The 24 September 2001 Fabius, NY Tornado
Ray Brady, NWS Binghamton
A brief, weak (F0) tornado touched down on September 24th 2001 near the town of Fabius in south central Onondaga county (Fig. 1 - map of northern Cortland and southern Onondaga county. Fabius is located east of Route 81 on State Route 80 in the upper right hand corner of the map). This tornado was the only severe weather reported on this afternoon, although flooding occurred later that evening over Oneida and Madison counties in central New York. This case was very interesting for a number of reasons: 1) tornadoes are relatively uncommon in upstate New York, especially during the "cool season" or fall months; 2) the synoptic system responsible for this storm was rather intense, with the potential for not only severe convection but also for flooding, which provided additional challenges for the forecasters on this day; and 3) the rapidly evolving tornadic storm displayed high precipitation (HP) supercell-like characteristics prior to touchdown.
This note will describe the development of the Fabius tornado. The tornado track and damage caused by it will be described in Section 2, and the synoptic setting will be discussed in Section 3. WSR-88D reflectivity and velocity data will be used to describe the evolution of the parent storm in Section 4, and comparisons to other HP supercell events will be shown in Section 5. A summary of the event concludes this note in Section 6.
2. The Fabius tornado
An F0 tornado touchdown on September 24th 2001 near the town of Fabius was confirmed by NWS meteorologists and WIXT meteorologist Dave Eichorn.(see the LSR and PNS following this paragraph). The tornado crossed Route 80 from the south near Goodrich Road around 2:35 pm (time tweeked from LSR after perusal of radar data) with a spotty damage path of 1/2mile and width of approximately 75 yards (Fig.2 - detailed map of the tornado track as indicated by the purple triangles). Maximum wind speeds were estimated to be 50 to 70 mph. Some of the nearby residents reported a loud roar and rotation in the clouds as the storm passed by. No other reports of severe weather were received on this day.
NWUS51 KBGM 251931
PRELIMINARY LOCAL STORM REPORT
NATIONAL WEATHER SERVICE BINGHAMTON, NY
330 PM EDT TUE SEP 25 2001
TIME(EDT) .....CITY LOCATION.....STATE ...EVENT/REMARKS... ....COUNTY LOCATION....
0245 PM FABIUS NY TORNADO ONONDAGA
BRIEF F0 TOUCHDOWN AT INTERSECTION OF GOODRICH RD AND ROUTE 80. TREES DOWNED...SMALL STORAGE SHED DESTROYED.
TTAA00 KBGM DDHHMM
PUBLIC INFORMATION STATEMENT
NATIONAL WEATHER SERVICE BINGHAMTON NY
350 PM EDT TUE SEP 25 2001
...NATIONAL WEATHER SERVICE VERIFIES A BRIEF F0 TORNADO TOUCHDOWN IN SOUTHERN ONONDAGA COUNTY ON SEPTEMBER 24TH...
A BRIEF TORNADO TOUCHDOWN WAS VERIFIED NEAR THE INTERSECTION OF GOODRICH ROAD AND ROUTE 80 IN THE TOWN OF FABIUS. THE TORNADO OCCURRED AROUND 245 PM ON MONDAY AFTERNOON. SPOTTY DAMAGE WAS OBSERVED ALONG A 1/2 MILE PATH...WITH AN AVERAGE WIDTH OF 75 YARDS. MOST OF THE DAMAGE WAS CONCENTRATED AT A RESIDENCE NEAR THE INTERSECTION OF ROUTE 80 AND GOODRICH ROAD WHERE SEVERAL LARGE TREES WERE DOWNED AND A SMALL SHED WAS DESTROYED. MAXIMUM SUSTAINED WINDS WERE ESTIMATED TO BE 50 TO 70 MPH.THE NATIONAL WEATHER SERVICE WOULD LIKE TO THANK WIXT-TV METEOROLOGIST DAVE EICHORN AND LOCAL RESIDENTS FOR ASSISTING IN THE STORM SURVEY.
The damage path began immediately south of Route 80 where a tractor trailer containing a small payloader was overturned. A small aluminum shed was also destroyed; a piece of it was wrapped around a telephone pole on Route 80. (Fig. 3 -Picture of damage at the start of the tornado track looking north. Note the piece of siding wrapped around the telephone pole above Bob's head! Residence that received the brunt of the storm can be seen to the north across the street).
