Browse
Abstract
Lightning strikes pose a hazard to human life and property, and can be difficult to forecast in a timely manner. In this study, a satellite-based machine learning model was developed to provide objective, short-term, location-specific probabilistic guidance for next-hour lightning activity. Using a convolutional neural network architecture designed for semantic segmentation, the model was trained using GOES-16 visible, shortwave infrared, and longwave infrared bands from the Advanced Baseline Imager (ABI). Next-hour GOES-16 Geostationary Lightning Mapper data were used as the truth or target data. The model, known as LightningCast, was trained over the GOES-16 ABI contiguous United States (CONUS) domain. However, the model is shown to generalize to GOES-16 full disk regions that are outside of the CONUS. LightningCast provides predictions for developing and advecting storms, regardless of solar illumination and meteorological conditions. LightningCast, which frequently provides 20 min or more of lead time to new lightning activity, learned salient features consistent with the scientific understanding of the relationships between lightning and satellite imagery interpretation. We also demonstrate that despite being trained on data from a single geostationary satellite domain (GOES-East), the model can be applied to other satellites (e.g., GOES-West) with comparable specifications and without substantial degradation in performance. LightningCast objectively transforms large volumes of satellite imagery into objective, actionable information. Potential application areas are also highlighted.
Significance Statement
The outcome of this research is a model that spatially forecasts lightning occurrence in a 0–60-min time window, using only images of clouds from the GOES-R Advanced Baseline Imager. This model has the potential to provide early alerts for developing and approaching hazardous conditions.
Abstract
Lightning strikes pose a hazard to human life and property, and can be difficult to forecast in a timely manner. In this study, a satellite-based machine learning model was developed to provide objective, short-term, location-specific probabilistic guidance for next-hour lightning activity. Using a convolutional neural network architecture designed for semantic segmentation, the model was trained using GOES-16 visible, shortwave infrared, and longwave infrared bands from the Advanced Baseline Imager (ABI). Next-hour GOES-16 Geostationary Lightning Mapper data were used as the truth or target data. The model, known as LightningCast, was trained over the GOES-16 ABI contiguous United States (CONUS) domain. However, the model is shown to generalize to GOES-16 full disk regions that are outside of the CONUS. LightningCast provides predictions for developing and advecting storms, regardless of solar illumination and meteorological conditions. LightningCast, which frequently provides 20 min or more of lead time to new lightning activity, learned salient features consistent with the scientific understanding of the relationships between lightning and satellite imagery interpretation. We also demonstrate that despite being trained on data from a single geostationary satellite domain (GOES-East), the model can be applied to other satellites (e.g., GOES-West) with comparable specifications and without substantial degradation in performance. LightningCast objectively transforms large volumes of satellite imagery into objective, actionable information. Potential application areas are also highlighted.
Significance Statement
The outcome of this research is a model that spatially forecasts lightning occurrence in a 0–60-min time window, using only images of clouds from the GOES-R Advanced Baseline Imager. This model has the potential to provide early alerts for developing and approaching hazardous conditions.
Abstract
This research begins the process of creating an ensemble-based forecast system for smoke aerosols generated from wildfires using a modified version of the National Severe Storms Laboratory (NSSL) Warn-on-Forecast System (WoFS). The existing WoFS has proven effective in generating short-term (0–3 h) probabilistic forecasts of high-impact weather events such as storm rotation, hail, severe winds, and heavy rainfall. However, it does not include any information on large smoke plumes generated from wildfires that impact air quality and the surrounding environment. The prototype WoFS-Smoke system is based on the deterministic High-Resolution Rapid Refresh-Smoke (HRRR-Smoke) model. HRRR-Smoke runs over a continental United States (CONUS) domain with a 3-km horizontal grid spacing, with hourly forecasts out to 48 h. The smoke plume injection algorithm in HRRR-Smoke is integrated into the WoFS forming WOFS-Smoke so that the advantages of the rapidly cycling, ensemble-based WoFS can be used to generate short-term (0–3 h) probabilistic forecasts of smoke. WoFS-Smoke forecasts from three wildfire cases during 2020 show that the system generates a realistic representation of wildfire smoke when compared against satellite observations. Comparison of smoke forecasts with radar data show that forecast smoke reaches higher levels than radar-detected debris, but exceptions to this are noted. The radiative effect of smoke on surface temperature forecasts is evident, which reduces forecast errors compared to experiments that do not include smoke.
