Collaborative Effort to Account for the Impact of the August 21st Solar Eclipse on Operational Forecasts in the Mid Atlantic and Southeast

Map of the percentage of sun obscuration during the eclipse shown in AWIPS.

Map of the percentage of sun obscuration during the eclipse shown in AWIPS.

Meteorologists recognize that solar eclipses in the past have had a notable impact on the sensible weather in the regions in which they occur. These impacts can include a decrease in surface temperature, reduction and changes to surface winds, lowering of surface pressure, changes in cloud cover and more. National Weather Service (NWS) meteorologist in the Southeast and mid-Atlantic have been working collaboratively to account for some of these impacts on official NWS forecasts during the eclipse on Monday, August 21, 2017.

eclipse temp training

Image highlighting the some details of the Eclipse Temperature smartTool and Procedure for GFE.

Surprisingly, most operational numerical weather prediction (NWP) systems do not account for the changes in incoming solar radiation from the sun during an eclipse and the resultant changes in the weather. This is an issue as NWS forecasters provide forecasts of temperature, winds, and other fields at hourly time steps and the eclipse impacts need to be captured by forecaster intervention over model guidance. The effort to provide details on the eclipse impact on weather in our region began with the development of hourly temperature reductions based on past eclipse events and factoring local climatology proposed by Frank Alsheimer at WFO CAE. Additional WFOs in the CIMMSE area collaborated and provided input on reductions while working within the temporal confines of the National Digital Forecast Database. Joshua Weiss at WFO ILM, examined data from the 1970 and 1984 eclipses in the Southeast and created a GFE smartTool and Procedure to provide a more scientifically sound and consistent process to edit the hourly temperature forecasts.


Official hourly forecasts of temperature and sky cover from three locations in adjacent WFOs CAE, ILM, and RAH. The temperatures were adjusted using the Eclipse Temperature GFE procedure. Columbia, SC is in the eclipse totality while Lumberton, NC and Wadesboro, NC reach 97% obscuration. Differing temperature reductions during and after the eclipse are influenced by various factors including differing amounts of cloud cover.

The GFE Procedure incorporates the forecast temperature and diurnal range without the eclipse impact, whether a location is in the total or partial eclipse, and the amount of cloud cover. Nearly a half dozen WFOs in the Southeast will be using this tool which should lead to a more consistent, more scientifically sound, and accurate forecast. In addition, shapefiles of eclipse information were developed for use in some applications, GIS maps detailing eclipse obscuration percentages were installed in AWIPS, and finally methodologies and strategies were shared via a Google document. This effort builds on the history and relationships built across the CIMMSE domain. The event is a great example of many NWS meteorologists and WFOs working together to provide enhanced forecasts and service.


Screen shot of the High-Resolution Rapid Refresh (HRRRx) experimental real-time weather forecast web page the eclipse.

It is also worth noting that some experimental and non-operational NWP systems will account for the eclipse. NOAA/ESRL/GSD made changes to the 3km experimental HRRR (HRRRx) to include code to account for the sun-obscuration from the eclipse (details… For real-time HRRRx experimental forecasts, including the effects of the eclipse starting Saturday night looking ahead 48 hours with the 00 UTC model run, visit Some of the selected weather fields available include downward solar radiation, cloud fields and 2-meter temperature, for the HRRRx (with eclipse effect), the operational HRRR-NCEP (without eclipse effects and some other differences) and HRRRx – HRRR-NCEP difference fields.

Posted in CIMMSE, NWP | Leave a comment

A pair of HSLC-related presentations: NOAA VLab Forum and AMS Mesoscale Conference

Over the last few months, I have provided two presentations on recent HSLC-related research. In June, I was the presenter at NOAA’s VLab Forum, where I provided an overview of the ongoing HSLC CSTAR project based at NC State University. A recording of this presentation can be found here.

The next month, I traveled to the AMS Conference on Mesoscale Processes in San Diego, where I presented an update on my ongoing idealized modeling work. Of particular interest, I have identified the chain of processes appearing to result in the development of strong, low-level vortices within simulated HSLC QLCSs and determined how low-level shear vector magnitude and low-level lapse rates could affect these processes.