Most of the damage occurred at a residence near the intersection of Route 80 and Goodrich Road where several, large soft maple and pine trees (mainly spruce and soft pine) were felled (Fig. 4 - picture of tree damage at the residence across the street. Note the shearing evident in the damaged trees and the different directions in which they were felled. Also note the white pole from the shed across the street at the bottom of the picture). However, several of the large soft maples situated in front of the house were rotted on the inside (Fig. 5 - Picture of damaged large maple trees - note the large bees nest in the tree directly in front of Dave Eichorn which became exposed when the tree split) . Damage to the house was minor. The tornado ascended a small hill north of this residence but the damage was spotty; pieces of aluminum from the shed were strewn here (about 1/4 mile north of the shed's location) and some of the larger trees were sheared off.
3. Synoptic Setting
A potent storm system was approaching the region on 24 September. The 12Z surface chart (Fig. 6 - large scale 12Z sfc chart) showed low pressure over southern Ontario, with a trailing cold front extending southward into the Gulf States. Surface charts at 13Z (Fig. 7 - local surface chart) and 21Z (Fig. 8 - local surface chart) show the progression of this front into western New York by late afternoon. Note that surface dewpoints across central New York, which were in the upper 50s at 13Z, rose into the lower and mid 60s by afternoon. The cold front was forecast to move slowly across central New York during the overnight hours.
Upper level charts (Fig. 9 - the upper left panel shows 850 mb heights/temps and winds; the upper right panel shows 300 mb heights and winds; the lower left panel shows surface pressure and thickness; and the lower right panel shows lifted indices and 850 mb RH) indicated a deep 300 mb trough centered over the Great Lakes with a 110 kt jet lifting into Ontario. Central New York was forecast to be in the right entrance region of this jet during the afternoon, providing strong upper level divergence. A closed 850 mb low was located near Georgian Bay with a +12 to +14 thermal ridge extending into central New York. The eta lifted index chart showed an axis of 0 to -2 LI's extending from the mid Atlantic states into western and central New York. The 12Z IR satellite picture (Fig. 10 - a rather poor 12Z national IR image ) showed a plume of moisture associated with the front, extending northward from the Gulf of Mexico into the eastern Great Lakes.
Forecast soundings from the 12Z Eta run valid at 18Z for Elmira (ELM, Fig.11), Ithaca (ITH, Fig. 12), and Syracuse (SYR, Fig. 13) indicated the potential for strong and possibly severe convection. Forecast CAPES were modest but in the range of other cool season severe weather events, ranging from 200-300 j/kg at ELM and ITH, to just over 600 j/kg at SYR. Forecast lifted indices ranged from around 0 at ITH and ELM to -2 at SYR. Wind profiles displayed strong directional and speed shear in the 0-3 km layer. Bulk Richardson Numbers (BRN's) at Elm and ITH were around 3, indicating that the wind shear might overwhelm the available bouyancy, effectively "shearing off" any storms. However, the BRN at SYR was 8, which indicated the potential for supercells. Helicities ranged from 150 to 200, at the lower end of the range indicating the potential for mesocyclone formation. Energy-helicity index (EHI) values, which relate CAPE to helicity, were below 0.5 at ITH and ELM, but up to 0.65 at SYR. Values less than 2 normally indicate the mesocyclone tornadoes are unlikely (based on Great Plains research), but we've seen more than one case where values approaching 0.75 or 1.0 were associated with severe weather and tornadoes in central New York and northeast Pennsylvania.
Given the ample supply of moisture, a slow moving front nearly parallel to the upper level flow and strong dynamics, the main concern for the mid shift forecaster was the threat of heavy rain and flash flooding (see below).