Abstract
This research begins the process of creating an ensemble-based forecast system for smoke aerosols generated from wildfires using a modified version of the National Severe Storms Laboratory (NSSL) Warn-on-Forecast System (WoFS). The existing WoFS has proven effective in generating short-term (0–3 h) probabilistic forecasts of high-impact weather events such as storm rotation, hail, severe winds, and heavy rainfall. However, it does not include any information on large smoke plumes generated from wildfires that impact air quality and the surrounding environment. The prototype WoFS-Smoke system is based on the deterministic High-Resolution Rapid Refresh-Smoke (HRRR-Smoke) model. HRRR-Smoke runs over a continental United States (CONUS) domain with a 3-km horizontal grid spacing, with hourly forecasts out to 48 h. The smoke plume injection algorithm in HRRR-Smoke is integrated into the WoFS forming WOFS-Smoke so that the advantages of the rapidly cycling, ensemble-based WoFS can be used to generate short-term (0–3 h) probabilistic forecasts of smoke. WoFS-Smoke forecasts from three wildfire cases during 2020 show that the system generates a realistic representation of wildfire smoke when compared against satellite observations. Comparison of smoke forecasts with radar data show that forecast smoke reaches higher levels than radar-detected debris, but exceptions to this are noted. The radiative effect of smoke on surface temperature forecasts is evident, which reduces forecast errors compared to experiments that do not include smoke.
Abstract
Storm surge caused by tropical cyclones can cause overland flooding and lead to loss of life while damaging homes, businesses, and critical infrastructure. In 2018, Hurricane Michael made landfall near Mexico Beach, Florida, on 10 October with peak wind speeds near 71.9 m s−1 (161 mph) and storm surge over 4.5 m NAVD88. During Hurricane Michael, water levels and waves were predicted near–real time using a deterministic, depth-averaged, high-resolution ADCIRC+SWAN model of the northern Gulf of Mexico. The model was forced with an asymmetrical parametric vortex model [generalized asymmetric Holland model (GAHM)] based on Michael’s National Hurricane Center (NHC) forecast track and strength. The authors report errors between simulated and observed water level time series, peak water level, and timing of peak for NHC advisories. Forecasts of water levels were within 0.5 m of observations, and the timing of peak water levels was within 1 h as early as 48 h before Michael’s eventual landfall. We also examined the effect of adding far-field meteorology in our TC vortex model for use in real-time forecasts. In general, we found that including far-field meteorology by blending the TC vortex with a basin-scale NWP product improved water level forecasts. However, we note that divergence between the NHC forecast track and the forecast track of the meteorological model supplying the far-field winds represents a potential limitation to operationalizing a blended wind field surge product. The approaches and data reported herein provide a transparent assessment of water level forecasts during Hurricane Michael and highlight potential future improvements for more accurate predictions.
Abstract
Storm surge caused by tropical cyclones can cause overland flooding and lead to loss of life while damaging homes, businesses, and critical infrastructure. In 2018, Hurricane Michael made landfall near Mexico Beach, Florida, on 10 October with peak wind speeds near 71.9 m s−1 (161 mph) and storm surge over 4.5 m NAVD88. During Hurricane Michael, water levels and waves were predicted near–real time using a deterministic, depth-averaged, high-resolution ADCIRC+SWAN model of the northern Gulf of Mexico. The model was forced with an asymmetrical parametric vortex model [generalized asymmetric Holland model (GAHM)] based on Michael’s National Hurricane Center (NHC) forecast track and strength. The authors report errors between simulated and observed water level time series, peak water level, and timing of peak for NHC advisories. Forecasts of water levels were within 0.5 m of observations, and the timing of peak water levels was within 1 h as early as 48 h before Michael’s eventual landfall. We also examined the effect of adding far-field meteorology in our TC vortex model for use in real-time forecasts. In general, we found that including far-field meteorology by blending the TC vortex with a basin-scale NWP product improved water level forecasts. However, we note that divergence between the NHC forecast track and the forecast track of the meteorological model supplying the far-field winds represents a potential limitation to operationalizing a blended wind field surge product. The approaches and data reported herein provide a transparent assessment of water level forecasts during Hurricane Michael and highlight potential future improvements for more accurate predictions.
