An approach to reconstruction of temperature on a seasonal basis using historical documents from China

A methodology is presented for reconstructing past climates on a seasonal basis using historical data from chronicles and documents in China. This procedure consists of three steps: first, individual climate indicators are identified and scaled by ordinal numbers (or grades) based on their individual statistics; secondly, the cold (warm) climate index is defined as a function of the scaled indicators. Finally, the climate index is calibrated in terms of temperatures by regression with recent observations. This methodology is applied to constructing the temperature series of the last 500 years in China, based on which climate variations are studied. Spectrum analysis shows variability in about 200, 80, 30, and 22–year cycles.