FXUS61 KBGM 240708
AREA FORECAST DISCUSSION
NATIONAL WEATHER SERVICE BINGHAMTON NY
310 AM EDT MON SEP 24 2001
RADAR IMAGERY SHOWG LARGE MASS OF PRECIP ACRS OH/TENN VLLY WORKG NE IN ADVANCE OF A DIGGING UPR LVL TROF AND LL FRNTL SYS. THE WHOLE TROF AND FRNT SYS SLIDES FARTHER E TODAY AND BEGINS TO DEEPEN IN TIME. A LLJ ACCELERATES IN RESPONSE TO THIS AND RUNS THRU THE REGION BY TONIGHT. THIS LLJ FOCUSES A RIBBON OF HI THETA-E AIR INTO THE SLOW MOVG FRNTL SYS WHICH IS ALMOST PARRALLEL TO THE UPR LVL FLO. AT UPR LVLS THERE IS A STRG JET ENTRANCE REGION TO CREATE THE NECCESSARY UPR DVRG IN SUPPORT OF SIGNFT RAINFALL. MODELS ARE PAINTING A BAND OF HVY RAINFALL ASSCTD WITH THIS DYNAMICAL SET-UP. ETA HAS THE BAND SPLITING THE CWA IN HALF WITH AMNTS OVR 2 INCHES. THE AVN HAS SIMILAR AMNTS BUT IS FARTHER WEST WITH THE BAND OF HVY PRECIP. THE NGM HAS SIMILAR POSITIONING OF THIS FEATURE AS THE ETA BUT HAS LESS QPF (TYPICAL MODEL BIAS). STILL FEEL POTENTIAL IS THERE FOR HVY RAIN DESPITE IN CWA DISAGREEMENT AMONG AVN AND ETA/NGM IN THE QPF FIELDS. THINK MOST AREAS SEE 1-2 INCHES OF RAIN WITH LOCALLY 3-5 INCH AMNTS AS POTENTIAL IS HI FOR A NARROW BAND OF HEAVY RAINFALL AND THUNDERSTORMS SOMEWHERE. WHEREEVER THE BAND OF HVY RAIN SETS-UP WILL WHERE THE HEAVIEST RAIN FALLS. AT THIS POINT IT IS TOO EARLY TO PINPOINT WHERE THIS WILL BE. THEREFORE NO WATCHES YET. WILL CONT FLD POTENTIAL OUTLOOK.
THINK A FEW SCTRD SHRA WILL DVLP FM W-E THIS MORNING INTO THE AFTRN AS MAIN BULK OF PRECIP REACHES WRN CWA BY MID AFTRN AND EASTERN CWA BY TONIGHT. MODELS ARE SHOWG MODEST INSTABILITY THIS AFTRN AS THE FIRST MAIN BATCH OF PRECIP ARRIVES SO WITH MENTION CHC TSRA THIS AFTRN AND EVE. MODELS DO STABILIZE THE AMS A BIT BY TONIGHT BUT STABILITY REMAINS LO (SLIGHTLY ABV ZERO) AND GIVEN THE STRG LLJ THUNDER WILL BE PSBL TONIGHT.
The forecast discussion issued for the morning update recognized the potential for isolated severe weather (see below).
FXUS61 KBGM 241407
AREA FORECAST DISCUSSION
NATIONAL WEATHER SERVICE BINGHAMTON NY
1005 AM EDT MON SEP 24 2001
WILL MAKE SOME TWEAKS TO CLEAN UP WORDING AND ADJUST TIMING OF PCPN IN SOME ZONES BUT OVERALL FORECAST IS IN GOOD SHAPE. MODELS CONTINUE TO HAVE THEIR DIFFERENCES BUT ARE IN GENERAL AGREEMENT IN BRINGING HEAVY PCPN TO THE REGION LATE TODAY AND TONIGHT. THUNDERSTORMS ARE ABOUT TO ENTER THE WRN CWA AND EXPECT INSTABILITY AND APPROACHING STRONG DYNAMICS TO KEEP SOME THUNDER AROUND THE REGION THROUGHOUT THE EVENT. WIND PROFILES AND CONVECTIVE PARAMETERS INDICATE THE POTENTIAL FOR SOME OF THE CONVECTION TO BECOME ORGANIZED ON A LOCAL LEVEL. WITH STRONG DYNAMICS PRESENT THERE WILL BE A CHANCE FOR ISOLATED SVR WX. WILL HOLD OFF ON DECISION TO ISSUE FFA UNTIL NEW MODEL GUIDANCE COMES IN TO SEE IF THERE WILL BE MORE OF A CONSENSUS ON EXACTLY WHERE THE HEAVY PCPN WILL BE FOUND.
A severe weather outlook was subsequently issued discussing the possibility of strong and damaging winds with thunderstorms (see below).
WWUS81 KBGM 241428
SEVERE WEATHER OUTLOOK
NATIONAL WEATHER SERVICE BINGHAMTON NY
1025 AM EDT MON SEP 24 2001
...THERE IS A SLIGHT RISK OF SEVERE THUNDERSTORMS TODAY...
