| description abstract | Vegetation around the Daihai Lake, northern China, is very sensitive to climate changes. In this paper, pollen-based quantitative climate reconstructions using three methods [weighted averaging partial least squares method (WAPLS), modern analog technique (MAT), and pollen response surface method (PRS)] were conducted to obtain robust reconstructions of Holocene climate changes in the Daihai Lake area. The result obtained by the three methods all consistently show the annual precipitation to have been 50?100 mm lower in the early Holocene, 100?200 mm higher in the Mid-Holocene, and 50?100 mm lower again in the late Holocene than at present. The WAPLS and the MAT methods also show quasi-synchronous oscillations of the mean annual temperature (Ta); 1°?2°C lower in the Early Holocene and 1°?3°C higher in the Mid-Holocene than today. The time period from 6200 to 5100 cal yr BP was the wettest and the warmest interval, with an annual precipitation (Pa) greater than 550 mm and mean annual temperature Ta higher than 6.5°C. Several cold and dry events can be identified to occur about 8200, 6000, and 4400 cal yr BP, with an annual precipitation less than 400 mm and a mean annual temperature colder than 4.5°C, respectively. The mean temperature of the warmest month (Tw) as reconstructed using both WAPLS and MAT methods was relatively stable during the Holocene, fluctuating about ±2°C relative to the present level, but the PRS method suggests more varied Tw values in both amplitude and frequency. After 1500 cal yr BP, no consistent pattern can be observed from these three different analyses, probably because of the impact of intensified human disturbances on the natural vegetation. The fluctuations of annual precipitation (Pa) correspond to that observed in Dongge Cave in southern China. The differences might be linked to Indian monsoon and East Asia monsoon climates or caused by the different degree of dating precision, different temporal resolution, and different sensitive response of climate proxies to the climate variations. | |
