Browse
Abstract
A user-focused verification approach for evaluating probability forecasts of binary outcomes (also known as probabilistic classifiers) is demonstrated that (i) is based on proper scoring rules, (ii) focuses on user decision thresholds, and (iii) provides actionable insights. It is argued that when categorical performance diagrams and the critical success index are used to evaluate overall predictive performance, rather than the discrimination ability of probabilistic forecasts, they may produce misleading results. Instead, Murphy diagrams are shown to provide a better understanding of the overall predictive performance as a function of user probabilistic decision threshold. We illustrate how to select a proper scoring rule, based on the relative importance of different user decision thresholds, and how this choice impacts scores of overall predictive performance and supporting measures of discrimination and calibration. These approaches and ideas are demonstrated using several probabilistic thunderstorm forecast systems as well as synthetic forecast data. Furthermore, a fair method for comparing the performance of probabilistic and categorical forecasts is illustrated using the fixed risk multicategorical (FIRM) score, which is a proper scoring rule directly connected to values on the Murphy diagram. While the methods are illustrated using thunderstorm forecasts, they are applicable for evaluating probabilistic forecasts for any situation with binary outcomes.
Significance Statement
Recently, several papers have presented verification results for probabilistic forecasts using so-called categorical performance diagrams, which summarize multiple verification metrics. While categorical performance diagrams measure discrimination ability, we demonstrate how they can potentially lead to incorrect conclusions when evaluating overall predictive performance of probabilistic forecasts. By reviewing recent advances in the statistical literature, we show a comprehensive approach for the meteorological community that (i) does not reward a forecaster who “hedges” their forecast, (ii) focuses on the importance of the forecast user’s decision threshold(s), and (iii) provides actionable insights. Additionally, we present an approach for fairly comparing the skill of categorical forecasts to probabilistic forecasts.
Abstract
A user-focused verification approach for evaluating probability forecasts of binary outcomes (also known as probabilistic classifiers) is demonstrated that (i) is based on proper scoring rules, (ii) focuses on user decision thresholds, and (iii) provides actionable insights. It is argued that when categorical performance diagrams and the critical success index are used to evaluate overall predictive performance, rather than the discrimination ability of probabilistic forecasts, they may produce misleading results. Instead, Murphy diagrams are shown to provide a better understanding of the overall predictive performance as a function of user probabilistic decision threshold. We illustrate how to select a proper scoring rule, based on the relative importance of different user decision thresholds, and how this choice impacts scores of overall predictive performance and supporting measures of discrimination and calibration. These approaches and ideas are demonstrated using several probabilistic thunderstorm forecast systems as well as synthetic forecast data. Furthermore, a fair method for comparing the performance of probabilistic and categorical forecasts is illustrated using the fixed risk multicategorical (FIRM) score, which is a proper scoring rule directly connected to values on the Murphy diagram. While the methods are illustrated using thunderstorm forecasts, they are applicable for evaluating probabilistic forecasts for any situation with binary outcomes.
Significance Statement
Recently, several papers have presented verification results for probabilistic forecasts using so-called categorical performance diagrams, which summarize multiple verification metrics. While categorical performance diagrams measure discrimination ability, we demonstrate how they can potentially lead to incorrect conclusions when evaluating overall predictive performance of probabilistic forecasts. By reviewing recent advances in the statistical literature, we show a comprehensive approach for the meteorological community that (i) does not reward a forecaster who “hedges” their forecast, (ii) focuses on the importance of the forecast user’s decision threshold(s), and (iii) provides actionable insights. Additionally, we present an approach for fairly comparing the skill of categorical forecasts to probabilistic forecasts.
