Close Window

Univ. of Massachusetts - Lowell: "The development, propagation, and erosion of severe convection in the Boston-Providence corridor"

Final Report


The objective of this proposal was to determine the influence of the coastal environment on the behavior of summertime convection in Southern New England. A previous study had looked only at rawinsonde data from a coastal location on Cape Cod (Chatham, MA), and determined that if the low level winds were strongly on-shore, then convection would tend to dissipate as it reached Southeastern Massachusetts (SEMass). Our intention was to use current observations and model simulations to develop a deeper understanding of this influence, and to develop a way for forecasters to better anticipate convective behavior in this region.

A total of 28 cases from 2001 and 2002 were analyzed, as well as a few special days from 2000. Thirteen of the cases were examples of convection which remained strong or severe all the way to the SEMass Atlantic coastline, while the other fifteen were examples in which the convection weakened or died as it approached the southeastern coastline of Massachusetts.

The Penn State/NCAR Mesoscale Model version 5 (MM5) was used in a quasi-operational mode as the main tool of analysis. The model was run for 24 hours of simulated time, over three nested grids with grid sizes of 36 km, 12 km, and 4 km respectively. Initialization and boundary conditions were provided by National Weather Service AVN or Eta model output, depending upon model availability.

The cases from 2001 were separated into two groups: sustained convection and dissipating convection. Each group was made into composites, from which several conclusions were drawn. Patterns of low-level wind, latent heat flux (lhf), and equivalent potential temperature (theta-e) were examined for differences between the two groups. Post processing included retrieval of averaged model soundings from various regions in southern New England, computing convective available potential energy (CAPE) and convective inhibition (CIN), and plotting these quantities as a function of time.

The plots of CAPE and CIN implied that the vertical stability over SEMass changed to late to sustain vigorous convection in the dissipating convection cases, but the differences between the two groups was not distinctive enough to warrant this extensive post processing for forecast purposes. However, the pattern differences were significant, and required little post processing to produce. After testing them with the 2002 cases, the following results emerged:

* low level winds from the south over SEMass led to dissipating convection over SEMass, whereas low level westerly or west southwesterly winds were favorable for sustained convection.
* theta-e tended to have maximum values over all of Massachusetts, or over SEMass for sustained convection. Maximum values over central or northeastern Massachusetts led to dissipating convection over SEMass.
* lhf patterns mirrored those of theta-e. Maximum values over central or northeastern Massachusetts led to dissipating convection over SEMass.
* cross sections of theta-e over SEMass must show low-level instability (theta-e decreasing upwards) for convection to be sustained.
* the model computed composite radar reflectivity showed high values over SEMass in most cases of sustained convection. While this last point seems obvious, we did not assume that the model would correctly simulate the convection in the still relatively coarse 4 km grid.
* inhomogeneity in the windfield with respect to direction and/or speed or both over several hours prior to the convective developments, combined with a - relative to the ground - stronger, southwesterly wind flow aloft, are favorable for sustained convection. (This, of course, is also supported by the theory that stronger windshear between the surface and higher levels is needed for convection to get organized)

In general, examination of these cases of convection demonstrated how complicated the convective process is, and how important it is for forecasters to have a good understanding of the process itself, rather than relying upon a checklist.


One member of the research team, Ms. Susanna Hopsch, used this research for her Masters thesis, and the Chief Meteorologist, Mr. Robert Thompson, from the Taunton (Massachusetts) National Weather Service (NWS) Forecast Office was on her thesis committee. There were a number of visits relating to thesis reports and questions which involved informal exchanges between Mr. Thompson, Ms. Hopsch and the rest of the thesis committee, both at the University as well as the NWS office. When Ms. Hopsch defended her thesis, Mr. Thompson attended.

In April of 2003, Dr. Frank Colby, Ms. Hopsch’s thesis advisor and the University P.I for the project, spent a day at the Taunton NWS Forecast Office as part of a seminar for Severe Weather Training. Dr. Colby presented the results of this study, and a copy of the presentation was provided to the Forecast Office.