Simulation-based research continues, and I intend to complete and defend my dissertation later this fall. Please let me know if you have any questions or comments!

Posted in CIMMSE, Convection, CSTAR, High Shear Low Cape Severe Wx | Leave a comment

Operational NWP Resolution and Sensitivities Study Using HSLC Event Hindcasts Summary

At last month’s CSTAR workshop, I shared the results and analysis of my thesis project as well as some thoughts on related future work. I would like to provide a summary of my and Dr. Lackmann’s work with this post for those who could not be there and as a refresher for those who were.

The problem I focused on is that severe convection that forms in  High Shear, Low CAPE (HSLC) environments is difficult to predict and is dangerous. There are two approaches to improve HSLC severe convection predictability: the first is the use of environmental predictors such as the SHERB (Sherburn and Parker 2014) and the MOSH (Sherburn et al. 2016) and the second is using Numerical Weather Prediction (NWP) models to predict convection which is what myself and Jessica King did (King et al. 2017). This research project sought to answer two research questions:

1. At what horizontal grid spacing, if any, does an NWP model provide operationally useful information about explicit low instability severe convection?
2. Can a NWP model properly differentiate between a HSLC event case and a HSLC null event case given the proper initial conditions?

Dr. Lackmann and I  looked at two cases, a event case February 24-25, 2011, a null case, December 25-26, 2009. I ran these cases using the WRF model at convection-permitting grid spacings (Figure 1) and then evaluated how well each domain represented the severe weather using severe proxy metrics: half-hourly maximum updraft helicity (1-4km), half-hourly maximum 10-meter wind speed, half-hourly maximum updraft speed, and composite radar reflectivity (Figure 2). In order to compare the different domains on an “equal playing field” so to speak, we conformed the 1.2-km and 400-m domains to a grid with a 3.6-km grid spacing over the area encompassed by the 400-m domain. The severe proxy metrics were calculated on the higher resolution domain first before this interpolation. This process is effectively sub-sampling. The distributions of the severe proxy values of these interpolated domains were compared to one another both in the same case and between the two cases (Figure 3).

The results of this analysis are as follows:
– Based on this study, the recommended horizontal grid spacing to run a NWP model in a way that provides operationally useful information about low instability convection is 3.6-km.
– A larger gain in detail is observed between the 3.6-km and 1.2-km domains than between the 1.2-km and 400-m domains.
– Overall, the event case and null event case are statistically significantly different. It is important to note that there is a difference between differentiating and distinguishing; the model had difficulty distinguishing between the two cases.
– Thresholds of severe proxy parameters, especially updraft helicity, should be adjusted to handle finer model resolution.

I look forward to receiving your questions and comments as well as participating in any discussion that follows. Please feel free to ask me questions in the comments section or to send an email to Thank you!

Figure 1: Model domain set up for the event case February 24-25, 2011

Figure 2: NEXRAD and conformed model domain composite radar reflectivity for the event case. Time shown is 03:00 UTC on 02/25/2011.

Figure 3: The severe proxy conformed half-hourly maximum updraft helicity distributions for the event case and null event case at each convective permitting domain.


King, J.R., M.D. Parker, K.D. Sherburn, and G.M. Lackmann 2017: Rapid Evolution of Cool Season, Low CAPE Severe Thunderstorm Environments. Wea. Forecasting. doi: 10.1175/WAF-D-16-0141.1

Sherburn, K.D., and M.D. Parker, 2014: Climatology and Ingredients of Significant Severe Convection in High-Shear, Low-CAPE Environments. Wea. Forecasting, 29, 854-877, doi: 10.1175/WAF-D-13-00041.1

———-, ———–, G.M. Lackmann, and J.R. King, 2016: Composite environments of severe and non-severe high-shear, low-CAPE convective events. Wea. Forecasting. Doi: 1175/WAF-D-16-0086.1

Posted in CIMMSE, Convection, CSTAR, High Shear Low Cape Severe Wx, NWP, Uncategorized | Tagged , | Leave a comment

Updated List of NC State-NWS CSTAR Publications


Figure depicting the change in surface based CAPE in the hours prior to severe convection in HSLC environments as described in King et al., 2017.

