The relative abundance of individual archaeal membrane lipids, namely of glycerol dialkyl glycerol tetraethers (GDGTs) with different numbers of cyclopentane rings, varies with temperature, which enables their use as a paleotemperature proxy index. The first GDGT-based index in marine sediments called TEX86 is believed to reflect mean annual sea surface temperature (maSST). The TEXL 86 is an alternative temperature proxy for lowerature regions (< 15 °C), where the original TEX86 proxy calibration shows a larger scatter. However, TEXL 86-derived temperatures still display anomalous estimates in polar regions. In order to elucidate the potential cause of the disagreement between the TEXL 86 estimate and SST, we analyzed GDGT fluxes and TEXL 86-derived temperatures in sinking particles collected with time-series sediment traps in high-northern- and highsouthern-latitude regions. At 1296 m depth in the eastern Fram Strait (79° N), a combination of various transporting mechanisms for GDGTs might result in seasonally different sinking velocities for particles carrying these lipids, resulting in strong variability in the TEXL 86 signal. The similarity of flux-weighted TEXL 86 temperatures from sinking particles and surface sediments implies an export of GDGTs without alteration in the Fram Strait. The estimated temperatures correspond to temperatures in water depths of 30-80 m, where nitrification might occur, indicating the favorable depth habitat of Thaumarchaeota. In the Antarctic Polar Front of the Atlantic sector (50° S), TEXL 86-derived temperatures displayed warm and cold biases compared to satellite-derived SSTs at 614 m depth, and its flux-weighted mean signal differs from the deep signal at 3196 m. TEXL 86-derived temperatures at 3196 m depth and the surface sediment showed up to 7 °C warmer temperatures relative to satellite-derived SST. Such a warm anomaly might be caused by GDGT contributions from Euryarchaeota, which are known to dominate archaeal communities in the circumpolar deep water of the Antarctic Polar Front. The other reason might be that a linear calibration is not appropriate for this frontal region. Of the newly suggested SST proxies based on hydroxylated GDGTs (OHGDGTs), only those with OH-GDGT-0 and crenarchaeol or the ring index (RI) of OH-GDGTs yield realistic temperature estimates in our study regions, suggesting that OH-GDGTs could be applied as a potential temperature proxy in highlatitude oceans.