High resolution approach to reconstruct Late Pleistocene and Holocene climate of the northern Arabian Sea, with special emphasis on the comparison of biomarker-based sea surface temperature proxies

Abstract

Multiple organic proxies have the potential to reconstruct sea surface temperature (SST), but their behavior is not completely understood within upwelling areas like the Arabian Sea. Here, we tested the applicability of three organic SST proxies: the UK’37 index based on long chain alkenones, the TetraEther indeX (TEXH86) and the Long chain Diol Index (LDI), supported by measures for surface productivity, bottom water oxygenation and upwelling intensity. Sediments from the northern Arabian Sea spanning 76 – 3 ka BP reveal that the three SST records differ in absolute reconstructed SST and furthermore show diverse SST variations. TEXH86-SSTs are highest (28.2 – 31.4 °C) and the record is characterized by a gradual warming between 76 and 3 ka BP, whereas LDI-SSTs are relatively stable around 26 °C with decreases in SST during the periods 22 - 3 ka BP, 45 – 37 ka BP and 57 – 51 ka BP. The UK’37-SST record shows a large SST range (17.5 to 26.8 °C). None of the SST proxies follow the stable oxygen isotopic composition (δ18O) of the Northern Hemisphere (NH) GRIP record. Instead, the TEXH86 shows a Southern Hemisphere (SH) influence as suggested by synchronous variation with the stable hydrogen isotopic composition (δD) of the EPICA Dome C record. The SH influence within the TEXH86 between 29 and 12 ka BP is probably related to increased influence of SH-sourced intermediate waters during periods of low upwelling intensity and sea-level lowstand. The gradual increase of TEXH86-SST from 76 to 3 ka BP might be related to a shift within the production pattern of isoprenoid glycerol dialkyl glycerol tetraethers (GDGTs) or could reflect a gradual warming towards the Holocene. LDI-SST is particularly forced by monsoon dynamics and therefore does not reflect annual mean SST. Changes in southwest monsoon (SWM) intensity have a dampening effect on LDI-SST differences between climatic states by inducing a shift in the blooming season of the biomarker producers. Furthermore, the lower temperature of upwelling waters can be recorded in the LDI during periods characteristic of an intense SWM. During periods of low SWM intensity, preferential degradation might cause an offset towards lower LDI-SST. The UK’37 record does not show the variability observed either within reference ice-core records or within the LDI and TEXH86 records. This differing behavior may be due to lateral transport of re-suspended alkenones. The results indicate that application of organic SST proxies within upwelling-influenced areas is complex and should be done with caution, since many processes exert an influence on reconstructed temperatures. Clearly, a multi-proxy approach combining SST proxies and information on upwelling-related parameters is required to understand SST dynamics within these regions.