A STRONG STORM SYSTEM APPROACHING FROM THE WEST WILL SPREAD SHOWERS AND THUNDERSTORMS INTO CENTRAL NEW YORK AND NORTHEAST PENNSYLVANIA. THE PRECIPITATION WILL SPREAD ACROSS THE CENTRAL SOUTHERN TIER OF NEW YORK THIS MORNING...REACH THE INTERSTATE 81 CORRIDOR EARLY THIS AFTERNOON...AND THEN ARRIVE IN THE POCONOS AND WESTERN CATSKILLS LATE IN THE DAY. THE PRECIPITATION WILL INCREASE IN INTENSITY AND BE HEAVY AT TIMES THROUGH MIDNIGHT. DURING THIS TIME...SOME OF THE THUNDERSTORMS WILL HAVE THE POTENTIAL TO PRODUCE LOCALIZED SEVERE WEATHER FROM STRONG DAMAGING WINDS. STAY TUNED TO NOAA WEATHER RADIO OR OTHER LOCAL MEDIA OUTLETS FOR THE LATEST ON THIS SITUATION. ADDITIONAL INFORMATION CAN BE FOUND ON THE NWS BINGHAMTON WEB SITE AT HTTP://www.erh.noaa.gov/er/bgm.
The afternoon forecast discussion also mentioned the continued threat for localized wind damage and flooding (see afdbgm below), and a flash flood watch was issued for the entire BGM county warning area at 320 pm (not shown).
FXUS61 KBGM 241833
AREA FORECAST DISCUSSION
NATIONAL WEATHER SERVICE BINGHAMTON NY
230 PM EDT MON SEP 24 2001
12Z RUN SHOWS MODELS COMING TO CONSENSUS WITH THE CURRENT EVENT. DEEPENING TROF MOVING OUT OF THE UPPER MIDWEST WILL CLOSE OFF OVER THE GREAT LAKES REGION AND DO THE SIT AND SPIN TO OUR WEST THROUGH
THE FORECAST PERIOD. ASSOCIATED SFC LOW WILL MOVE OVER THE REGION TONIGHT AND THEN GET SUCKED BACK TO THE NW ON TUESDAY. VERY STRONG DYNAMICS WILL BE CENTERED ON THE CWA BY 00Z WITH STRONG UVV...
Q-CONV...AND UPPER DIV DRIVING THE PCPN. THE SYSTEM ALSO HAS A DEEP MOISTURE FEED FROM THE GULF AND ATLANTIC. CURRENT RADARS SHOW HEAVIEST PCPN BISECTING THE REGION MORE OR LESS ALONG THE I81
CORRIDOR BUT AREAS TO THE EAST ARE STILL RELATIVELY DRY. THE CURRENT SITUATION HAS ENOUGH INGREDIENTS WARRANTS A FLASH FLOOD WATCH DESPITE THE RELATIVELY DRY ANTECEDENT CONDITIONS. THE HEAVIEST
RAIN FROM NOW THROUGH MIDNIGHT WILL LIKELY BE FROM I81 EAST. THE MOST VULNERABLE AREAS ARE ACROSS THE CATSKILLS AND POCONOS BECAUSE THEY HAD A COUPLE OF INCHES OF RAIN DURING THE LAST EVENT. HOWEVER
SIGNIFICANT PCPN THIS MORNING MAY HAVE BEEN ENOUGH TO PRIME THE WESTERN AREAS AND ADDITIONAL PCPN EXPECTED TONIGHT MAY BE ENOUGH TO CAUSE SOME PROBLEMS. WILL THERE FORE GO WITH FAA FOR THE ENTIRE CWA BUT DELINEATE I81 EAST AS THE PRIME THREAT AREA. WITH STRONG DYNAMICS EMBEDDED THUNDER IS A GOOD POSSIBILITY TONIGHT AND WITH GOOD WIND PROFILE SOME LOCALIZED WIND DAMAGE IS POSSIBLE IN THE STRONGER CELLS.
.BGM...FFA ALL ZONES TONIGHT.
In addition, a tornado watch was issued for northeast Pennsylvania at 353 pm. Although several cells displayed cyclonic rotation, no warnings were issued and no reports of severe weather were received. However, flash flood warnings were issued that evening for Oneida county at 1025 pm, and for Madison county at 1030 pm (not shown). Rainfall amounts were impressive, especially along the I-81 corridor where reports of 4 to 7 inches were common.
