Spatiotemporal climate change over the Pearl River basin inferred from observational and tree-ring data

Abstract

This study investigated the spatiotemporal climate change over the Pearl River basin based on observational and tree-ring data. Spatiotemporal climate changes over the Pearl River basin on the annual and seasonal scales were analyzed using the empirical orthogonal function (EOF) analysis, based on the gridded Climatic Research Unit time-series (CRU-TS) data during 1950–2016. Results show that both annual and seasonal temperatures have increased over the basin for the EOF1 mode. Annual total precipitation has increased over the basin for the EOF1 mode, while seasonal precipitation shows different trends, with increasing trend in summer and winter and decreasing trend in spring and autumn. Both annual and seasonal scPDSI shows drying trends over the basin for the EOF1 mode. To further investigate the summer and winter climate dynamics, relationships of the Pacific and Indian Ocean SSTs and climate indices such as ENSO, PDO and NAO, with the summer and winter temperature, precipitation and scPDSI variability in the Pearl River basin were analyzed. Results show that ENSO plays an important role in summer and winter climate changes over the basin, except in summer precipitation. Other internal climate forcings also have certain influences on summer and winter climate, in particular the PDO on summer precipitation variability. Based on the Pearson’s correlation analysis, growth–climate relationships of Guangdong pine (Pinus kwangtungensis), Bristlecone hemlock (Tsuga longibracteata), and Fujian Cypress (Fokienia hodginsii) were investigated in the Pearl River basin. Results suggest that drought has opposite influences on tree growth of Guangdong pine and Fujian Cypress. As for Guangdong pine, tree growth had negative correlation with previous July precipitation and positive correlations with previous summer temperature, suggesting that relatively warm-dry climate of the previous growing season can benefit tree growth in the following year. As for Fujian cypress, temperature in August had significant negative influence on tree growth, and precipitation in previous August and September and current October had significant positive influence on tree growth. These negative temperature and positive precipitation relations reflect typical drought responses, suggesting that water availability during warm season is crucial to tree growth. As for Bristlecone hemlock, trees at MRS site are most sensitive to climate, especially to the previous spring temperature. With tree-ring width chronologies of Guangdong pine, Fujian cypress, and Bristlecone hemlock in the Pearl River basin, annual and seasonal scPDSI and the spring temperature reconstructions were developed based on the significant growth-climate relationships. Multi-decadal variations can be detected in the scPDSI and temperature reconstructions. Two severe drought periods were found in the 1900s (1901, 1908, and 1909) and 1760s (1764, 1765, 1766, 1767, and 1769) based on the scPDSI reconstructions. Two extremely warm (2009 and 2008) and six extremely cold (1902, 1982, 1980, 1986, 1983, and 1981) years were found for spring temperature reconstruction. Significant correlations with the SSTs in western Pacific Ocean, Indian Ocean, and equatorial Atlantic Ocean suggest that the SST anomalies may influence regional scPDSI or temperature variability in the Pearl River basin.