Abstract
Nine sets of 36-h, 10-member, convection-allowing ensemble (CAE) forecasts with 3-km horizontal grid spacing were produced over the conterminous United States for a 4-week period. These CAEs had identical configurations except for their initial conditions (ICs), which were constructed to isolate CAE forecast sensitivity to resolution of IC perturbations and central initial states about which IC perturbations were centered. The IC perturbations and central initial states were provided by limited-area ensemble Kalman filter (EnKF) analyses with both 15- and 3-km horizontal grid spacings, as well as from NCEP’s Global Forecast System (GFS) and Global Ensemble Forecast System. Given fixed-resolution IC perturbations, reducing horizontal grid spacing of central initial states improved ∼1–12-h precipitation forecasts. Conversely, for constant-resolution central initial states, reducing horizontal grid spacing of IC perturbations led to comparatively smaller short-term forecast improvements or none at all. Overall, all CAEs initially centered on 3-km EnKF mean analyses produced objectively better ∼1–12-h precipitation forecasts than CAEs initially centered on GFS or 15-km EnKF mean analyses regardless of IC perturbation resolution, strongly suggesting it is more important for central initial states to possess fine-scale structures than IC perturbations for short-term CAE forecasting applications, although fine-scale perturbations could potentially be critical for data assimilation purposes. These findings have important implications for future operational CAE forecast systems and suggest CAE IC development efforts focus on producing the best possible high-resolution deterministic analyses that can serve as central initial states for CAEs.
Significance Statement
Ensembles of weather model forecasts are composed of different “members” that, when combined, can produce probabilities that specific weather events will occur. Ensemble forecasts begin from specified atmospheric states, called initial conditions. For ensembles where initial conditions differ across members, the initial conditions can be viewed as a set of small perturbations added to a central state provided by a single model field. Our study suggests it is more important to increase horizontal resolution of the central state than resolution of the perturbations when initializing ensemble forecasts with 3-km horizontal grid spacing. These findings suggest a potential for computational savings and a streamlined process for improving high-resolution ensemble initial conditions.
Abstract
Nine sets of 36-h, 10-member, convection-allowing ensemble (CAE) forecasts with 3-km horizontal grid spacing were produced over the conterminous United States for a 4-week period. These CAEs had identical configurations except for their initial conditions (ICs), which were constructed to isolate CAE forecast sensitivity to resolution of IC perturbations and central initial states about which IC perturbations were centered. The IC perturbations and central initial states were provided by limited-area ensemble Kalman filter (EnKF) analyses with both 15- and 3-km horizontal grid spacings, as well as from NCEP’s Global Forecast System (GFS) and Global Ensemble Forecast System. Given fixed-resolution IC perturbations, reducing horizontal grid spacing of central initial states improved ∼1–12-h precipitation forecasts. Conversely, for constant-resolution central initial states, reducing horizontal grid spacing of IC perturbations led to comparatively smaller short-term forecast improvements or none at all. Overall, all CAEs initially centered on 3-km EnKF mean analyses produced objectively better ∼1–12-h precipitation forecasts than CAEs initially centered on GFS or 15-km EnKF mean analyses regardless of IC perturbation resolution, strongly suggesting it is more important for central initial states to possess fine-scale structures than IC perturbations for short-term CAE forecasting applications, although fine-scale perturbations could potentially be critical for data assimilation purposes. These findings have important implications for future operational CAE forecast systems and suggest CAE IC development efforts focus on producing the best possible high-resolution deterministic analyses that can serve as central initial states for CAEs.
Significance Statement
Ensembles of weather model forecasts are composed of different “members” that, when combined, can produce probabilities that specific weather events will occur. Ensemble forecasts begin from specified atmospheric states, called initial conditions. For ensembles where initial conditions differ across members, the initial conditions can be viewed as a set of small perturbations added to a central state provided by a single model field. Our study suggests it is more important to increase horizontal resolution of the central state than resolution of the perturbations when initializing ensemble forecasts with 3-km horizontal grid spacing. These findings suggest a potential for computational savings and a streamlined process for improving high-resolution ensemble initial conditions.
