Adapted from Preprints, 23rd AMS Conference on Hurricanes and Tropical Meteorology, Dallas, 11-15 Jan 1998


Roger Edwards

Storm Prediction Center

Norman, OK


Forecasting tornadoes associated with tropical cyclones (TCs) has been a concern for National Weather Service (NWS) Severe Local Storms (SELS) unit forecasters at least since SELS' inception (Malkin and Galway, 1953). Several major tornado events have occurred in TCs, including an outbreak from Hurricane Carla (Sadowski 1962) with a violent F4 tornado (Grazulis 1993), 115 tornadoes from Hurricane Beulah in the second largest tornado outbreak by numbers in U.S. history (Grazulis 1993), a deadly regional outbreak of long-lived supercells and multiple-vortex tornadoes with the remains of Hurricane Danny (McCaul, 1987), and 37 tornadoes -- three with F3 damage -- from the remnants of Tropical Storm Beryl (Vescio et al., 1995).

As part of its mission of providing national severe and hazardous weather forecast guidance, the Storm Prediction Center (SPC, formerly SELS) is responsible for predicting certain events associated with TCs before, during and after landfall. Though severe thunderstorms with damaging downdrafts may occur almost anywhere within a TC, the damaging gust threat around landfall is generally intended to fall under the domain of tropical storm or hurricane watches and warnings. SPC products may deal with the peripheral damaging wind hazard, and severe gusts from core convection well inland while a cyclone remains tropical in character, e.g., Hurricane Andrew for several days after landfall (Weiss and Ostby 1993). Severe hail -- 19 mm or larger by NWS definition -- in tropical cyclones is rare but has occurred in clusters with some systems (e.g., Tropical Storm Beryl as documented in NCDC 1994). Significant hail minimum two inches or 51 mm (Hales 1988) or hail producing non-agricultural property damage at ground level, are historically undocumented in TCs. SPC forecasts for landfalling TCs, therefore, concentrate largely on the tornado threat, except for the category of Mesoscale Convective Discussions (MCDs) dealing with heavy rainfall.

This article discusses the purpose and philosophy for each SPC severe local storm product during TC events, and the role of SPC coordination with the National Hurricane Center (NHC) section of the Tropical Prediction Center.


SPC issues two outlooks which cover the second-day general- and severe-thunderstorm threats across the conterminous U.S. (times valid beginning in Feb. 1999) -- the first at 0730 UTC (during Daylight Time) or 0830 CDT (during Standard Time) and the next at 1730 UTC. Each is valid for 24 hours beginning at 1200 UTC the next day. There are currently five day-1 outlook issuance times (Table 1).

Table 1.

SPC convective outlooks are intended to cover broader spatial and temporal domains than severe weather watches, with longer lead time. [The general-thunderstorm line is intended for a minimum 10% thunderstorm coverage, and categorically include all severe risks.] SPC outlooks for TC tornado potential are based on the track forecast in NHC advisories, and contain verbiage referring to NHC guidance for further information on the cyclone. Since these forecasts should be consistent in placement and timing, SPC outlooks are adjusted with changes in NHC forecasts of landfall timing and intensity. Outlook areas are heavily weighted toward the azimuthal northeastern quadrant of TCs and their inland remnants, identified as the most favorable region for tornado occurrence by several decades of cumulative climatological study (e.g., Hill et al. 1966; Novlan and Gray 1974; Gentry 1983). Those works also incorporate climatological reasoning with respect to temporal distribution and cyclone strength (e.g., Weiss 1987), and more recently, physical concepts based on kinematic and thermodynamic analyses (McCaul 1991). See Fig. 1 for a graphic example of an SPC outlook area in a TC situation.

Figure 1. "Slight" risk outlook (inner curve, labeled SLGT) issued at 1452 UTC, 11 September 1998, for Tropical Storm Frances. Outer curve denotes general thunderstorm forecast.

Routine Day-2 outlook issuances may fall before or early within the first hurricane watch, when there is more uncertainty in the track forecast than during the hurricane warning. The outlooks typically become more precise as landfall time approaches. Given the associated margins for error, SPC outlooks will:

  1. generally be centered to the right of the landfall track, which will cover the climatologically and physically most-favorable quadrant of the TC;
  2. often spread beyond the area of a tropical cyclone landfall most likely to contain tornadoes, especially on day-2; and
  3. almost always begin as "slight" risks, also considering the limited spatial/temporal coverages of tornadoes in most landfalling TCs.

