Fig. 1: Study region and cross-validation results at five gauging stations. | Nature Communications

Fig. 1: Study region and cross-validation results at five gauging stations.

From: Reconstructed eight-century streamflow in the Tibetan Plateau reveals contrasting regional variability and strong nonstationarity

Fig. 1

a The red line shows the domain of the Tibetan Plateau. Red triangles denote gauging stations in headwater regions, i.e., from north to south, Tangnaihai (TNH) on the Yellow River, Zhimenda (ZMD) on the Yangtze River, Changdu (CD) on the Lancang-Mekong River, Jiayuqiao (JYQ) on the Nu-Salween River, and Nuxia (NX) on the Yarlung Zangbo-Brahmaputra River. Grey + symbols denote locations of Palmer Drought Severity Index (PDSI) predictors used in this study. Green trees denote open-sourced tree-ring chronologies located in or surrounding the South and East Tibetan Plateau (SETP). Shaded areas represent significantly correlated areas at a 0.05 significance level between reconstructed streamflow at the ZMD gauging station and Climate Research Unit (CRU) precipitation for each grid cell. b Medians of cross-validation results for 100 samples at each gauge, a perfect model would have all these five indices equal 1.0. c Radar plots showing means of five indices for all gauging stations. d The black box (a) denotes the area used to average precipitation to perform linear regression between annual streamflow at the ZMD gauge and spatially averaged annual CRU precipitation. Spatial correlation maps and linear regressions between CRU precipitation and annual streamflow for the other four gauges (TNH, CD, JYQ and NX) are shown in Supplementary Fig. 6.

Back to article page