Comparison of lake records for climate reconstructions: A case study from Hala Lake, northern Tibetan Plateau, China

Sediment cores from a lake are often interpreted in light of modern limnological and environmental conditions to infer past climate and hydrological conditions in a region. Records from the Tibetan Plateau and adjacent areas display a heterogeneous picture through space and time. The saline Hala Lake, located in the Qilian Mountains, Qinghai Province, China, at 4078 m a.s.l. was selected to prove the applicability of a selected single sediment core for a consistent inference of past hydrological and climate conditions. Based on nine sediment cores obtained from different locations and water depth, sedimentation patterns and depositional conditions within the lake were investigated in detail. Two long cores H7 and H8 from the center of the lake (65 m water depth) and from the western nearshore location (20 m water depth) were compared by means of sedimentary composition, minerals and geochemical data (X-ray diffraction, X-ray fluorescence, loss-on-ignition, and by CNS analyzer). The respective age model was based on seventeen AMS radiocarbon dates, indicating a negligible reservoir error for sediments from the lake center and approximately 1,000 years error for the near-shoreline sediments. Six cores from the littoral zone revealed a sedimentary succession from sand and silty clay to laminated clay at the southern side of the lake. Undisturbed finely laminated sediments were found from 15 m water depth down to the deepest part. Core H5 (2.5 m length) from 31 m water depth yielded abundant green algal mats mixed with clayey lake deposits. Algae occurred between 25 and 32 m water depth and influenced the dissolved oxygen content of the stratified lake. The comparison of cores H7 and H8 yielded prominent mismatches for different time periods, which may, in part, attributed to lake internal processes independent of climatic influence. We thus conclude that data from a single sediment core may lead to different climate inferences. Common shifts among proxy data, however, showed that major climate shifts, of regional to global significance, can be tracked and allow reconstruction of lake level changes over the last 24,000 years. The results indicate the advance of glaciers into the lake basin during the global LGM. The lake experienced lowest levels of 25-50 m below the present stage. A stepwise refilling started already at ca. 16 kyr BP and reached the -25 m level during the Bølling/Allerød warm phase at ca. 13.5 kyr BP. A desiccation episode falls within the Younger Dryas, followed by a substantial rise during the first millennium of the Holocene as a result of climate warming which enforced glacier melt. Disturbed sediments in cores H7 and H12 indicate a single mass flow event that was most likely triggered by a major seismic event at around 8.5 kyr BP, independent from climate. The occurrence of authigenic dolomite within the mass flow layer is unique and may be attributed to local formation in the littoral zones.