The "moderate" risk category is normally reserved for well-organized severe weather events associated with fronts, baroclinic-wave cyclones and drylines in the midlatitudes where the threat of severe weather of any type is significant or relatively concentrated. As applied solely for tornadoes, such as in a TC, it would mean an enhanced probability of tornadoes in series or clusters, and/or "significant" tornadoes (Hales 1988 and Grazulis 1993) producing F2 or greater damage.

A "moderate" risk in a TC would probably be relatively short-fuse -- in the day-1 period -- and would discuss an exceptional combination of observed or developing intense vertical shear parameters (such as described by McCaul 1991), with strong forecaster confidence for numerous tornadoes.


As the TC approaches land, guidance from the convective outlooks and short-range numerical models (now including the GFDL hurricane model) is increasingly blended with satellite, radar and other observational data. Observations are continuously monitored to assess the tornado threat, including rawinsonde soundings, WSR-88D velocity-azimuth display wind profiles (VWP), buoy and ship reports, and land-based surface observations. These can be analyzed directly to compute shear and instability under the landfalling TC envelope, and incorporated into modified forecast soundings from short-range models such as the Eta and Rapid Update Cycle (RUC). Such real-time information could be exceptionally beneficial in the event of a rare climatological anomaly which would normally not be included in an outlook.

In the several hours prior to landfall, before the periphery of the most favorable sector of a TC circulation approaches the coast, an MCD is usually issued outlining the area of heightened concern within the broader severe weather outlook. Graphic illustrations of the threat area (e.g., Fig. 2) are sent to the SPC website ( shortly thereafter, as time permits. The MCD will mention the possibility of tornado watch issuance, if warranted.

Figure 2. SPC MCD Internet graphic for the tornado threat from Tropical Storm Frances (1998).

SPC may also issue MCDs when a distinct heavy rainfall threat of at least 25.4 mm hr-1 appears on the meso-beta scale. Such an MCD typically includes several counties, with a short-term forecast time of 1-4 hours. Heavy to excessive rainfall is a common hazard in tropical cyclones as a whole, with the threat often considered too broad for the sort of precise nowcast guidance in an SPC heavy rain MCD. One might be written, however, for an intense and slow-moving rain band associated with a TC. Other possibilities for heavy rain TC MCDs include small zones where low level convergence and instability appear suitable for a heavy rain event well inland from landfall.


In close coordination with the mesoscale forecaster, the lead forecaster makes watch decisions using much of the same analyses and guidance as MCDs -- but with a slightly more precise mesoscale threat area (e.g., Fig. 3). Watches are drawn to tightly fit the corridor of highest tornado potential, based on current and forecast TC motion, climatology and aforementioned observational guidance. Since TC tornado threats typically last longer in time than those of non-tropical cyclones, the usual 4- to 6-hour time window for watches is often expanded. In these cases, tornado watches may be valid 6 to 12 hours. During landfalling TCs, the lead and/or mesoscale forecasters coordinate watch outlines with NHC and local NWS offices as detailed below.

Figure 3. SPC tornado watch number 824 valid for the landfall of Tropical Storm Frances (1998).

After a watch becomes valid, the mesoscale forecaster issues status reports roughly every two hours. These products are intended to provide updates on the general trends and spatial shift of tornado potential during the duration of the watch, to delineate specific spiral bands or zones of convection most likely to produce tornadoes, and to provide advance notice of the possibility of additional watch issuance. When possible, status reports also delineate the areas where the threat of severe weather will continue. Local NWS offices often use this guidance to clear counties out of the watch as necessary.


SPC and NHC may commence unscheduled, informal telephone coordination on potential tropical cyclone landfalls as early as two days beforehand, when the SPC lead or outlook forecaster prepares the initial Day-2 Convective Outlook. The coordination becomes more frequent as the expected landfall hour approaches, as the NHC location forecast becomes more precise through issuance of hurricane watches and warnings, and as the mention of tornadoes in NHC advisories becomes more likely. Formal coordination typically begins 6-12 hours before the onset of the tornado threat, prior to landfall, when the outermost spiral convective bands are still offshore. During the coordination, SPC will specify the extent of the tornado threat, including timing and general location of the affected areas, and whether or not NHC should mention the possibility of tornadoes in the Tropical Storm/Hurricane advisories.