Abstract
The High-Resolution Ensemble Forecast system (HREF) calibrated thunder guidance is a suite of probabilistic forecast products designed to predict the likelihood of at least one cloud-to-ground (CG) lightning flash within 20 km (12 miles) of a point during a given 1-, 4-, and 24-h time interval. This guidance takes advantage of a combination of storm attribute and environmental fields produced by the convection-allowing HREF to objectively improve upon lightning forecasts generated by the non-convection-allowing Short-Range Ensemble Forecast system (SREF). This study provides an overview of how the HREF calibrated thunder guidance was developed and calibrated to be statistically reliable against observed CG lightning flashes recorded by the National Lightning Detection Network (NLDN). Performance metrics for the 1-, 4-, and 24-h guidance are provided and compared to the respective SREF calibrated probabilistic lightning forecasts. The HREF calibrated thunder guidance has been implemented operationally within the National Weather Service and is now available to the public.
Significance Statement
The NOAA Storm Prediction Center has created a suite of new calibrated probabilistic thunderstorm guidance products from a convection-allowing model ensemble, the HREF. The new guidance is a notable improvement over the long-running SREF calibrated thunder guidance and is now operational across the National Weather Service.
Abstract
The High-Resolution Ensemble Forecast system (HREF) calibrated thunder guidance is a suite of probabilistic forecast products designed to predict the likelihood of at least one cloud-to-ground (CG) lightning flash within 20 km (12 miles) of a point during a given 1-, 4-, and 24-h time interval. This guidance takes advantage of a combination of storm attribute and environmental fields produced by the convection-allowing HREF to objectively improve upon lightning forecasts generated by the non-convection-allowing Short-Range Ensemble Forecast system (SREF). This study provides an overview of how the HREF calibrated thunder guidance was developed and calibrated to be statistically reliable against observed CG lightning flashes recorded by the National Lightning Detection Network (NLDN). Performance metrics for the 1-, 4-, and 24-h guidance are provided and compared to the respective SREF calibrated probabilistic lightning forecasts. The HREF calibrated thunder guidance has been implemented operationally within the National Weather Service and is now available to the public.
Significance Statement
The NOAA Storm Prediction Center has created a suite of new calibrated probabilistic thunderstorm guidance products from a convection-allowing model ensemble, the HREF. The new guidance is a notable improvement over the long-running SREF calibrated thunder guidance and is now operational across the National Weather Service.
Abstract
Atmospheric River Reconnaissance (AR Recon) is a targeted campaign that complements other sources of observational data, forming part of a diverse observing system. AR Recon 2021 operated for ten weeks from January 13 to March 22, with 29.5 Intensive Observation Periods (IOPs), 45 flights and 1142 successful dropsondes deployed in the northeast Pacific. With the availability of two WC-130J aircraft operated by the 53rd Weather Reconnaissance Squadron (53 WRS), Air Force Reserve Command (AFRC) and one National Oceanic and Atmospheric Administration (NOAA) Aircraft Operations Center (AOC) G-IVSP aircraft, six sequences were accomplished, in which the same synoptic system was sampled over several days.
The principal aim was to gather observations to improve forecasts of landfalling atmospheric rivers on the U.S. West Coast. Sampling of other meteorological phenomena forecast to have downstream impacts over the U.S. was also considered. Alongside forecast improvement, observations were also gathered to address important scientific research questions, as part of a Research and Operations Partnership.
Targeted dropsonde observations were focused on essential atmospheric structures, primarily atmospheric rivers. Adjoint and ensemble sensitivities, mainly focusing on predictions of U.S. West Coast precipitation, provided complementary information on locations where additional observations may help to reduce the forecast uncertainty. Additionally, Airborne Radio Occultation (ARO) and tail radar were active during some flights, 30 drifting buoys were distributed, and 111 radiosondes were launched from four locations in California. Dropsonde, radiosonde and buoy data were available for assimilation in real-time into operational forecast models. Future work is planned to examine the impact of AR Recon 2021 data on model forecasts.
Abstract
Atmospheric River Reconnaissance (AR Recon) is a targeted campaign that complements other sources of observational data, forming part of a diverse observing system. AR Recon 2021 operated for ten weeks from January 13 to March 22, with 29.5 Intensive Observation Periods (IOPs), 45 flights and 1142 successful dropsondes deployed in the northeast Pacific. With the availability of two WC-130J aircraft operated by the 53rd Weather Reconnaissance Squadron (53 WRS), Air Force Reserve Command (AFRC) and one National Oceanic and Atmospheric Administration (NOAA) Aircraft Operations Center (AOC) G-IVSP aircraft, six sequences were accomplished, in which the same synoptic system was sampled over several days.
