| Fenhe river is the second largest tributary of the Yellow river and the biggest river of Shanxi Province. Fenhe river basin is located in semi-arid and semi-humid region. The precipitation is a main replenishment source for surface water in the basin, and its spatial variation charicateristics affect the distribution of surface water resources, and thus have a great impact on the ecological environment and the socioeconomic development in this basin. A dataset of monthly precipitation series (1971-2011) from 154 precipitation stations was used to analyze the spatiotemporal distribution features of annual precipitation, the precipitation in spring (from March to May), summer (from June to August), autumn (from September to November), winter (from December to February of the following year) and the annual flood season (from June to September) rainfall, respectively, over the Fenhe river basin (FHRB), the upper reaches (TUR), the middle reaches (TMR) and the lower reaches (TLR) during the period 1971～2011. The nonparametric Mann-Kendall (MK) statistical method was used to test the trend change. The MK mutation test, cumulative anomaly method, Yamamoto method and moving t-test method were used to test the abrupt change. And the Morlet wavelet analysis was used to test the variation of precipitation for the different periods of the year. The Geostatistical Analyst Module of GIS and GS plus was used to compare and select the optimal interpolation model based on the geographical conditions and statistical data characteristics. The results were as follows:(1) On trend analysis--the liner trends of annual precipitation, the precipitation in spring, summer, autumn, winter and flood season rainfall in FHRB during 1971～2011 were-6.0 mm/10 a,5.3 mm/10 a,-10.0 mm/10 a, 6.6 mm/10 a,0.2 mm/10 a and-7.2 mm/10 a, respectively. Negative trends of annual precipitation and rainfall in summer and flood season were found, while positive trends of precipitation in spring, autumn and winter were found in FHRB, TMR and TLR. On the contrary, positive trends of annual precipitation and rainfall in spring, summer and autumn besides flood season were found, while negative trend of precipitation in winter were found in TUR. It was related to its type of mountain watershed, which had different geographical environment compared with TMR and TLR. All the decrease or increase trends were not significant (Z> 0.10).(2) On mutation analysis--during the studied period, two abrupt changes in annual precipitation occurred in 1979 and 1997. Two abrupt changes of rainfall in spring occurred in 1975 and 1980. Abrupt changes of rainfall in summer and autumn were detected in 1982 and 2000, respectively. Two abrupt changes of precipitation in winter occurred in 1980 and 2000. To sum up, the typical abrupt change of precipitation fall happened in 1979,1980 and 1997, and the typical abrupt change of precipitation rise happened in 2000 and 2007. No abrupt change of rainfall in flood season occurred in FHRB and no abrupt change of annual precipitation and rainfall in flood season in TLR. Those were related to the lower latitude of TLR and more annual precipitation besides rainfall in flood season, which caused the insensitivity to the climate change.(3) On periodic variation analysis--in order of priority, there were obvious periodic oscillation of 2 a,32 a,6 a and 4 a for the annual precipitation variation,32 a,22 a,7 a and 4 a for the rainfall variation in spring,6 a,4 a and 16 a for the rainfall variation in summer,4 a,32 a,6 a and 14 a for the rainfall variation in flood season,9 a,2 a,32 a and 22 a for the rainfall variation in autumn,15 a,6 a,4 a and 32 a for the precipitation variation in winter. There exists certain degree of similarities in periodic oscillation of precipitation among different seasonal scales in different regions. According to the trend of unclosed contours in rainy period or completely closed contours in rainless period, there would be more precipitation at different temporal scales in a long time scale after 2011 in FHRB, TUR, TMR and TLR compared with the previous two decades or more years before 2011.(4) On spatial variation analysis--the coefficients of correlation between precipitation and NDVI at different time scales in FHRB were all significant at 0.05 confidence level, which proved the coupling relationship between precipitation and vegetation growth. Precipitation in different time scales increased significantly at 0.01 confidence levels with increasing altitude. Precipitation in different time scales decreased significantly at 0.01 confidence levels with increasing latitude. Annual precipitation, rainfall in flood season and autumn decreased significantly at 0.05 or 0.01 confidence levels with increasing longitude. Similar spatial distribution patterns were found for annual precipitation and rainfall in summer and flood season in FHRB, TMR and TLR, with decreasing trends from the south to the north and from mountain areas to basin areas, and the lowest rainfall value was in the central Taiyuan basin. The trends of isohyets were from the northeast to the southwest. In spring and winter, the trend of isohyets was from the east to the west with decreasing trend from the south to the north. In autumn, the trend of isohyets was from the northwest to the southeast with decreasing trend from the southwest to the northeast. Spatial distribution of precipitation was homogeneous in spring and winter in TUR, while annual precipitation and rainfall in summer, autumn and flood season were more in western mountain areas than eastern mountain areas and central valley areas. |
PDF Full Text Request