4. Radar data
The radar evolution of the parent storm that produced the Fabius tornado was rather interesting. It was embedded within a large area of rain with embedded heavier convective cells (Fig. 14 - 1818Z larger scale 0.5o reflectivity - the storm of interest is over northwest Cortland county). A radar loop zoomed on the Fabius storm provides a good overview of the evolution of the low level reflectivity features (Fig. 15 - 0.5 reflectivity loop from 1818Z to 1906Z - times are at the bottom of the image - loop pauses on the 1st frame and will run continuously until you close out).
The storm looks structurally amorphous at 1818Z (Fig. 16 - 1818Z 0.5 reflectivity) but by 1824 (Fig 17 - 1824Z 0.5 reflectivity) and especially at 1829Z (Fig. 18 - 1829Z 0.5 reflectivity) the higher reflectivity area begins to display a "kidney-bean" appearance. The structure looks quite interesting at 1834Z (Fig. 19 - 1834Z 0.5 reflectivity) , when an appendage (or brief hook echo) is observed on the southeast flank of the storm. The storm loses its HP characteristics as it transitions to a more of a bow type configuration with a rotating comma head and trailing comma tail after 1840Z (more on this later).
Storm relative velocity data (default storm motion was 205o at 32 kts) indicated a fairly long-lived and well developed low level mesocyclone with the Fabius storm prior to touchdown (Fig. 20 - 0.5 deg storm relative velocity (SRV) loop from 1818Z to 1906Z). The low level circulation over north central Cortland county was initially broad (Fig.21 - 1818Z 0.5 deg SRV) but became smaller and strengthened rapidly after 1829Z as it approached the Cortland/Onondaga county border (Fig. 22 - 1829Z 0.5deg SRV), with strong gate to gate shear evident at 1834Z (Fig. 23 - 1834Z 0.5deg SRV - note the strong gate to gate shear over south central Onondaga county with a -22 to -30 pixel adjacent to a +22 to +30 pixel) and 1840Z (Fig. 24 - 1840Z 0.5deg SRV ) - signature has weakened a bit due to weaker outbound velocities) over south central Onondaga county.
A closer inspection of the upper level reflectivity and velocity data through the use of 4-panel displays reveals some interesting features prior to tornado development. (If the 4-panel base velocity displays look too dark, try increasing the brightness of your monitor). At 1802Z reflectivity data from 0.5 to 3.4o (Fig. 25 - 4 panel reflectivity display), showed an eastward tilt with height of the max reflectivity core and small areas of enhanced reflectivity gradient on the southern flank, but no other significant features were observed. (However, could the small area of lower reflectivity just to the NW of Homer at 1.5, 2.4 and 3.4o be a reflection of a weak echo region [WER?]). The base velocity data (Fig. 26 - 4 panel base velocity display) showed stronger southerly flow over central and eastern portions of Cortland county as compared to western sections, and implied the presence of a low level cyclonic shear/vorticity zone. The storm appeared to be situated along this shear zone, and cyclonic shear was associated with the storm, especially evident at 1.5 and 2.4o. These circulations show up a little better in the relative velocity data (Fig. 27 - 1802Z reflectivity and relative velocity at 0.5 and 2.4o).
Little change was noted in the reflectivity data at 1807Z (not shown), although velocity data indicated some strengthening of the southerly flow over central Cortland county to the east of Homer. Significant and rapid changes began to occur at 1813Z, when the high reflectivity cores at 1.5 and 2.4o began to take on a kidney-shaped appearance (Fig. 28 - 1813Z 4 panel reflectivity - note the well defined reflectivity signature at 1.5 deg). It is important to note that this reflectivity signature, one that is often noted with HP supercells, occured 5-10 minutes before the 0.5o reflectivity echo assumed a similar appearance. Also note the development of the N-S line of echoes extending from the extreme SE flank of the storm, which I'll refer to as the 'pendant' echo. This development of this echo may be related to convergence along or slightly behind the advancing rear flank downdraft, or convergence lines in the low-level storm inflow. As will be seen in the following paragraphs, it appears that this echo was situated near the leading edge of the rear flank downdraft.