The principal aim was to gather observations to improve forecasts of landfalling atmospheric rivers on the U.S. West Coast. Sampling of other meteorological phenomena forecast to have downstream impacts over the U.S. was also considered. Alongside forecast improvement, observations were also gathered to address important scientific research questions, as part of a Research and Operations Partnership.
Targeted dropsonde observations were focused on essential atmospheric structures, primarily atmospheric rivers. Adjoint and ensemble sensitivities, mainly focusing on predictions of U.S. West Coast precipitation, provided complementary information on locations where additional observations may help to reduce the forecast uncertainty. Additionally, Airborne Radio Occultation (ARO) and tail radar were active during some flights, 30 drifting buoys were distributed, and 111 radiosondes were launched from four locations in California. Dropsonde, radiosonde and buoy data were available for assimilation in real-time into operational forecast models. Future work is planned to examine the impact of AR Recon 2021 data on model forecasts.
Abstract
Recent research has shown that random forests (RFs) can create skillful probabilistic severe weather hazard forecasts from numerical weather prediction (NWP) ensemble data. However, it remains unclear how RFs use NWP data and how predictors should be generated from NWP ensembles. This paper compares two methods for creating RFs for next-day severe weather prediction using simulated forecast data from the convection-allowing High-Resolution Ensemble Forecast System, version 2.1 (HREFv2.1). The first method uses predictors from individual ensemble members (IM) at the point of prediction, while the second uses ensemble mean (EM) predictors at multiple spatial points. IM and EM RFs are trained with all predictors as well as predictor subsets, and the Python module tree interpreter (TI) is used to assess RF variable importance and the relationships learned by the RFs. Results show that EM RFs have better objective skill compared to similarly configured IM RFs for all hazards, presumably because EM predictors contain less noise. In both IM and EM RFs, storm variables are found to be most important, followed by index and environment variables. Interestingly, RFs created from storm and index variables tend to produce forecasts with greater or equal skill than those from the all-predictor RFs. TI analysis shows that the RFs emphasize different predictors for different hazards in a way that makes physical sense. Further, TI shows that RFs create calibrated hazard probabilities based on complex, multivariate relationships that go well beyond thresholding 2–5-km updraft helicity.
Abstract
Recent research has shown that random forests (RFs) can create skillful probabilistic severe weather hazard forecasts from numerical weather prediction (NWP) ensemble data. However, it remains unclear how RFs use NWP data and how predictors should be generated from NWP ensembles. This paper compares two methods for creating RFs for next-day severe weather prediction using simulated forecast data from the convection-allowing High-Resolution Ensemble Forecast System, version 2.1 (HREFv2.1). The first method uses predictors from individual ensemble members (IM) at the point of prediction, while the second uses ensemble mean (EM) predictors at multiple spatial points. IM and EM RFs are trained with all predictors as well as predictor subsets, and the Python module tree interpreter (TI) is used to assess RF variable importance and the relationships learned by the RFs. Results show that EM RFs have better objective skill compared to similarly configured IM RFs for all hazards, presumably because EM predictors contain less noise. In both IM and EM RFs, storm variables are found to be most important, followed by index and environment variables. Interestingly, RFs created from storm and index variables tend to produce forecasts with greater or equal skill than those from the all-predictor RFs. TI analysis shows that the RFs emphasize different predictors for different hazards in a way that makes physical sense. Further, TI shows that RFs create calibrated hazard probabilities based on complex, multivariate relationships that go well beyond thresholding 2–5-km updraft helicity.
