Convective rain with tropical character in Ontario

Aug 10, 2004
Dundas Ontario
This summer I have been struck by how much of our rainfall in Ontario has had tropical characteristics.

Many convective events occurring during the summer have produced rainfall amounts that match radar estimates produced by using the Rosenfeld Tropical Z/R relationship Z= 250®^1.2, rather then the usual summer deep convection equation used by the NWS Z=300®^1.4 or the general (Marshall/Palmer Stratiform Z/R) relation Z=200®^1.6 that is often used.

In the Hamilton area, 2 damaging convective rainfall events occurred, the first one on July 26 lasted up to 3 hours and was caused by a training event that erupted in an airmass that was moderately unstable and had very high PW values of ~50 mm or 2 inches. This event produced no hail or damaging wind reports that I am aware of. The lack of strong downdrafts could have been anticipated, considering that almost all of the troposphere below 500 mb was quite moist. Still, these storms produced abundant electrical activity, indicating that there was significant ice in the upper parts of the updrafts and that the cells were not exclusively the result of warm cloud layer rain coalescence processes. IR satellite indicated very cold anvils erupted over at least some of the cells in this training system. The Buffalo WSR-88D indicated very little area covered by reflectivity of > 50 DBZ but very large areas covered by 40-45 and 45-50 DBZ values. The Buffalo WSR-88D was using the summer deep convection Z/R relation during this storm, and rainfall totals were woefully underestimated. The Buffalo WSR-88D estimated a narrow axis of 1+ inch amounts over Hamilton Mountain, when actual amounts exceeded 2 inches.

On Aug 19, two long lived HP supercells tracked E across Southern Ontario. The northern one produced at least two long track F2 tornadoes that were so hidden by rotating shafts of rain and hail that almost no one actually saw the tornadoes. Very low LCL's and haze made visibility even worse.

The south suppercell's meso crossed the south edge of the City of Hamilton. As the storm slowly passed through the city it brought torrential rainfall and damaging flooding to the City. There were reports of isolated marble sized hail on the southern edge of the city, nearest the meso, but as the FFD passed right over the city it seemed not to produce much straight line wind damage.

On the Buffalo WSR-88D, this storm had a persistent but relatively small core of 50-55 DBZ reflectivities, while the area of 45-50 DBZ reflectivity’s was much larger. The radar again was using the summer deep convection relation for rainfall estimation, but again woefully underestimated precipitation, even although some hail was present. The rainfall estimate indicated a narrow strip of 1+ inch amounts, while people in Hamilton reported 2.0-2.4 inches of rain in less then 60 minutes, in step with the dangerous and destructive flooding that was observed. Again, the rainfall would have been better estimated by applying the Rosenfield tropical Z/R relation to this storm.

These were but two of the more conspicuous examples of rainfall showing tropical characteristics in Ontario this summer. Other convective events dropped hints of tropical behavior...suffocating deluges of warm rain that seemed to be composed of relatively small drops, high ratios of rainfall to lightning flash density, and a lack of significant cool outflow from convective cells.

I am left wondering if the summer deep convection Z/R relation is more appropriate for thunderstorms that occur in environments conductive to classic severe convective weather.....very high (3000 J/Kg) mixed layer CAPE, abundant dry air above the mixed layer, with moderate (1-1.5 inch) PW's. Thunderstorms producing large amounts of damaging hail and severe, very cold downdrafts occur in these environments.

Certainly this environment is not characteristic of most convection E of the great plains during the summer. Even the tornado outbreak of August 19 differed from the classic Great Plains setup. The 0-5 Km SRH was quite high (300-500 m2/s2) but the mixed layer cape was not extreme, perhaps only 1000-1500 J/Kg. Also the air above the ~100 mb thick surface based mixed layer was not very dry, lots of alto-cumulus was visible at times, and RUC and NAM model output supported PW's of 2 inches or more and surface dewpoints were very high...about 75 (24C).

I would be glad to hear from anyone who has a hunch about the applicability of the summer deep convection Z/R vs. the Rosenfeld Tropical Z/R. in different convective environments.