Shown below is a list of publications and abstracts developed as a part of the most recent NC State-NWS CSTAR project entitled “Understanding and Prediction of High Impact Weather Associated with Low-Topped Severe Convection in the Southeastern U.S.” Note that additional presentations are scheduled for upcoming conferences including the AMS 24th Conference on Probability and Statistics in July and the 17th Conference on Mesoscale Processes in July. In addition, manuscripts are being composed for multiple projects including the predictability study using ensembles and dynamical-statistical downscaling project.

Publications (most recent first):

1) King, J. R., M. D. Parker, K. D. Sherburn, and G. M. Lackmann, 2017: Rapid evolution of cool season, low CAPE severe thunderstorm environments. Wea. Forecasting, 32, 763-779. | PDF |

2) Sherburn, K. D., M. D. Parker, J. R. King, and G. M. Lackmann, 2016: Composite environments of severe and non-severe high-shear, low-CAPE convective events. Wea. Forecasting, 31, 1899-1927. | PDF |

Conference abstracts (most recent first):

1) Sherburn, K. D., and M. D. Parker, 2016: The origins of rotation within high-shear, low-CAPE mesovortices and mesocyclones. 28th Conference on Severe Local Storms, AMS, 7-11 November 2016, Portland, OR. | Recorded presentation | Manuscript |

2) Sherburn, K. D., and M. D. Parker, 2016: Insights from composite environments of high-shear low-CAPE severe convection. 28th Conf. on Severe Local Storms, AMS, 7-11 November 2016, Portland, OR. | Poster handout | Manuscript |

3) Blank, L., and G. Lackmann, 2016: Operational predictability of explicit high shear, low CAPE convection. 6th Conference on Transition of Research to Operations, AMS, 11-14 January 2016, New Orleans, LA. | Recorded presentation |

4) King, J. R., and M. D. Parker, 2015: Conditioning and evolution of high shear, low CAPE severe environments. 16th Conference on Mesoscale Processes, AMS, 2-6 August 2015, Boston, MA. | Recorded presentation | Manuscript |

5) Sherburn, K. D., and M. D. Parker, 2015: Examining the sensitivities of high-shear low-CAPE convection to low-level hodograph shape. 16th Conference on Mesoscale Processes, AMS, 2-6 August 2015, Boston, MA. | Recorded presentation | Manuscript |

6) King, J. R., and M. D. Parker, 2014: Synoptic influence on high shear, low CAPE convective events. 27th Conference on Severe Local Storms, AMS, 2-7 November 2014, Madison, WI. | Poster handout | Manuscript |

7) Sherburn, K. D., and M. D. Parker, 2014: High-shear, low-CAPE environments: What we know and where to go next. 27th Conference on Severe Local Storms, AMS, 2-7 November 2014, Madison, WI. | Recorded presentation | Manuscript |

8) Sherburn, K. D., and M. D. Parker, 2014: On the usage of composite parameters in high-shear, low-CAPE environments. 27th Conference on Severe Local Storms, AMS, 2-7 November 2014, Madison, WI. | Poster handout | Manuscript |

Posted in Uncategorized | Leave a comment

The Utility of Non-NWS Upper Air Observations – NC State Soundings Support Severe Weather Operations at NWS Raleigh on May 4th and 5th, 2017

NC State sounding from 0103 UTC 05/05/2017 during a severe weather event.

There are several non-NWS organizations that take upper-air observations and share them with NWS forecasters. These organizations largely consist of universities, military installations or research laboratories. The motivation for taking these observations include research efforts as well as educational and training activities.

One example of such an organization is the “Sounding Club” at NC State University which is lead by students and overseen by professors within the Atmospheric Science department.  Some of the goals of the organization according to their website are to help students obtain hands-on experience in addition to collecting data for projects. In addition to professional development and project support, the soundings have been used by NWS forecasters to support operational needs.