A visit from Mr. David Vallee is planned to demonstrate the NWS Weather Event Simulator to a class at the University during the 2003-2004 academic year.

A joint presentation to the next Conference on Severe Local Storms is being planned.


Colby, F. P. Jr., 2003: Forecasting the development, propagation, and erosion of severe convection in the Boston-Providence corridor. Presented at the Severe Weather Training Session, April 1, 2003, Taunton, Ma.

Hopsch, S. B., 2002: A study of severe convection over Southern New England. M.S. thesis, Dept. of Env., Earth, & Atmos. Sciences, University of Massachusetts Lowell, 160 pp. [Available from Dept. of Env., Earth & Atmos. Sciences, University of Massachusetts Lowell, 1 University Ave, Lowell, MA 01854.]

Hopsch, S. B., 2001: A Study of Severe Convection Over Southern New England: Preliminary Results. Second Southern New England Conference and Conference Preprint, December 1, 2001, Worcester, MA.

Hopsch, S. B., 2002: Common Characteristics of Severe Convection Over Southern New England. Southern New England Chapter of the NWA quarterly meeting, June 8, 2003, Worcester, MA.


4.1 This project has reinforced a growing relationship with the Taunton NWS Forecast Office. We have been working on informal research projects for the last few years, but the interactions fostered by this project have already led to definite plans for new research, including another M.S. thesis. The exchange of personnel has been very useful too. Visits to the NWS office have kept the University faculty up to date on current NWS forecast procedures, and we are really looking forward to the visit by Mr. Vallee to the University in the coming year.

The only problems we encountered in this project revolved around data. Ms. Hopsch had hoped to apply Fourier transforms to look for patterns in hourly metar observations. Unfortunately, there were enough missing observations to make this effort impossible. In addition, the NWS Forecast Office had some trouble retrieving WSR-88D data from their unreliable archive, since updated. This made verification of some of the model runs more difficult.

4.2 WFO Taunton, in recent years, has struggled with the dilemma of whether or not severe convection would be a significant threat to the Boston-Providence metropolitan corridor. Dr. Colby’s research has not only identified a variety or critical elements, but has resulted in a very concise forecast approach to assist in our decision making process.

Each fall the Science and Operations Officer, at WFO Taunton, composes a training program for the forthcoming fiscal year. For the 2002-2003 season, this plan centered its severe weather training sessions around Dr. Colby’s research. Dr. Colby and David Vallee (SOO WFO BOX) worked together on developing the day-long training content, and took the staff through two Weather Event Simulations (WES) related to severe convection and the complexities of forecasting the evolution in the Boston-Providence metropolitan corridor. These simulation cases included two particular severe weather outbreaks; May 31, 2002 which did not make it into the coastal corridor and July 27, 2002, which did make it to the coastal corridor.

The training schedule for the day included the following topics:

1. WFO Taunton Severe Weather Verification and Operations
2. Buoyancy and Shear Considerations with Regard to Convective Evolution (D. Vallee)
3. A Study of Severe Weather in Southern New England (F. Colby)
4. Two WES cases following the lunch break (Dr. Colby and D. Vallee)

From the WFO perspective, the process of taking applied research and moving it swiftly into forecast operations was effectively demonstrated from our collaborative efforts. Likewise, Dr. Colby’s participation was very well received by WFO Taunton forecasters. Many have commented on how enlightening his information was and that his interactive participation was a valued enhancement to the training effort. Dr. Colby and David Vallee see a tremendous opportunity for future work and consider this effort a success at crossing that bridge from research to operations. Likewise, we see a tremendous opportunity for the sponsoring University by having NWS forecasters become more involved with the educational efforts through on-site visits, student volunteer projects, and near real-time sharing of datasets once the modified version of the WES software is available to the academic community. This type of collaboration helps the participating university design curriculum and projects that are relevant to present day operational issues.

The most significant draw back was no fault of our own. There were several large gaps in available Archive IV data, which limited the amount of model verification that would have been desired. The NCDC on-line archive of level II and level III products will hopefully reduce the likelihood of this impacting future radar related research.