The strengthening of the southerly flow observed in the base velocity data continued (Fig. 29 - 1813Z 4 panel base velocity), with an area of enhanced "away" velocities (36 to 50 kts) just east of Homer especially evident at 1.5 and 2.4o. This mid-level (2.4 deg) circulation was located on the SE flank of the storm near the mid-level appendage (Fig. 30 - reflectivity and relative velocity at 1813Z), and the data indicated that the circulation was contracting, especially when compared to the 2.4o circulation observed at 1802Z ( see Fig. 27).
By 1818Z (Fig. 31 - 1818Z 4 panel reflectivity) the low level reflectivity began to display more of a kidney-bean appearance, and a notch of lower reflectivity, possibly signaling the development of a rear-flank downdraft, began to form on the southwest flank of the storm, just to the northeast of Homer. The maximum reflectivity and areal coverage of the 3.4o echo began to decrease, suggesting that the upper level core may have been descending, therefore supporting the development of the rear flank downdraft. (Yes, the storm was moving away from the radar, but I don't believe this was the main reason the upper level returns decreased). The mesocyclone continued to intensify, and pixels of 50-64 kts were now observed on the base velocity data between Homer and Truxton at 1.5 and 2.4o (Fig. 32 - 1818Z 4 panel base velocity).
The apparent collapse of the 3.4o reflectivity core continued at 1824Z (Fig. 33 - 1824Z 4 panel reflectivity). A channel of lower reflectivity (15-20 dBZ) now split what remained of the 3.4o reflectivity core and the 50+ dBZ returns at 2.4o also began to shrink. The notch on the SW flank of the storm remained will defined up to 2.4o, as did the kidney bean-shaped appearance at 0.5o. The mid level mesocyclone appeared to have weakened a bit as pixels indicating flow of 50-64 kts were no longer evident at 1.5o and 2.4o, but strengthening was noted at 0.5o, especially in the immediate Truxton vicinity (Fig. 34 - 1824Z 4 panel base velocity). This is more readily apparent in the storm relative velocity data (Fig. 35 - 1824Z reflectivity and storm relative velocity).
The reflectivity cores at 2.4o and 3.4o continue to decrease at 1829Z (Fig. 36 - 1829Z 4-panel reflectivity). The weak echo reflectivity channel (i.e., the "split") which was first noted at 3.4o had developed at 2.4o between Truxton and Tully (Fig. 36), and was also beginning to appear in the 1.5o data. The 0.5o deg echo now displays an appendage or developing hook echo on the SE flank of the storm. The base velocity at 1829Z (Fig. 37 - 1829Z 4-panel base velocity ) indicates little change in the strength of the mid-level mesocyclone, but the circulation at 0.5o has tightened up considerably while strengthening just to the southeast of Tully (Fig. 38- 1829Z reflectivity and storm relative velocity).
Upper level reflectivities continued to decrease at 1834Z (Fig.39 - 1834Z 4-panel reflectivity); the weak echo reflectivity channel was still evident at 1.5 and 2.4o (and the argument could be made that it is also now developing at 0.5o). The 0.5o echo still showed a hook echo on the SE flank of the storm and although less well-defined at this time, a weak HP structure is still evident at 0.5o. The pendant echo extending SSW from the hook also remained well defined (it is this echo that develops into the southern portion of the soon-to-be line). The base velocity data (Fig. 40 - 1834Z 4-panel base velocity data) showed a tight 0.5o circulation with strong gate-to-gate shear located just a few miles east-northeast of Tully, nearly over Fabius. Based on this data, it appears that the F0 tornado touched down near Fabius at approximately 1835Z, about 10 minutes earlier than the initial storm report indicated. Figure 41 (Fig. 41- 1834Z 4-panel reflectivity and storm relative velocity) shows the 0.5o circulation was located on the SE extension of the low-level hook echo.
The 1.5o weak echo reflectivity channel remained well defined at 1840Z (Fig. 42 - 1840Z 4 - panel reflectivity) , and also showed a better-defined reflection at 0.5o at 1840Z. The overall structure at this point looks rather unorganized, although the remnants of the hook echo are still visible at 0.5o. The circulation remains well defined and quite strong at 0.5o although the pixels of strongest inbound and outbound velocities are no longer horizontally adjacent. However, it has weakened considerable aloft (see Fig. 43 - 1840Z 4-panel base velocity and Fig. 44 - 1840Z reflectivity and storm relative velocity). Note development of another circulation associated with the cell over extreme southeast Onondaga county (i.e., along the old pendant echo), especially evident in the 1.5o velocity data as a separate area of enhanced "away" velocities.
