Recent observations in the Green River Formation suggest that ancient “Lake Gosiute” was a playa-lake complex (Eugster and Surdam, 1973). In this paper, the new playa-lake model is tested in a basin-wide study of surface and subsurface observations. The rocks deposited in and around “Lake Gosiute” can be divided into three distinct facies: (1) marginal silt and sand, (2) carbonate mud flat, and (3) lacustrine. Each lithologic facies has a characteristic carbonate mineral assemblage. The marginal facies is characterized by calcite concretions and calcareous cements. The mud-flat facies is characterized by calcite and (or) dolomite. The lacustrine facies is characterized either by trona (sodium carbonate) or by oil shale (either calcitic or dolomitic). The regional distribution pattern of lithologic facies and mineral zones in the Green River Formation of Wyoming is identical with that of modern playa-lake complexes. Moreover, in the Tipton Shale Member, once supposed to have been deposited in a large, deep, open, fresh-water lake (Bradley, 1963), there is strong evidence demonstrating large fluctuations in the position of the shoreline and progressive increases in salinity and alkalinity of the lake water. By mapping the regional distribution and types of lateral changes characterizing individual stromatolite units, the fluctuations in shoreline position can be quantified. The vertical distribution of fossils and ooliths in the Green River Formation allows an evaluation of water chemistry. In addition, the assemblage of sedimentary structures in the Tipton Shale Member is compatible only with a sedimentologic model characterized by shallow-water deposition and frequent subaerial exposure.
Thus, the deep-water stratified-lake model is untenable not only for the Wilkins Peak Member but also for the Tipton Shale Member of the Green River Formation. In contrast, the playa-lake model is consistent with the physical, chemical, and paleontologic aspects of the Green River Formation of Wyoming.