of Oklahoma

  20 shared

• Christopher D. Karstens

  University of Oklahoma

  4 shared

• Douglas A. Speheger

  NOAA National Weather Service

  4 shared

• Makenzie Krocak

  University of Oklahoma

  3 shared

Labs

• Climate Impacts GroupPI

Education

• PhD Meteorology, Meteorology

  Pennsylvania State University

  2017

• MSc Atmospheric Science, Atmospheric and Oceanic Sciences

  McGill University

  2014

• BSc Honours Atmospheric Science, Atmospheric Science

  University of Alberta

  2010

Awards & honors

• NRC Research Associateship (declined, 2018)
• Al and Betty Blackadar Graduate Scholarship in Meteorology,…
• Postgraduate Scholarship, Natural Sciences and Engineering R…
• Anne C. Wilson Graduate Student Research Award, PSU (2012-20…
• Dr. Dennis W. and Joan S. Thomson Distinguished Graduate Fel…