Abstract
The Global Ensemble Forecast System (GEFS) is upgraded to version 12, in which the legacy Global Spectral Model (GSM) is replaced by a model with a new dynamical core—the Finite Volume Cubed-Sphere Dynamical Core (FV3). Extensive tests were performed to determine the optimal model and ensemble configuration. The new GEFS has cubed-sphere grids with a horizontal resolution of about 25 km and an increased ensemble size from 20 to 30. It extends the forecast length from 16 to 35 days to support subseasonal forecasts. The stochastic total tendency perturbation (STTP) scheme is replaced by two model uncertainty schemes: the stochastically perturbed physics tendencies (SPPT) scheme and stochastic kinetic energy backscatter (SKEB) scheme. Forecast verification is performed on a period of more than two years of retrospective runs. The results show that the upgraded GEFS outperforms the operational-at-the-time version by all measures included in the GEFS verification package. The new system has a better ensemble error–spread relationship, significantly improved skills in large-scale environment forecasts, precipitation probability forecasts over CONUS, tropical cyclone track and intensity forecasts, and significantly reduced 2-m temperature biases over North America. GEFSv12 was implemented on 23 September 2020.
Abstract
The Global Ensemble Forecast System (GEFS) is upgraded to version 12, in which the legacy Global Spectral Model (GSM) is replaced by a model with a new dynamical core—the Finite Volume Cubed-Sphere Dynamical Core (FV3). Extensive tests were performed to determine the optimal model and ensemble configuration. The new GEFS has cubed-sphere grids with a horizontal resolution of about 25 km and an increased ensemble size from 20 to 30. It extends the forecast length from 16 to 35 days to support subseasonal forecasts. The stochastic total tendency perturbation (STTP) scheme is replaced by two model uncertainty schemes: the stochastically perturbed physics tendencies (SPPT) scheme and stochastic kinetic energy backscatter (SKEB) scheme. Forecast verification is performed on a period of more than two years of retrospective runs. The results show that the upgraded GEFS outperforms the operational-at-the-time version by all measures included in the GEFS verification package. The new system has a better ensemble error–spread relationship, significantly improved skills in large-scale environment forecasts, precipitation probability forecasts over CONUS, tropical cyclone track and intensity forecasts, and significantly reduced 2-m temperature biases over North America. GEFSv12 was implemented on 23 September 2020.
Abstract
Accurately representing boundary layer turbulent processes in numerical models is critical to improve tropical cyclone forecasts. A new turbulence kinetic energy (TKE)-based moist eddy-diffusivity mass-flux (EDMF-TKE) planetary boundary layer scheme has been implemented in NOAA’s Hurricane Analysis and Forecast System (HAFS). This study evaluates EDMF-TKE in hurricane conditions based on a recently developed framework using large-eddy simulation (LES). Single-column modeling tests indicate that EDMF-TKE produces much greater TKE values below 500-m height than LES benchmark runs in different high-wind conditions. To improve these results, two parameters in the TKE scheme were modified to ensure a match between the PBL and surface-layer parameterizations. Additional improvements were made by reducing the maximum allowable mixing length to 40 m based on LES and observations, by adopting a different definition of boundary layer height, and by reducing nonlocal mass fluxes in high-wind conditions. With these modifications, the profiles of TKE, eddy viscosity, and winds compare much better with LES results. Three-dimensional idealized simulations and an ensemble of HAFS forecasts of Hurricane Michael (2018) consistently show that the modified EDMF-TKE tends to produce a stronger vortex with a smaller radius of maximum wind than the original EDMF-TKE, while the radius of gale-force wind is unaffected. The modified EDMF-TKE code produces smaller eddy viscosity within the boundary layer compared to the original code, which contributes to stronger inflow, especially within the annulus of 1–3 times the radius of maximum wind. The modified EDMF-TKE shows promise to improve forecast skill of rapid intensification in sheared environments.
Abstract
Accurately representing boundary layer turbulent processes in numerical models is critical to improve tropical cyclone forecasts. A new turbulence kinetic energy (TKE)-based moist eddy-diffusivity mass-flux (EDMF-TKE) planetary boundary layer scheme has been implemented in NOAA’s Hurricane Analysis and Forecast System (HAFS). This study evaluates EDMF-TKE in hurricane conditions based on a recently developed framework using large-eddy simulation (LES). Single-column modeling tests indicate that EDMF-TKE produces much greater TKE values below 500-m height than LES benchmark runs in different high-wind conditions. To improve these results, two parameters in the TKE scheme were modified to ensure a match between the PBL and surface-layer parameterizations. Additional improvements were made by reducing the maximum allowable mixing length to 40 m based on LES and observations, by adopting a different definition of boundary layer height, and by reducing nonlocal mass fluxes in high-wind conditions. With these modifications, the profiles of TKE, eddy viscosity, and winds compare much better with LES results. Three-dimensional idealized simulations and an ensemble of HAFS forecasts of Hurricane Michael (2018) consistently show that the modified EDMF-TKE tends to produce a stronger vortex with a smaller radius of maximum wind than the original EDMF-TKE, while the radius of gale-force wind is unaffected. The modified EDMF-TKE code produces smaller eddy viscosity within the boundary layer compared to the original code, which contributes to stronger inflow, especially within the annulus of 1–3 times the radius of maximum wind. The modified EDMF-TKE shows promise to improve forecast skill of rapid intensification in sheared environments.
