Search

The recently proposed global and local calibrations for the mean annual air temperature (MAT) reconstruction on the basis of 5- and 6-methyl brGDGTs have rarely been applied to the Chinese LPS yet, leaving the applicability of these calibrations unclear. Here, we used the improved chromatography method to analyze 198 loess-paleosol samples from the Weinan section in the southern CLP for the past 350 kyr. The 6-methyl brGDGTs comprise a major proportion of total brGDGTs, pointing to alkaline conditions for most soil samples from the Weinan LPS. The decoupled profile variation of MBT′ and MBT′ suggests their response to different enviromental factors, possibly soil pH (or soil moisture) and temperature, respectively. This discrimination further corraborates that temperature and monsoonal precipitation were not in phase during the last four deglaciations on the CLP. Temperature estimates for the six calibrations tested show similar trends but remarkably differ in amplitudes. The soil moisture appears to affect the global MBT′/CBT calibration, the global MAT calibration (a calibration based on the multiple linear regression) and the Chinese local SSM (Stepwise Selection Method) calibration, resulting in a significant underestimation of late Holocene temperature. In contrast, the dry climate has no effect on the global MBT′ calibration. Of the six calibrations, only the Chinese local SSM calibration and global MAT calibration produce temperature variation amplitude over the past 350 ka that is consistent with other independent proxy data. The recently proposed local calibrations based on 5- and 6-methyl brGDGTs for North China yield the amplitude of temperature changes that is much larger than the results determined by other approaches. All the six calibrations have their own weakness in the MAT reconstruction, due partly to the inconsistency between the brGDGT distribution in the Weinan LPS and the modern soils used to establish these calibrations. [ABSTRACT FROM AUTHOR]