Tornadoes in a Deceptively Small CAPE Environment:  The "Surprise" 4/20/04 Outbreak in Illinois and Indiana On the late afternoon and early evening of 20 April 2004, an outbreak of tornadoes occurred across northern Illinois and central Indiana, including an F3 tornado that killed eight people.  Over a four-hour period, 31 tornadoes developed from discrete ‘low-topped’ supercells that interacted with a rapidly advancing warm front characterized by a weak thermal gradient.  Several of the supercells produced more than five tornadoes each, some of which could be classified as large and long tracked (e.g., 0.8 km wide with path lengths > 24 km).  The event was largely unanticipated and one of the more difficult outbreaks in recent history to forecast, based on short-range numerical weather prediction data less than six hours prior to thunderstorm initiation.  However, in retrospect, subtle clues were found to be present for forecasters who monitored trends and evolution of observed data in combination with model data.  This case also highlights some important issues regarding tornado events that occur with relatively small CAPE.

ETA and RUC model forecasts from the morning of 20 April 2004 did a poor job of handling the position of the warm front by mid and late afternoon, and, as a result, instability along and near the front was under-forecast and poorly represented.  In particular, the lowest 100 mb mean layer convective available potential energy (CAPE) was forecast to be < 250 J kg^-1 for the region and time period of interest.  This small and misleading CAPE appeared to result from a fictitious, shallow dry layer above the surface that was not representative of observed low-level, near-saturated conditions that existed along and close to the warm front.  Although surface-based computations often overestimate instability by not accounting for low-level mixing of lifted parcels, a shallower computation was more appropriate in this case to eliminate some of the "false" mixing that possibly contaminated the deeper mixed layer.  In lifting the lowest 50 mb mixed parcel derived from RUC soundings valid at the time of convective initiation near and along the warm front, CAPE was on the order of 1500 J kg^-1.  Nearly all CAPE was located below 500 mb, with the maximum buoyancy centered below 600 mb, an unusual situation compared to more “typical” supercell tornado environments from a recent RUC sounding database study where maximum buoyancy was usually centered near or above 400 mb.  Furthermore, the distribution of CAPE was co-located within a favorable vertical wind profile and hodograph which may have increased tornado potential; observed 0-1 and 0-3 km storm-relative helicity was >300 and 450 m² s^-2, respectively.  Another factor favorable for tornadoes may have been storm motion that was largely coincident with movement of the advancing warm front, keeping the supercells in an environment supportive of tornadoes for an extended period of time.

Ideas and suggestions will be offered in this paper to help forecasters detect and become aware of deceptive "small CAPE" environments in short-range, model-derived products, with an emphasis on the increased potential for tornadoes in these settings.