Abstract
We utilized high temporal resolution, near-surface observations of sustained winds and gusts from two networks, the primarily airport-based Automated Surface Observing System (ASOS) and the New York State Mesonet (NYSM), to evaluate forecasts from the operational High-Resolution Rapid Refresh (HRRR) model, versions 3 and 4. Consistent with past studies, we showed the model has a high degree of skill in reproducing the diurnal variation of network-averaged wind speed of ASOS stations, but also revealed several areas where improvements could be made. Forecasts were found to be underdispersive, deficient in both temporal and spatial variability, with significant errors occurring during local nighttime hours in all regions and in forested environments for all hours of the day. This explained why the model overpredicted the network-averaged wind in the NYSM because much of that network’s stations are in forested areas. A simple gust parameterization was shown not only to have skill in predicting gusts in both networks but also to mitigate systemic biases found in the sustained wind forecasts.
Significance Statement
Many users depend on forecasts from operational models and need to know their strengths, weaknesses, and limitations. We examined generally high-quality near-surface observations of sustained winds and gusts from the nationwide Automated Surface Observing System (ASOS) and the New York State Mesonet (NYSM) and used them to evaluate forecasts from the previous (version 3) and current (version 4) operational High-Resolution Rapid Refresh (HRRR) model for a selected month. Evidence indicated that the wind forecasts are excellent yet imperfect and areas for further improvement remain. In particular, we showed there is a high degree of skill in representing the diurnal variation of sustained wind at ASOS stations but insufficient spatial and temporal forecast variability and overprediction at night everywhere, in forested areas at all times of day, and at NYSM sites in particular, which are more likely to be sited in the forest. Gusts are subgrid even at the fine grid spacing of the HRRR (3 km) and thus must be parameterized. Our simple gust algorithm corrected for some of these systemic biases, resulting in very good predictions of the maximum hourly gust.
Abstract
We utilized high temporal resolution, near-surface observations of sustained winds and gusts from two networks, the primarily airport-based Automated Surface Observing System (ASOS) and the New York State Mesonet (NYSM), to evaluate forecasts from the operational High-Resolution Rapid Refresh (HRRR) model, versions 3 and 4. Consistent with past studies, we showed the model has a high degree of skill in reproducing the diurnal variation of network-averaged wind speed of ASOS stations, but also revealed several areas where improvements could be made. Forecasts were found to be underdispersive, deficient in both temporal and spatial variability, with significant errors occurring during local nighttime hours in all regions and in forested environments for all hours of the day. This explained why the model overpredicted the network-averaged wind in the NYSM because much of that network’s stations are in forested areas. A simple gust parameterization was shown not only to have skill in predicting gusts in both networks but also to mitigate systemic biases found in the sustained wind forecasts.
Significance Statement
Many users depend on forecasts from operational models and need to know their strengths, weaknesses, and limitations. We examined generally high-quality near-surface observations of sustained winds and gusts from the nationwide Automated Surface Observing System (ASOS) and the New York State Mesonet (NYSM) and used them to evaluate forecasts from the previous (version 3) and current (version 4) operational High-Resolution Rapid Refresh (HRRR) model for a selected month. Evidence indicated that the wind forecasts are excellent yet imperfect and areas for further improvement remain. In particular, we showed there is a high degree of skill in representing the diurnal variation of sustained wind at ASOS stations but insufficient spatial and temporal forecast variability and overprediction at night everywhere, in forested areas at all times of day, and at NYSM sites in particular, which are more likely to be sited in the forest. Gusts are subgrid even at the fine grid spacing of the HRRR (3 km) and thus must be parameterized. Our simple gust algorithm corrected for some of these systemic biases, resulting in very good predictions of the maximum hourly gust.