Abstract
Though discrete supercells are usually emphasized in severe weather forecasting, hazard production is often preceded by their interaction with external features. Past studies have examined the impacts of cell mergers, boundaries, other supercells, convective systems, etc., but usually in isolation. Here, we investigate 230 significant tornadoes, 246 significant hail events, and 191 null cases across the United States using WSR-88D data. We find that in over 90% of cases, supercells that produced significant hazards were accompanied by external features. These features varied between hazards; for example, hailstorms were more frequently near boundaries than tornadic storms. That said, the positions of these features with respect to the storm (and storm-relative inflow) distinguished between hazard potential and type. For example, tornadic storms were predominantly on the more-unstable side of a boundary, while non-tornadic storms and hailstorms were on the less-unstable side. Similarly, tornadic storms had more cells in their rear flanks than forward flanks, while hailstorms had more cells in their forward flanks than rear flanks. Although these conditions were observed regardless of the background environment, they were affected by certain variables in the vertical profile, especially in tornadic cases. Namely, when storm-relative inflow was stronger and lifting condensation level (LCL) was lower, tornadic storms were accompanied by more rear-flank cells that were closer to the storm, more directly opposite the storm-relative inflow, for a longer period of time. We propose that these interactions likely modulate hazard potential, in ways that are not accounted for in traditional environmental parameter-based forecasting.
Abstract
Though discrete supercells are usually emphasized in severe weather forecasting, hazard production is often preceded by their interaction with external features. Past studies have examined the impacts of cell mergers, boundaries, other supercells, convective systems, etc., but usually in isolation. Here, we investigate 230 significant tornadoes, 246 significant hail events, and 191 null cases across the United States using WSR-88D data. We find that in over 90% of cases, supercells that produced significant hazards were accompanied by external features. These features varied between hazards; for example, hailstorms were more frequently near boundaries than tornadic storms. That said, the positions of these features with respect to the storm (and storm-relative inflow) distinguished between hazard potential and type. For example, tornadic storms were predominantly on the more-unstable side of a boundary, while non-tornadic storms and hailstorms were on the less-unstable side. Similarly, tornadic storms had more cells in their rear flanks than forward flanks, while hailstorms had more cells in their forward flanks than rear flanks. Although these conditions were observed regardless of the background environment, they were affected by certain variables in the vertical profile, especially in tornadic cases. Namely, when storm-relative inflow was stronger and lifting condensation level (LCL) was lower, tornadic storms were accompanied by more rear-flank cells that were closer to the storm, more directly opposite the storm-relative inflow, for a longer period of time. We propose that these interactions likely modulate hazard potential, in ways that are not accounted for in traditional environmental parameter-based forecasting.
Abstract
Recent operationally driven research has generated a framework, known as the three ingredients method and mesovortex warning system, that can help forecasters anticipate mesovortex development and issue warnings within quasi-linear convective systems (QLCSs). However, dual-polarization radar data has not yet been incorporated into this framework. Therefore, several dual- and single-polarization radar signatures associated with QLCS mesovortices were analyzed to determine if they could provide additional information about mesovortex development and intensity. An analysis of 167 mesovortices showed that 1) K DP drops precede ∼95% of mesovortices and provide an initial indication of where a mesovortex may develop; 2) midlevel K DP cores are a potentially useful precursor signature because they precede a majority of mesovortices and have higher magnitudes for mesovortices that produce wind damage or tornadoes; 3) low-level K DP cores and areas of enhanced spectrum width have higher magnitudes for mesovortices that produce wind damage or tornadoes but tend to develop at about the same time as the mesovortex, which makes them more useful as diagnostic than as predictive signatures; and 4) as range from the radar increases, the radar signatures become less useful in anticipating mesovortex intensity but can still be used to anticipate mesovortex development or build confidence in mesovortex existence.
Significance Statement
The purpose of this study is to look at weather radar features that might help forecasters predict the development and intensity of tornadoes and strong winds within linear thunderstorm systems. Our results show that the intensity and trends of some radar features are helpful in showing when these hazards might develop and how strong they might be, while other radar features are less helpful. This information can help forecasters focus on the most useful radar features and ultimately provide the best possible warnings.
