干旱区生态环境知识资源中心

The Yellow River basin could be divided into three sub-regions, which makes it as the ideal target for studying regional climate change. On the basis of daily precipitation at 62 meteorological stations in the Yellow River basin, spatial distribution and temporal trends of annual and monthly mean precipitation and extremes were analyzed during 1961-2012. The Mann-Kendall trend test and linear least-square method were used to detect trends and magnitudes in annual and monthly mean precipitation and extremes. The results indicate that mean precipitation and extremes have different trends, and the three sub-regions also have distinct trends. Annual average precipitation shows a slight decrease in the whole basin with a trend of -8.8 mm/decade, a significant decrease in the eastern monsoon sub-region with trends of -14.4 mm/decade but increases in the high-elevation sub-region with trends of 1.3 mm/decade. Monthly precipitation in the Yellow River basin shows a different seasonality, December and June have largest positive trends, while November and October have largest negative trends. The change degree of annual precipitation extremes in the whole Yellow River basin decreased, reflected by seven indices (CWD, SDII, R10, R95p, R99p, Rx1day, and Rx5day) having negative trends but significantly different in the three sub-regions. Specifically, trends of all the ten annual precipitation extremes indices in the eastern monsoon sub-region were dominant negative, while mainly positive in the arid and semi-arid sub-region and high-elevation sub-region. The four monthly precipitation indices (PRCPTOT, SDII, Rx1day, and Rx5day) have main positive trends in February, May, June, and December, while negative trends in April, August, September, October, and November, in which the months having the most dominant positive trends are distinctly different (in February or June or December), while months with the most dominant negative trends are the same (in November). In the whole basin, eight indices (PRCPTOT, SDII, R10, R20, R95p, R99p, Rx1day, and Rx5day) have positive relations with elevation, while two indices (CDD and CWD) have negative relationship with elevation, but in the three sub-regions, relations between the ten indices and elevation are not significant. Relationship between extremes indices and large-scale atmospheric circulations show that, in the whole basin, all the ten annual indices have little relationship with Northern Hemisphere Subtropical High (NHSH) and Northern Hemisphere Polar Vortex (NHPV). But for the four monthly precipitation indices (i.e., Rx1day, Rx5day, PRCPTOT, and SDII), there were significant positive relationships with NHSH but significant negative relationships with NHPV. The results of this study are useful to master change rule of local mean precipitation and extremes change, which will help to prevent natural hazards caused by precipitation extremes.