A pair of soundings from the NC State Sounding Club were very helpful during the evening and overnight hours on May 4th and 5th when there was concern for severe convection (SPC reports).  NWS Forecaster Michael Strickler noted the utility of the soundings at approximately 01Z and 03Z on 05 May…

The soundings revealed that the low 70s/low 60s surface temperatures and dew points were already supportive of surface-based convection even prior to the arrival of the mT air mass from the southeast, but with only weak instability and low equilibrium levels. This information provided radar operators confidence of severe wind and tornado potential. The sounding also revealed the character and strength of the low level shear, with a greater degree of confidence relative to that of WSR-derived VWP winds/hodographs that tend to fluctuate/become noisy and are consequently often times less reliable. Pre-QLCS storm motion vectors, as indicated by the soundings, also proved beneficial from a pre-warning situational awareness standpoint.

NWS Chat message from 0525 UTC (0125 AM EDT) 05/05/2017 describing the near term environment.

UNC Asheville sounding from 0300 UTC 01/07/2017 during a winter storm that affected much of the Virginias and Carolinas.

NWS Raleigh forecasters obtain the soundings from the NC State Sounding Club in near real-time either through email, posts to the CSTAR mailing list, or via social media.  In addition to NC State, there are several other organizations that take upper air observations.  UNC Asheville has taken upper air observations during many winter storms in recent years as a part of their Sounding-based Experiment on Mixed Precipitation Events (SEMPE) project. These soundings have been used by the NWS offices in Greenville-Spartanburg, Blacksburg and Raleigh among others.

In the Southeast, we are aware of several organizations taking supplemental observations; some of them are shown in the list below:

  • NC State
  • UNC Asheville
  • UNC Charlotte
  • University of South Carolina
  • University of Alabama Huntsville
  • Mississippi State
  • University of Louisiana-Monroe
  • Simmons AAF Fort Bragg
  • Redstone Arsenal
  • NOAA Lab in Oak Ridge, TN

These observations have a great deal of potential to support NWS warning and forecast operations. Unfortunately, they are often underutilized for a variety of reasons including a lack of awareness and coordination of the sounding operations, limitations in the dissemination and display of the soundings, and an inability to ingest the data into AWIPS.

There is a desire to collect and organize these grass roots observations into a process to improve the awareness, communication, access and display of these observations among NWS forecasters. Collaborators with the NC State CSTAR group will be reaching out to those taking the observations to gauge interest in improve the flow of the data to the NWS and perhaps elsewhere.

Posted in CSTAR, Winter Weather | Leave a comment

2017 CSTAR Workshop and Mid-Atlantic and Southeast Sub-regional SOO Meeting Held in Raleigh


Photo of the 2017 CSTAR workshop attendees.

On April 26th through 28th over 20 meteorologists from the National Weather Service (NWS) as well as faculty and students from N.C. State University (NCSU) gathered in Raleigh to get updates on various collaborative research activities, share operational and training successes and to look to the future toward planned and potential projects across our region.

Meteorologists from 9 NWS WFOs along with NWS Eastern Region Headquarters (ERH) met during a portion of Wednesday and Friday for a NWS Sub-regional SOO Meeting. This time was dedicated to meet some of the new SOOs/staff from around the area, discuss best practices and challenges unique to the region, and to look to the future of the science program. Thursday was largely dedicated to collaborative research with NC State including updates on a 3 year Collaborative Science, Technology, and Applied Research (CSTAR) project focused on high impact weather associated with low-topped severe convection in the Southeastern U.S.” In addition, other research activities at NC State including heavy banded snowfall, predictive modeling for storm surge and flooding, and quantifying radar uncertainty and ensemble QPE were presented.

Finally, a new CSTAR project entitled “Understanding fundamental processes and evaluating high-resolution model forecasts in high-shear low-CAPE severe storm environments” was introduced. This project will build off of previous collaborative research between NCSU and the NWS, which has had very successful research to operation results.

Presentations from the workshop are available via Google Drive at this URL:

Look for additional details on the workshop including follow up activities, highlights and updates in future blog posts. A virtual workshop is tentatively planned for October with a desire for another in person meeting next fall as the next CSTAR effort matures.

Posted in CIMMSE, CSTAR | Leave a comment

Hurricane Matthew Gust Factors


Tropical cyclone best track and wind radii data for Hurricane Matthew provided by the National Hurricane Center from 07 October to 09 October 2016.