Abstract
Recent operationally driven research has generated a framework, known as the three ingredients method and mesovortex warning system, that can help forecasters anticipate mesovortex development and issue warnings within quasi-linear convective systems (QLCSs). However, dual-polarization radar data has not yet been incorporated into this framework. Therefore, several dual- and single-polarization radar signatures associated with QLCS mesovortices were analyzed to determine if they could provide additional information about mesovortex development and intensity. An analysis of 167 mesovortices showed that 1) K DP drops precede ∼95% of mesovortices and provide an initial indication of where a mesovortex may develop; 2) midlevel K DP cores are a potentially useful precursor signature because they precede a majority of mesovortices and have higher magnitudes for mesovortices that produce wind damage or tornadoes; 3) low-level K DP cores and areas of enhanced spectrum width have higher magnitudes for mesovortices that produce wind damage or tornadoes but tend to develop at about the same time as the mesovortex, which makes them more useful as diagnostic than as predictive signatures; and 4) as range from the radar increases, the radar signatures become less useful in anticipating mesovortex intensity but can still be used to anticipate mesovortex development or build confidence in mesovortex existence.
Significance Statement
The purpose of this study is to look at weather radar features that might help forecasters predict the development and intensity of tornadoes and strong winds within linear thunderstorm systems. Our results show that the intensity and trends of some radar features are helpful in showing when these hazards might develop and how strong they might be, while other radar features are less helpful. This information can help forecasters focus on the most useful radar features and ultimately provide the best possible warnings.
Abstract
On 8 August 2023, a wind-driven wildfire pushed across the city of Lahaina, located in West Maui, Hawaii, resulting in at least 100 deaths and an estimated economic loss of 4–6 billion dollars. The Lahaina wildfire was associated with strong, dry downslope winds gusting to 31–41 m s−1 (60–80 kt; 1 kt ≈ 0.51 m s−1) that initiated the fire by damaging power infrastructure. The fire spread rapidly in invasive grasses growing in abandoned agricultural land upslope from Lahaina. This paper describes the synoptic and mesoscale meteorology associated with this event, as well as its predictability. Stronger-than-normal northeast trade winds, accompanied by a stable layer near the crest level of the West Maui Mountains, resulted in a high-amplitude mountain-wave response and a strong downslope windstorm. Mesoscale model predictions were highly accurate regarding the location, strength, and timing of the strong winds. Hurricane Dora, which passed approximately 1300 km to the south of Maui, does not appear to have had a significant impact on the occurrence and intensity of the winds associated with the wildfire event. The Maui wildfire was preceded by a wetter-than-normal winter and near-normal summer conditions.
Significance Statement
The 2023 Maui wildfire was one of the most damaging of the past century, with at least 100 fatalities. This paper describes the meteorological conditions associated with the event and demonstrates that excellent model forecasts made the threat foreseeable.
Abstract
On 8 August 2023, a wind-driven wildfire pushed across the city of Lahaina, located in West Maui, Hawaii, resulting in at least 100 deaths and an estimated economic loss of 4–6 billion dollars. The Lahaina wildfire was associated with strong, dry downslope winds gusting to 31–41 m s−1 (60–80 kt; 1 kt ≈ 0.51 m s−1) that initiated the fire by damaging power infrastructure. The fire spread rapidly in invasive grasses growing in abandoned agricultural land upslope from Lahaina. This paper describes the synoptic and mesoscale meteorology associated with this event, as well as its predictability. Stronger-than-normal northeast trade winds, accompanied by a stable layer near the crest level of the West Maui Mountains, resulted in a high-amplitude mountain-wave response and a strong downslope windstorm. Mesoscale model predictions were highly accurate regarding the location, strength, and timing of the strong winds. Hurricane Dora, which passed approximately 1300 km to the south of Maui, does not appear to have had a significant impact on the occurrence and intensity of the winds associated with the wildfire event. The Maui wildfire was preceded by a wetter-than-normal winter and near-normal summer conditions.
Significance Statement
The 2023 Maui wildfire was one of the most damaging of the past century, with at least 100 fatalities. This paper describes the meteorological conditions associated with the event and demonstrates that excellent model forecasts made the threat foreseeable.
