Surface Boundary & Air Mass Considerations for Storms Near Philadelphia, PA: 8/28/20 - Operations Proving Ground

On the evening of August 27th, decaying convection, which had produced severe wind across New York, Pennsylvania, and northern New Jersey earlier in the day (see previous blog post), had produced an outflow boundary which was advancing southward through central New Jersey (Figure 1).

Figure 1: 0.5° Base Reflectivity from KDIX valid from 0100z to 02:44Z 8/28 with annotation of the outflow boundary (red) at 02:44 8/28. Radar imagery viewed with GR software.

By late on the evening of 8/27 it was clear that the convectively driven outflow had stalled across central New Jersey. The question then became, how would this affect convection for the following day? Would the deposited cold pool have time to modify which would allow convection to penetrate further north? These were some of the uncertainties being considered locally at this time.

On the following morning, 12Z Mixed Layer (ML) CAPE values based on SPC mesoanalysis data revealed a northwest to southeast stability gradient across Pennsylvania into the Delmarva region (Figure 2). This gradient was along the boundary left in wake of the previous day’s convection. MLCAPE values over western and south-central Pennsylvania ranged from 1000-1500 J/kg (fairly high for 8 am local time) while northeast Pennsylvania and practically all of New Jersey were stable with values less than 100 J/kg. In many instances, the preference is to view MLCAPE, in lieu of surface-based (SB) CAPE, as it tends to give a more complete view (assessing the mean parcel in the lowest 50-100 MB) of the stability profile, as opposed to focusing solely on a surface-based value. 

Observed regional soundings from 12Z also provided insight into the environment on either side of the boundary. The 12Z 8/28 KALB (Albany, NY) sounding (Figure 3) sampled relatively cool and stable conditions in the lowest layers of the atmosphere (surface - 850 MB), with a pronounced thermal inversion just above 850 MB. By contrast, the 12Z 8/28 KPIT (Pittsburg, PA) sounding (Figure 4) painted a much different picture, with moist and unstable conditions noted above an early morning, surface-based inversion. In fact, conditionally unstable lapse rates and strong speed shear are noted from about 925 MB all the way up to just below 500 MB. Given the likelihood for the atmosphere to further destabilize with daytime heating, this environment would favor the development of robust and locally severe convection within the next 6 hours or so, assuming sufficient lifting mechanisms would be in place.

The starkness between the two aforementioned soundings (12Z KALB & 12Z KPIT), and the different air masses on either side of the boundary, can be contextualized based on the precipitable water (PWAT) value. For example, the 12Z KALB sounding had a PWAT of 0.80”, which is around the 25th percentile for Albany in August. Contrast this to the 12Z KPIT sounding which had a PWAT of 1.74”, which is above the 90th percentile for August.

The impetus for the convection was a low-level wave approaching from the Great Lakes region (Figure 5). Meanwhile, Tropical Cyclone Laura’s remnants approached from the southwest. An enhanced south-southwest flow on the forward side of Laura’s circulation center resulted in a push of warm, moist, and unstable air into the low-level boundary from the Ohio Valley into western Pennsylvania. However, farther downstream in eastern Pennsylvania and New Jersey, the combined effects of cooler, denser, and stable air to the north and northeast, plus the nearly parallel orientation of the low-level boundary to the mean steering flow, resulted in a quasi-stationary frontal zone. An animation of Air Mass RGB imagery from 12Z 8/28 through 00z 8/29 showed the contrasting air mass characteristics on either side of the frontal zone (Figure 6).

Figure 6: GOES-16 Air Mass RGB valid 14:41Z - 22:56Z 8/28/20 with annotation at 14:41Z and 20:01Z.

At this point the main question was whether the stable air mass deposited by the cold pool would have a chance to modify over the next 12 hours? If so, would full modification or partial modification occur? In particular, would just the southern edge of the that air mass modify? The overall thinking was that modification would occur across the southern fringe of the stable region, while the cold pool would be harder to dislodge from locations in northern New Jersey and northeast Pennsylvania given the proximity to the stationary boundary and 300 MB jet streak (Figure 7). This would favor surface-based convection across central and southern New Jersey and southeast Pennsylvania.