To evaluate the performance of CSTAR related research to operations activities, the sustained winds, wind gusts, and gust factors for Hurricane Matthew (2016) were examined across coastal and eastern Georgia, South Carolina, North Carolina, and Virginia. The image above or to the right is the tropical cyclone best track and wind radii data for Hurricane Matthew which shows the track of the storm near and along much of the southeast U.S. coast. The track data is a subjectively smoothed representation of a tropical cyclone’s history over its lifetime, based on a post-storm analysis of all available data. The data also contains wind radii information which is the farthest distance from the cyclone’s center where sustained winds of 34-, 50-, and 64-kts are occurring in each of four quadrants about the storm (NE, SE, SW, and NW).

The map below is a subjective analysis of the maximum wind gusts observed across NC during Hurricane Matthew. Many locations across the immediate coast observed wind gusts in excess of 70 MPH with a few locations across the Outer Banks reporting wind gust greater than 80 MPH. Wind gusts of 50 MPH or more extended inland to the southern and central I-95 corridor.


Subjective analysis of the maximum wind gusts (MPH) observed across North Carolina from 07 October to 09 October 2016 during Hurricane Matthew.

NWS Raleigh volunteer Victoria Oliva, examined the sustained winds, wind gusts, and gust factors for Matthew across coastal and eastern Georgia, South Carolina, North Carolina and Virginia. Hourly observations of winds and wind gusts from 60 regular ASOS or AWOS METAR locations impacted by the over land wind field associated with Tropical Storm Matthew were examined. The locations examined in this analysis extended from KSVN (Hunter Army Airfield near Savannah, Georgia) northeast along and just inland of the coast to KWAL (Wallops Island, Virginia).

Only observations from routine hourly METARs were used (special observations and observations not at the top of the hour were excluded). In addition, gust factors were only calculated for sustained winds of 10 kts or greater. For each observation, the hourly wind gust factor was computed. The gust factor is defined as the ratio between the wind gust of a specific duration to the mean (sustained) wind speed for a period of time. A total of 979 gust factors were computed for Matthew.

The maximum sustained wind contained in the Matthew data set was only 53 kts and the maximum wind gust was 70 kts with both observations from KHXD (Hilton Head, SC). The data set contained a large number of lower end sustained winds and wind gusts. Nearly 79% or 769 out of the 979 observations, contained in this data-set had sustained winds less than 25 kts. Only 85 observations, or around 9% of all observations, had sustained winds of 40 kts or more and only 8 observations had sustained winds of 40 kts or more.


Scatter plot of sustained winds and gust factors for Hurricane Matthew for 60 METAR locations across coastal and eastern Georgia, South Carolina, North Carolina and Virginia.

The chart to the right is a scatter plot of the sustained winds in kts versus gust factors for the 979 observations included in the study along with a best fit regression curve (y = -0.231ln(x) + 2.2498). In general, the chart demonstrates an inverse relationship between the wind speed and gust factor. Not surprisingly, the gust factors with Matthew were rather variable at low sustained wind speeds and generally converged and decreased with increasing sustained wind speed. This chart is similar to the database of 15 storms used to develop the CSTAR TCM wind technique.

mathew.histogramA histogram of the frequency of gust factors for Matthew is shown to the right. The average gust factor for Matthew was 1.58 which is somewhat higher than the average of 1.53 for the database of 15 tropical cyclones used to develop the CSTAR TC wind technique. The histogram data noted that the gust factors were most frequently noted between 1.5 and 1.6.

Comparison of gust factor regression equations associated with Hurricane Matthew and the 15 storm database used to develop the CSTAR TCM wind technique.

We generated a regression equation using the gust factor data set associated with Matthew and compared it to the regression equation used in the 15 storm database used to develop the CSTAR TCM wind technique. The figure to the right compares the Matthew regression equation (shown in red) to the 15 storm equation (shown in blue). They show a similar trend but the gust factors with Matthew are consistently a little higher than the 15-storm dataset.

Posted in CSTAR, TC Inland and Marine Winds, Uncategorized | Leave a comment