Higher outflow channel dissection in the martian region of southern circum-Chryse appears to have extended from the Late Hesperian to the Middle Amazonian Epoch. These outflow channels were excavated within the upper 1. km of the cryolithosphere, where no liquid water is expected to have existed during these geologic epochs. In accordance with previous work, our examination of outflow channel floor morphologies suggests the upper crust excavated by the studied outflow channels consisted of a thin (a few tens of meters) layer of dry geologic materials overlying an indurated zone that extends to the bases of the investigated outflow channels (1. km in depth). We find that the floors of these outflow channels contain widespread secondary chaotic terrains (i.e., chaotic terrains produced by the destruction of channel-floor materials). These chaotic terrains occur within the full range of outflow channel dissection and tend to form clusters. Our examination of the geology of these chaotic terrains suggests that their formation did not result in the generation of floods. Nevertheless, despite their much smaller dimensions, these chaotic terrains are comprised of the same basic morphologic elements (e.g., mesas, knobs, and smooth deposits within scarp-bound depressions) as those located in the initiation zones of the outflow channels, which suggests that their formation must have involved the release of ground volatiles. We propose that these chaotic terrains developed not catastrophically but gradually and during multiple episodes of nested surface collapse. In order to explain the formation of secondary chaotic terrains within zones of outflow channel dissection, we propose that the regional Martian cryolithosphere contained widespread lenses of volatiles in liquid form. In this model, channel floor collapse and secondary chaotic terrain formation would have taken place as a consequence of instabilities arising during their exhumation by outflow channel dissection. Within relatively warm upper crustal materials in volcanic settings, or within highly saline crustal materials where cryopegs developed, lenses of volatiles in liquid form within the cryolithosphere could have formed, and/or remained stable.In addition, our numerical simulations suggest that low thermal conductivity, dry fine-grained porous geologic materials just a few tens of meters in thickness (e.g., dunes, sand sheets, some types of regolith materials), could have produced high thermal anomalies resulting in subsurface melting. The existence of a global layer of dry geologic materials overlying the cryolithosphere would suggest that widespread lenses of fluids existed (and may still exist) at shallow depths wherever these materials are fine-grained and porous. The surface ages of the investigated outflow channels and chaotic terrains span a full 500 to 700. Myr. Chaotic terrains similar in dimensions and morphology to secondary chaotic terrains are not observed conspicuously throughout the surface of Mars, suggesting that intra-cryolithospheric fluid lenses may form relatively stable systems. The existence of widespread groundwater lenses at shallow depths of burial has tremendous implications for exobiological studies and future human exploration. We find that the clear geomorphologic anomaly that the chaotic terrains and outflow channels of southern Chryse form within the Martian landscape could have been a consequence of large-scale resurfacing resulting from anomalously extensive subsurface melt in this region of the planet produced by high concentrations of salts within the regional upper crust. Crater count statistics reveal that secondary chaotic terrains and the outflow channels within which they occur have overlapping ages, suggesting that the instabilities leading to their formation rapidly dissipated, perhaps as the thickness of the cryolithosphere was reset following the disruption of the upper crustal thermal structure produced during outflow channel ex