As the morning progressed into the early afternoon it became clear that the cold pool deposited from the previous evening had modified slightly on the southern fringe, but not very much across central/ northern New Jersey. The GOES 16 Day Cloud Phase imagery revealed laminar-looking billow clouds across central/ northern New Jersey (indicative of a statically-stable Planetary Boundary Layer [PBL]) while across the Delmarva, Horizontal Convective Rolls (HCRs) were noted (indicative of mixing and a more unstable PBL) (Figure 8). Based on the SPC’s mesoanlaysis data at 18Z, MLCAPE corroborated the satellite analysis of stability (Figure 9).

Figure 8:  GOES 16 Day Cloud Phase imagery showing stable billow clouds across northern New Jersey and northeast Pennsylvania, while HCRs were noted over the Delmarva.

The warm, moist, tropical nature of the air mass just south and west of the aforementioned low-level frontal boundary would likely become a limiting factor for the potential of severe hail. A lack of dry air and the absence of substantial buoyancy in the hail growth zone (between -10 and -30C) would provide an unfavorable environment for hail production. This is a good example of where keen mesoanalysis could also narrow down the most likely individual storm threats. The convection was going to be weakly forced and with mostly uniform tropospheric flow, some organization into clusters was expected. (For a more detailed discussion on storm mode, see the separate blog post focused on the day prior with similar conditions). The primary threat looked to be damaging wind gusts. 

A major difference between the environment on this day and that of the day prior was the less favorable effective shear values in relation to the convective placement. On August 27th, effective bulk shear values were around 50 kts across New York, while effective shear values were only around 30 kts across the PHI CWA on 8/28 (Figure 10). This left some uncertainty on how impactful the event would be for the Mid-Atlantic. The shear was lower but still appeared to be sufficient. The overall thinking at this time though, was that the impacts to the region would be less than what was observed across Pennsylvania and New York on August 27th.

Through careful mesoanalysis, forecasters pinpointed the location and potential impacts of an outflow boundary which delineated a region of stable, convectively processed air and conditionally unstable air that could support organized deep, moist convection that afternoon. Further, by correctly diagnosing the thermodynamic profiles, they were able to identify damaging wind as the most likely threat. Finally, by recognizing the somewhat marginal magnitude of effective bulk shear, they determined the threat of widespread severe weather to be more limited than the previous day. All these factors provided them with the insight necessary to target their IDSS threat messaging to partners (Figure 11). In the future, WFO PHI will look to continue to provide more spatially targeted messages.

By the end of the day, much of the deep convection and heavy rainfall occurred near the leading edge of the slowly northward advancing tropical moisture plume associated with Laura (Figures 8 & 12). The advancement of the tropical air mass could be tracked using the Air Mass RGB referenced earlier (Figure 7).

The change in the moisture profile was impressive and can be quantified by comparing the 12Z 8/28 and the 00Z 8/29 KIAD (Washington Dulles) soundings. There is a significant increase in observed PWAT, from near 1.5” at 12Z 8/28 to over 2” by 00Z 8/29. (Note: KIAD, KPIT, and KALB are all labeled on the annotated images in the Air Mass RGB loop [Figure 7]). By the time the event was over there were multiple severe wind reports focused along and west of the fall line (Figure 13). And as anticipated, the severe hail threat was limited by the tropical air mass; very few severe hail reports occurred across the Mid-Atlantic Region. In totality, WFO PHI issued 14 severe thunderstorm warnings, 9 special marine warnings, and issued 94 local storm reports.

Authored by Brian Haines (WFO Philadelphia/Mt. Holly) and Micheal Jurewicz (WFO State College); edited by Ryan Difani (OU CIMMS/OPG); satellite imagery and animations courtesy of Matt Foster (OPG). Special thanks to the OPG team for their contributions, reviews, and support.