Abstract
Development of an impact-based decision support forecasting tool for surface-transportation hazards requires consideration for what impacts the product is intended to capture and how to scale forecast information to impacts to then categorize impact severity. In this first part of the series, we discuss the motivation and intent of such a product, in addition to outlining the approach we take to leverage existing and new research to develop the product. Traffic disruptions (e.g., crashes, increased travel times, roadway restrictions, or closures) are the intended impacts, where impact severity levels are intended to scale to reflect the increasing severity of adverse driving conditions that can correlate with a need for enhanced mitigation efforts by motorists and/or transportation agencies (e.g., slowing down, avoiding travel, and imposing roadway restrictions or closures). Previous research on how weather and road conditions impact transportation and novel research herein to create a metric for crash impact based on precipitation type and local hour of the day are both intended to help scale weather forecasts to impacts. Impact severity classifications can ultimately be determined through consideration of any thresholds used by transportation agencies, in conjunction with the scaling metrics.
Significance Statement
Weather can profoundly impact surface transportation and motorist safety. Because of this and because there are no explicit tools available to forecasters to identify and communicate potential impacts to surface transportation, there is a desire for the development of such a forecast product. However, doing so requires careful consideration for what impacts are intended to be included, how weather corresponds to impacts, and how thresholds for impact severity should be defined. In this first part of the paper series, we outline each of these aspects and present novel research and approaches for the development of an impact-based forecast product specifically tailored to surface-transportation hazards. The product is ultimately intended to improve motorist safety and mobility on roads.
Abstract
Development of an impact-based decision support forecasting tool for surface-transportation hazards requires consideration for what impacts the product is intended to capture and how to scale forecast information to impacts to then categorize impact severity. In this first part of the series, we discuss the motivation and intent of such a product, in addition to outlining the approach we take to leverage existing and new research to develop the product. Traffic disruptions (e.g., crashes, increased travel times, roadway restrictions, or closures) are the intended impacts, where impact severity levels are intended to scale to reflect the increasing severity of adverse driving conditions that can correlate with a need for enhanced mitigation efforts by motorists and/or transportation agencies (e.g., slowing down, avoiding travel, and imposing roadway restrictions or closures). Previous research on how weather and road conditions impact transportation and novel research herein to create a metric for crash impact based on precipitation type and local hour of the day are both intended to help scale weather forecasts to impacts. Impact severity classifications can ultimately be determined through consideration of any thresholds used by transportation agencies, in conjunction with the scaling metrics.
Significance Statement
Weather can profoundly impact surface transportation and motorist safety. Because of this and because there are no explicit tools available to forecasters to identify and communicate potential impacts to surface transportation, there is a desire for the development of such a forecast product. However, doing so requires careful consideration for what impacts are intended to be included, how weather corresponds to impacts, and how thresholds for impact severity should be defined. In this first part of the paper series, we outline each of these aspects and present novel research and approaches for the development of an impact-based forecast product specifically tailored to surface-transportation hazards. The product is ultimately intended to improve motorist safety and mobility on roads.
Abstract
In line with the continued focus of the National Weather Service (NWS) to provide impact-based decision support services (IDSS) and effectively communicate potential impacts, a new IDSS forecasting tool for surface-transportation hazards is in development at the Weather Prediction Center: the hourly winter storm severity index (WSSI-H). This second part of the series outlines the current algorithms and thresholds for the components of the WSSI-H, which has been developed in line with the approach and considerations discussed in Part I of this series. These components—snow amount, ice accumulation, snow rate, liquid rate, and blowing and drifting snow—each address a specific hazard for motorists. The inclusion of metrics related to driving conditions for untreated road surfaces and time-of-day factoring for active precipitation types helps directly tie forecasted weather conditions to transportation impacts. Impact severity level thresholds are approximately in line with thresholds used by transportation agencies when considering various mitigation strategies (e.g., imposing speed restrictions or closing roadways). Whereas the product is not meant to forecast specific impacts (e.g., road closure or pileup), impact severity levels are designed to scale with increasingly poor travel conditions, which can prompt various mitigation efforts from motorists or transportation agencies to maintain safety. WSSI-H outputs for three winter events are discussed in depth to highlight the potential utility of the product. Overall, the WSSI-H is intended to provide high-resolution situational awareness of potential surface-transportation-related impacts and aid in enhanced collaborations between NWS forecasters and stakeholders like transportation agencies to improve motorist safety.