The Fabius storm began its transition into a bowing line segment at 1845Z as it continued to move northeast (Fig. 44 - 1845Z 4-panel reflectivity). Base velocity data (Fig. 45 - 1845Z 4-panel velocity) at 1.5o still showed 2 separate areas of enhanced "away" velocities across eastern Onondaga county. By 1850Z, the comma head echo and circulation (mesocyclone) associated with the original Fabius storm had moved to a position just south of Manlius (Fig. 46 - 1850Z 4-panel reflectivity), while the other circulation maintained itself near or just south of the "break" in the line over extreme southeast Onondaga county (Fig. 47 - 1850Z 4-panel base velocity). It was this latter storm which now displayed the most impressive reflectivities aloft.
This latter storm (Fig. 48 - 1906Z reflectivity) continued to be associated with a low level circulation (Fig. 49 - 1906Z SRV) as it moved northeast into west central Madison county. This circulation was also sampled in the 0.2o base velocity data from the WIXT Doppler Radar at 305 pm (Fig. 50 - 1905Z WIXT base velocity - radar location is over east central Onondaga county - velocity scale is on the bottom of the figure). Although the corresponding reflectivity data was unavailable, this base velocity data clearly showed the original circulation with the Fabius storm just NW of Chittenango (large area of enhanced away velocities next to much weaker 'away' velocities), and the 2nd circulation represented by the velocity couplet mid-way between Chittenango and Cazenovia. This "secondary" circulation was located near a break (area of lower reflectivity) within the line, and is strikingly similar to the "broken-S" signature, which has been observed in other cool-season tornado or severe wind producing thunderstorms. However, no tornado or wind damage was reported with this storm as it headed northeast into Madison county. Please see the 4-page conference paper by McAvoy et al. for more details on this radar signature, which can be accessed here.
5. Lightning data and WSR-88d algorithm performance
A few words on the lightning data (not archived unfortunately) and WSR-88D algorithm performance on this day. Earlier in the day, the lightning data showed frequent strikes with a cluster of thunderstorms over eastern Lake Ontario which moved ashore into Jefferson county around noontime. Severe thunderstorm warnings were issued for these storms (we were in backup mode for BUF). During the afternoon hours when the Fabius storm developed, C-G lightning strikes were very infrequent and isolated. However, it is interesting to note that during the 30 minutes preceding the touchdown, this was the only storm producing C-G lightning, albeit only a stroke or two every 5 to 10 minutes.
The WSR-88D never identified a TVS with this storm, but the mesocyclone algorithm did an admirable job identifying the parent circulation associated with the Fabius storm. In the hour surrounding the tornado touchdown, the only mesocyclone identified in the BGM CWA was associated with this storm, with a positive identification for 3 consecutive volume scans prior to touchdown. It is interesting to note the algorithm trends concerning the depth and width of the mesocyclone during this period. As with any algorithm output, you must be cognizant of the errors introduced by radar sampling and the shortcomings of the algorithm itself. However, the algorithm appears to have correctly detected the trend of mesocyclone intensification, followed by a weakening phase (associated with concurrent collapse in storm top as implied by the upper level reflectivity data). At 1813Z, (Fig. 51 - 1813Z 1.5deg SRV with mesocyclone algorithm output - algorithm output at the top of the page), the algorithm indicated a base at 4.9kft and top at 7.7kft for a depth of 2.8kft, with an average diameter of 3.05 n mi. At 1818Z (Fig. 52 - 1818Z meso output) the algoroithm indicated that the circulation had expanded vertically, with a base at 1.9 kft and top of 8.5kft, for a depth of 6.6kft. The estimated diameter was 2.5 n mi. The 1824Z algorithm output (Fig. 53 - 1824Z output) suggested that circulation was decreasing in depth from the top down. The base remained near 2kft but the top of the circulation was estimated at only 5.6kft. It was at this time that the radar data implied the development of a rear flank downdraft at low levels.
5. Comparisons to other HP storms
The structure of the storm in the base reflectivity data displays some similarities to one of the "composite life cycles" of an HP supercell as illustrated by Moller et al. (1990) (Fig.54 - figure from Moller et al. showing 2 "composite life cycles of an high precipitation (HP) supercell storm), especially to stages 1 through 8a. Also note the similarity in the appearance between the plan view schematic of an HP supercell storm as shown by Moller et al. (1994) (Fig. 55 - Moller schematic of an HP supercell) to the reflectivity pattern observed at 1.5o at 1813Z (Fig 28 - 1813Z 4-panel reflectivity) or at 1818Z at all elevation angles (Fig. 31 - 1818Z 4-panel reflectivity).
