Previous simulations of derecho-producing squall lines (Schmidt, 1991; Bernardet and Cotton, 1998) suggested that air contributing to strong surface winds originates in front-to-rear storm relative flow which is then lifted upon encountering an upward-directed pressure gradient force (PGF), associated with an elevated induced low formed by a rotating updraft or a mesocyclone. Once the stable air leaves the region of uplift, it then descends rapidly back towards the surface with melting and evaporation of precipitation enhancing its negative buoyancy. It is hypothesized that this front-to-rear, up-down downdraft (Knupp, 1987) is a major driving force of derechos rather than the descent of the rear-to-front downdraft (RFD), which is normally thought to be the main source of severe winds.
In this study, ELDORA dual-Doppler data from during BAMEX are used to examine if those observations support or refute the hypothesis that 1). Front-to-rear up-down downdrafts (Knupp 1987) do exist in observations of bow-echoes and 2). they contribute to derecho events. Data from three case studies from the seventeen cases in which the NRL P-3 participated during BAMEX were chosen from among several derecho-like, bow-echo MCS cases that produced extended tracks of severe surface winds, and which were well observed by the NRL P-3 aircraft's ELDORA Doppler radar. Two different dual-Doppler analyses programs were used to analyze the ELDORA Doppler data. CEDRIC, developed by NCAR, was used to analyze the 5 July case, with data provided by Hanne Murphy and Dr. Roger Walkimoto (UCLA). The other two case studies, the 24 and 10 June 2003, were analyzed using a program developed by Dr. John Gamache (NOAA, Hurricane Research Division). Using these data, over 100 back-trajectory analyses were calculated for each of the three cases. Parcels were released at the three lowest levels of each analysis (.5 km, 1 km and 1.5 km AGL), at numerous locations associated with both well-defined low-level downdraft and strong horizontal wind speed at the lowest level, i.e., locations in the Doppler analyses which are likely linked to severe surface winds. It was found that one-half to three-quarters of all parcels released at such locations originated in front-to-rear "up-down" downdraft trajectories. Whereas strong surface winds and low-level downdrafts are often generally attributed to descending rear-to-front inflow, we conclude from these analyses that front-to-rear "up-down" downdrafts are likely a major contributor to severe surface winds. Ongoing analysis involving surface wind reports, damage surveys, and theta-e analysis of mesonet data and proximity soundings should further establish the link between the front-to-rear "up-down" downdrafts and severe surface winds.