Significance Statement
A new impact-based forecast product designed to aid in situational awareness of potential impacts from surface-transportation-related hazards is in development. In this second part of the series, we outline the algorithms and thresholds for the various components of the product, where each component addresses a unique hazard. Product outputs for three winter events are presented to highlight the potential utility of the product in an operational forecast setting. Ultimately, enhanced collaboration between forecasters and transportation agencies alongside guidance from this product will bolster consistent messaging to motorists and improve safety and mobility on roads.
Abstract
In line with the continued focus of the National Weather Service (NWS) to provide impact-based decision support services (IDSS) and effectively communicate potential impacts, a new IDSS forecasting tool for surface-transportation hazards is in development at the Weather Prediction Center: the hourly winter storm severity index (WSSI-H). This second part of the series outlines the current algorithms and thresholds for the components of the WSSI-H, which has been developed in line with the approach and considerations discussed in Part I of this series. These components—snow amount, ice accumulation, snow rate, liquid rate, and blowing and drifting snow—each address a specific hazard for motorists. The inclusion of metrics related to driving conditions for untreated road surfaces and time-of-day factoring for active precipitation types helps directly tie forecasted weather conditions to transportation impacts. Impact severity level thresholds are approximately in line with thresholds used by transportation agencies when considering various mitigation strategies (e.g., imposing speed restrictions or closing roadways). Whereas the product is not meant to forecast specific impacts (e.g., road closure or pileup), impact severity levels are designed to scale with increasingly poor travel conditions, which can prompt various mitigation efforts from motorists or transportation agencies to maintain safety. WSSI-H outputs for three winter events are discussed in depth to highlight the potential utility of the product. Overall, the WSSI-H is intended to provide high-resolution situational awareness of potential surface-transportation-related impacts and aid in enhanced collaborations between NWS forecasters and stakeholders like transportation agencies to improve motorist safety.
Significance Statement
A new impact-based forecast product designed to aid in situational awareness of potential impacts from surface-transportation-related hazards is in development. In this second part of the series, we outline the algorithms and thresholds for the various components of the product, where each component addresses a unique hazard. Product outputs for three winter events are presented to highlight the potential utility of the product in an operational forecast setting. Ultimately, enhanced collaboration between forecasters and transportation agencies alongside guidance from this product will bolster consistent messaging to motorists and improve safety and mobility on roads.
Abstract
The impact of assimilating targeted Uncrewed Aircraft System (UAS) observations on the prediction of radiation and river-valley fog is assessed using Observing System Experiments (OSEs). Multi-rotor UAS were deployed during FOGMAP (Frequent in situ Observations Above Ground for Modeling and Advanced Prediction of fog) which took place during the summer of 2022 in northern Kentucky. Targeted UAS missions were flown to sample the spatio-temporal variability of temperature and moisture in the vicinity of the Cincinnati/Northern Kentucky International Airport. During each mission the UAS performed near-continuous profiling at two locations between the surface to 120 m AGL throughout the night. Data denial experiments were performed using the Ensemble Kalman Adjustment Filter available in NSF NCAR’s Data Assimilation Research Testbed (DART) to determine the impact of assimilating UAS observations on the skill of analyses and forecasts issued during potential fog events. Simulations that only assimilated conventional observations tended to have a dry bias in the analyses and forecasts. The dry bias in the analyses was reduced in experiments that assimilated UAS observations leading to improved probabilistic predictions of fog. Sensitivity tests revealed that the ensemble mean analyses were improved when assimilating UAS observations of specific humidity rather than relative humidity (RH) due to the existence of a cold bias near the surface and the negative covariance between RH and temperature. It was also found that either the assumed observation error variance of (1 g kg−1)2 or the ensemble spread of the background specific humidity was too large since their sum tended to overestimate the Root Mean Squared Error (RMSE) of the predicted ensemble mean values.
