Upland glaciation in tropical pangaea: Geologic evidence and implications for late paleozoic climate modeling

Gerilyn S. Soreghan, Dustin E. Sweet, Nicholas G. Heavens

Research output: Contribution to journalArticlepeer-review

30 Scopus citations


The late Paleozoic archives a prolonged icehouse, long recognized by means of Gondwanan continental glaciation. In contrast, the paleotropics have long been considered warm. Here we present the hypothesis of upland glaciation in the Ancestral Rocky Mountains (ARM) of western equatorial Pangaea, a region located within 11° of the paleoequator. The data to support this hypothesis include (a) a Permo-Pennsylvanian valley with glacial attributes and diamictite exhibiting rare striated clasts; (b) coarse-grained lacustrine strata onlapping the valley and preserving lonestones in Gilbert-type deltaic deposits proximally, along with (c) coarse-grained fluvial siliciclastic strata with microstriae and evidence for widespread flood deposition; (d) polygonally cracked paleosurfaces inferred to reflect frozen ground; and (e) voluminous paleoloess. Tropical glaciation occurs today at altitudes >4500 m and descended to 2100-3000 m at the last glacial maximum (LGM). However, ARM depositional systems terminating at sea level and emanating from inferred ice-contact facies indicate that ice-terminus elevations were lower (<1200-1600 m) than those of the LGM. If valid, tropical temperatures were ~15°C cooler than today during intervals recording hypothesized tropical glacial conditions. This implies at least episodic cold within western tropical Pangaea, which conflicts with inferences from oxygen isotope paleothermometry. Furthermore, climate models for the late Paleozoic cannot reproduce tropical upland glaciation except under prohibitively low Pco2, implying the need to consider other forcings, such as cloud and aerosol behavior. Upland glaciation in the Permo-Pennsylvanian tropics was potentially widespread, given the global orogenesis accompanying Pangaean assembly. However, testing this hypothesis requires identification of pro- and periglacial indicators, owing to widespread erosion of upland (glaciated) regions. Midlatitude glaciation in both hemispheres also was likely, challenging climate models and paleogeographic consensus for this period.

Original languageEnglish
Pages (from-to)137-163
Number of pages27
JournalJournal of Geology
Issue number2
StatePublished - Mar 2014


Dive into the research topics of 'Upland glaciation in tropical pangaea: Geologic evidence and implications for late paleozoic climate modeling'. Together they form a unique fingerprint.

Cite this