A 1,000-Year History of Typhoon Landfalls in Guangdong, Southern China, Reconstructed from Chinese Historical Documentary Records



In China, the abundance of historical documentary records in the form of Fang Zhi (semiofficial local gazettes) offers an extraordinary opportunity for providing a high-resolution historical dataset for the frequency of typhoon strikes. We have reconstructed a 1,000-year time series of typhoon landfalls in the Guangdong Province of southern China since AD 975 based on data compiled from Fang Zhi. Even though the 571 typhoon strikes recorded in the historical documents probably underrepresent the total number of typhoon landfalls in Guangdong, calibration of the historical data against the observations during the instrumental period 1884–1909 suggests that the trends of the two datasets are significantly correlated (r= 0.71), confirming that the time series reconstructed from historical documentary evidence contains a reliable record of variability in typhoon landfalls. On a decadal timescale, the twenty-year interval from AD 1660 to 1680 is the most active period on record, with twenty-eight to thirty-seven typhoon landfalls per decade. The variability in typhoon landfalls in Guangdong mimics that observed in other paleoclimatic proxies (e.g., tree rings, ice cores) from China and the northern hemisphere. Remarkably, the two periods of most frequent typhoon strikes in Guangdong (AD 1660–1680, 1850–1880) coincide with two of the coldest and driest periods in northern and central China during the Little Ice Age. Conceivably, the predominant storm tracks shifted to the south during these cold periods, resulting in fewer landfalls in Japan and the east-central Chinese coast but more typhoons hitting Guangdong. Spectral analysis of the Guangdong time series reveals an approximately fifty-year cycle in typhoon landfall frequency. While the physical mechanism remains to be identified, it is tempting to relate this periodicity to the pentadecadal variability identified in the North Pacific Index (NPI) time series.