Abstract
The impact of assimilating targeted Uncrewed Aircraft System (UAS) observations on the prediction of radiation and river-valley fog is assessed using Observing System Experiments (OSEs). Multi-rotor UAS were deployed during FOGMAP (Frequent in situ Observations Above Ground for Modeling and Advanced Prediction of fog) which took place during the summer of 2022 in northern Kentucky. Targeted UAS missions were flown to sample the spatio-temporal variability of temperature and moisture in the vicinity of the Cincinnati/Northern Kentucky International Airport. During each mission the UAS performed near-continuous profiling at two locations between the surface to 120 m AGL throughout the night. Data denial experiments were performed using the Ensemble Kalman Adjustment Filter available in NSF NCAR’s Data Assimilation Research Testbed (DART) to determine the impact of assimilating UAS observations on the skill of analyses and forecasts issued during potential fog events. Simulations that only assimilated conventional observations tended to have a dry bias in the analyses and forecasts. The dry bias in the analyses was reduced in experiments that assimilated UAS observations leading to improved probabilistic predictions of fog. Sensitivity tests revealed that the ensemble mean analyses were improved when assimilating UAS observations of specific humidity rather than relative humidity (RH) due to the existence of a cold bias near the surface and the negative covariance between RH and temperature. It was also found that either the assumed observation error variance of (1 g kg−1)2 or the ensemble spread of the background specific humidity was too large since their sum tended to overestimate the Root Mean Squared Error (RMSE) of the predicted ensemble mean values.
Abstract
Short-lived and poorly organized convective cells, often called weakly forced thunderstorms (WFTs), are a common phenomenon during the warm season across the eastern and southeastern United States. While typically benign, wet downbursts emanating from such convection can have substantial societal impacts, including tree, power line, and property damage from strong outflow winds. Observational studies have documented the occurrence of severe (25.7 m s−1 or higher) wind speeds from wet downbursts, but the frequency of severe downbursts, including the spatial extent and temporal duration of severe winds, remains unclear. The ability for modern observing networks to reliably observe such events is also unknown; however, answering these questions is important for improving forecast skill and verifying convective warnings accurately. This study attempts to answer these questions by drawing statistical inferences from 97 high-resolution idealized simulations of single-cell downburst events. It was found that while 35% of the simulations featured severe winds, the spatial and temporal extent of such winds is limited—O(10) km2 or less and persisting for around 5 min on average. Furthermore, through a series of simulated network experiments, it is postulated that the probability that a modern mesonet observes a severe wind gust given a severe downburst is around 1%. From these results, a statistical argument is made that most tree impacts associated with pulse convection are likely caused by subsevere winds. Several implications for forecasting, warning, and verifying WFT events fall out from these discussions.
Abstract
Short-lived and poorly organized convective cells, often called weakly forced thunderstorms (WFTs), are a common phenomenon during the warm season across the eastern and southeastern United States. While typically benign, wet downbursts emanating from such convection can have substantial societal impacts, including tree, power line, and property damage from strong outflow winds. Observational studies have documented the occurrence of severe (25.7 m s−1 or higher) wind speeds from wet downbursts, but the frequency of severe downbursts, including the spatial extent and temporal duration of severe winds, remains unclear. The ability for modern observing networks to reliably observe such events is also unknown; however, answering these questions is important for improving forecast skill and verifying convective warnings accurately. This study attempts to answer these questions by drawing statistical inferences from 97 high-resolution idealized simulations of single-cell downburst events. It was found that while 35% of the simulations featured severe winds, the spatial and temporal extent of such winds is limited—O(10) km2 or less and persisting for around 5 min on average. Furthermore, through a series of simulated network experiments, it is postulated that the probability that a modern mesonet observes a severe wind gust given a severe downburst is around 1%. From these results, a statistical argument is made that most tree impacts associated with pulse convection are likely caused by subsevere winds. Several implications for forecasting, warning, and verifying WFT events fall out from these discussions.
