One of the greatest challenges for an operational rainfall enhancement program is to identify those days for which seeding efficacy might be optimal. At first glance the problem seems to be the challenge of making an accurate convective forecast. However differing environmental conditions along with differing convective modes complicate the forecast problem. The question of optimal seeding with respect to convective mode is beyond the scope of this paper. However, ambient conditions in environments that sustained deep convection versus those that did not will be looked at. Using an approach taken by previous investigators, thunderstorm indices are calculated for two separate datasets from the 2003 operational year in the SOAR target area (April to September). Level III reflectivity data and data from a 5 cm radar in Plains, Texas are used to discriminate thunderstorm from non-thunderstorm days within a 70 nm range ring from the SOAR station (located in Plains, Texas). Conventional thunderstorm indices, such as the lifted index and the showalter index, will be looked at for significant predictability signals. In addition, seeding related parameters such as the first echo depth and the index of coalescence activity will be examined to deduce if these parameters have any correlation with conventional sounding parameters and their ability to distinguish different seeding regimes.