Reconstruction of the Northern Hemisphere temperature from 1500 to 1949 by optimal regional averaging method

The rareness and inhomogeneity of the data points cause difficulties in the reconstruction of past average temperature. Optimal regional averaging is a method that can overcome these difficulties and obtain the average temperature of target area by means of optimal weights using limited temperature data. In this paper, the average temperature in the Northern Hemisphere is calculated by the optimal regional averaging method using two types of data: temperature data from Climatic Research Unit from 1901 to 2000 and maximum latewood density dataset of tree from 1500 to 1949. Five, ten, fifteen data points from CRU and forty data points from MXD are used in our research. The results show that even with the relatively less data used in this reconstruction, the method allows the reconstruction of the average temperature of the Northern Hemisphere more accurately, which provides the temperature information for palaeoclimate reconstruction.