The Fabius storm also exhibited some similarity to an event described by Funk et al. (1999), where an embedded HP supercell-like structure evolved into rotating comma head-comma tail pattern. The Fabius storm shows some similarities to the 1600Z-1624Z sequence shown in his paper (Fig. 56 - Sequence of 0.5o reflectivity images - contoured at 10 dBZ starting at 30 dBZ - returns >40 shaded).
Moller at al. (1990), discussing the operational real-time detection of HP supercells note that "severe weather forecasters must be aware of the variations of rotating storm reflectivity patterns. It is critical to understand that a storm with persistent 'inflow notch' (WER) or kidney-bean shape on the east flank may be an HP supercell..... Forecasters should be alert for a rapid HP-BE (bow echo) transition. Echo distortion due to intrusion of the rear-inflow jet and resultant upshear weak-echo channel could mean an increase in violent downburst potential". These statements appears to summarize our scenario rather nicely. They also state in this paper that other life cycles and events are possible within the supercell continuum. Moller et al. (1994) state that "as the mesocyclone moves with respect to the precipitation, inflow notches, spiral bands, and other complicated reflectivity structures can arise....(see Fig. 57 for some examples of these echo configurations). The key , of course, is the presence of a mesocyclone....". Forecasters should also recognize that "HP storms may exhibit their mesocyclones closer to the front flank (relative to storm motion)".
6. Summary and Discussion
A rapidly evolving storm displayed HP supercell characteristics before producing a brief and weak (F0) tornado in the town of Fabius on September 24th 2001. Radar data showed a kidney-bean appearance to the low and mid reflectivity core, a weak reflectivity echo channel aloft which subsequently developed at lower levels, a pendant and hook echo, and indications of a collapsing top prior to the touchdown. The tornado appeared to be of the "classic" non-descending type, as WSR-88D velocity data indicated the mesocyclone was initially strongest aloft, with subsequent strengthening of the mesocyclone at low levels, and ultimately the development of a TVS-like signature at 0.5o. The WSR-88D mesocyclone algorithm identified a persistent circulation prior to the occurrence of the severe weather. Unfortunately, severe weather warnings were not issued for this storm , but it appears that we could have issued a Severe thunderstorm or tornado warning with a 10-15 minute lead time if the HP supercell signatures had been quickly identified in real-time (for example, with the 1818Z data), especially given the the synoptic environment, which was favorable for this type of low- topped supercell development. (Of course it's easy to be a Monday Morning Quarterback when it's quiet, you have all the time need to peruse the data, and you don't have several storms on the scope at once).
The only report of severe weather occurred at Fabius, and it could be debated whether a tornado or severe thunderstorm warning would have been more appropriate, since the straight line wind damage was as impressive as the damage caused by the tornado. There are several opinions on this issue; mine is that a severe thunderstorm warning would have been fine for this storm since the tornado was brief and weak. But of course we always need to upgrade a severe thunderstorm warning if radar data (or other data sources) indicate a greater threat for additional tornado development or damage greater than F0 in intensity.
Moller, A.R., C.A.Doswell 3rd, and R. Przybylinski, 1990: High-Precipitation Supercells: A conceptual model and documentation. Preprints, 16th Conference on Severe Local Storms, Kananaskis Park,Alberta, Canada, Amer. Meteor. Soc., 53-57.
Moller, A.R., C.A. Doswell 3rd, and M.P. Foster, and G.R. Woodall, 1994: The operational recognition of supercell thunderstorm environments and storm structures. Wea. Forecasting, 9, 327-347.
Funk, T.W., K.E. Darmofal, J.D. Kirkpatrick, V.L. DeWald, R.W. Przybylinski, G.K. Schmocker, and Y. Lin, 1999: Storm reflectivity and mesocyclone evolution associated with the 15 April 1994 squall line over Kentucky and southern Indiana. Wea. Forecasting, 14, 976-993.
McAvoy, B.P., W.A. Jones, and P.D. Moore, 2000: Investigation of an unusual storm structure associated with weak to occasionally strong tornadoes over the eastern United States. Preprints, 20th Conference on Severe Local Storms, 182-185. Can be accessed at broken-S.pdf.