Abstract
There is growing interest in impact-based decision support services to address complex decision-making, especially for winter storm forecasting. Understanding users’ needs for winter storm forecast information is necessary to make such impact-based winter forecasts relevant and useful to the diverse regions affected. A mixed-method social science research study investigated extending the winter storm severity index (WSSI) [operational for the contiguous United States (CONUS)] to Alaska, with consideration of the distinct needs of Alaskan stakeholders and the Alaskan climate. Data availability differences suggest the need for an Alaska-specific WSSI, calling for user feedback to inform the direction of product modifications. Focus groups and surveys in six regions of Alaska provided information on how the WSSI components, definitions, and categorization of impacts could align with stakeholder expectations and led to recommendations for the Weather Prediction Center to consider in developing the WSSI Alaska product. Overall, wind (strength and direction) and precipitation are key components to include. Air travel is a critical concern requiring wind and visibility information, while road travel is less emphasized (contrasting with CONUS needs). Special Weather Statements and Winter Storm Warnings are highly valued, and storm trajectory and transition (between precipitation types) information are the important contexts for decision-makers. Alaska is accustomed to and prepared for winter impacts but being able to understand how components (wind, snow, and ice) contribute to overall impact enhances the ability to respond and mitigate damage effectively. The WSSI adapted for Alaska can help address regional forecast needs, particularly valuable as the climate changes and typical winter conditions become more variable.
Significance Statement
Impact-based support services can assist decision-makers in prioritizing preparedness and mitigation actions related to winter storm events. The winter storm severity index adapted for specific considerations in Alaska (such as including wind and visibility components) can extend winter weather impact-based forecasting’s utility. Additionally, lessons learned from the process of adapting a national product to specific regional needs may inform best practices for gathering stakeholder input and feedback.
Abstract
There is growing interest in impact-based decision support services to address complex decision-making, especially for winter storm forecasting. Understanding users’ needs for winter storm forecast information is necessary to make such impact-based winter forecasts relevant and useful to the diverse regions affected. A mixed-method social science research study investigated extending the winter storm severity index (WSSI) [operational for the contiguous United States (CONUS)] to Alaska, with consideration of the distinct needs of Alaskan stakeholders and the Alaskan climate. Data availability differences suggest the need for an Alaska-specific WSSI, calling for user feedback to inform the direction of product modifications. Focus groups and surveys in six regions of Alaska provided information on how the WSSI components, definitions, and categorization of impacts could align with stakeholder expectations and led to recommendations for the Weather Prediction Center to consider in developing the WSSI Alaska product. Overall, wind (strength and direction) and precipitation are key components to include. Air travel is a critical concern requiring wind and visibility information, while road travel is less emphasized (contrasting with CONUS needs). Special Weather Statements and Winter Storm Warnings are highly valued, and storm trajectory and transition (between precipitation types) information are the important contexts for decision-makers. Alaska is accustomed to and prepared for winter impacts but being able to understand how components (wind, snow, and ice) contribute to overall impact enhances the ability to respond and mitigate damage effectively. The WSSI adapted for Alaska can help address regional forecast needs, particularly valuable as the climate changes and typical winter conditions become more variable.
Significance Statement
Impact-based support services can assist decision-makers in prioritizing preparedness and mitigation actions related to winter storm events. The winter storm severity index adapted for specific considerations in Alaska (such as including wind and visibility components) can extend winter weather impact-based forecasting’s utility. Additionally, lessons learned from the process of adapting a national product to specific regional needs may inform best practices for gathering stakeholder input and feedback.