The mention of a tornado threat in NHC advisories does not depend on the issuance of a tornado watch; but it should be consistent with the latest SPC Day-1 outlook. For example, if the Day-1 outlook delineates a "slight risk" area beginning tonight over southeastern Louisiana and southern Mississippi for the possibility of TC-spawned tornadoes, SPC may request verbiage in the NHC advisories starting at 1800 UTC similar to this: "There is the possibility of isolated tornadoes over southeastern Louisiana and southern Mississippi tonight.

If the SPC lead forecaster deems a tornado watch necessary, SPC coordinates the valid period and location (end points) with NHC and affected NWS offices through the Hurricane Hotline -- the most efficient means currently available for such communication. Except for quickly evolving or unexpected situations, such as an abrupt change in TC track, the watch area will be consistent with the latest SPC day 1 outlook. [In the rare exceptions, the outlook is amended to cover at least the watch domain.] This hotline coordination can be done during regular NHC conference calls, or whenever a watch needs to be issued or replaced. Since a tornado watch implies an increased risk of tornadoes, the wording in the NHC advisory reflects this, e.g., "There is the possibility of a few tornadoes over southeast Texas and southern Louisiana tonight...and a tornado watch has been issued for this area."

Watch clearance is done only by local NWS offices, is usually based on the status reports, and often involves phone coordination with SPC. The Hurricane Hotline may be used for this and for consultation between SPC, NHC and field offices on cancellation of a tornado watch. When the TC has either weakened sufficiently, or moved back offshore, and the tornado threat has diminished, the SPC lead forecaster may either cancel any remaining watches, or allow them to expire as scheduled. If NHC is still writing advisories, they may then remove mention of a tornado threat. If responsibility for tracking the system has been turned over to the Hydrometeorological Prediction Center (HPC), SPC discusses the severe weather threat with them. Then, as with other non-tropical systems, further severe weather watch coordination involves SPC and affected local NWS offices.


The author thanks Dave Imy, Bob Johns, Joe Schaefer and Mike Vescio of SPC, as well as Jack Beven, Max Mayfield and Richard Pasch of NHC, for their guidance and helpful discussions in compiling this article.


Gentry, R.C., 1983: Genesis of tornadoes associated with hurricanes. Mon Wea. Rev., 111, 1793-1805.

Grazulis, T.P., 1993: Significant Tornadoes: 1680-1991. Environmental Films, St Johnsbury VT, 1326 pp.

Hales, J.E., 1988: Improving the watch/warning system through use of significant event data. Preprints, 15th Conf. Severe Local Storms, Amer. Meteor. Soc., Baltimore, 165-168.

Hill, E.L., W. Malkin and W.A. Schulz, Jr., 1966: Tornadoes associated with cyclones of tropical origin -- practical features. J. Appl. Meteor., 5, 745-763.

Malkin, W. and J. G. Galway, 1953: Tornadoes associated with hurricanes: As illustrated by the Franconis, Va., tornado, September 1, 1952. Mon Wea. Rev., 81, 299-303.

McCaul, E.W., Jr., 1987: Observations of the Hurricane "Danny" tornado outbreak of 16 August 1985. Mon. Wea. Rev., 115, 1206-1223.

_____, 1991: Buoyancy and shear characteristics of hurricane-tornado environments. Mon Wea. Rev., 119, 1954-1978.

NCDC, 1994: Storm Data, 36, 8, 155 pp.

Novlan, D.J., and W.M. Gray, 1974: Hurricane-spawned tornadoes. Mon Wea. Rev., 102, 476-488.

Sadowski, A., 1962: Tornadoes associated with Hurricane Carla, 1961. Mon Wea. Rev., 90, 514-516.

Vescio, M.D., S.J. Weiss, and F.P. Ostby, 1995: Tornadoes associated with Tropical Storm Beryl. Preprints, 21st Conf. Hurricanes and Tropical Meteor., Amer. Meteor. Soc., Miami, 469-471.

Weiss S.J., 1987: Some climatological aspects of forecasting tornadoes associated with tropical cyclones. Preprints, 17th Conf. Hurricanes and Tropical Meteor., Amer. Meteor. Soc., Miami, 160-163.

Weiss, S.J., and F.P. Ostby, 1993: Synoptic and mesoscale environment associated with severe local storms produced by Hurricane Andrew. Preprints, 17th Conf. Severe Local Storms, Amer. Meteor. Soc., St. Louis, 267-271.