| Global warming and the increasing human activities influence hydrological cycles,resulting in more frequent and severe hydroclimatic extremes with devastating impacts on agriculture,ecosystems,industry,and socio-economic sustainable development.In general,the occurrences of extreme hydroclimatic events are closely related to nonstationarity,whilst the assumption of stationarity in hydrological series is the foundation for hydrological frequency analysis.Therefore,it is of important significance to study the variations of extreme hydroclimatic events under the changing environments,which would provide useful information for regional water resources management,adaptive strategies and eco-enviroment protection as well as the sustainable development of socio-economy.In this study,the Wei River Basin(WRB),a typical semi-humid and semi-arid and eco-environmentally vulnerable region was selected as the study area.The changing patterns,stationarity and the underlying causes of extreme temperature,extreme precipitation,flood and low flow series were investigated with an aim to legitimately,systematically,and comprehensively reveal the hydrological response of the WRB to the changing environments.The primary findings obtained in this paper are as follows:(1)Based on the land transfer matrix method and land use/cover area change rate index,the evolution characteristics of the underlying surface of the WRB were studied.Results show that cultivated land,woodland and grassland are the main land use/cover modes in the basin.The past decades witnessed the convert of water area,towns,rural areas and construction land into cultivated land by a large margin.Overall,the area of cultivated land,high-cover grassland and urban land in the basin showed an increasing trend,while the area of woodland,shrub,low grass,water area,rural land and bare land showed a decreasing trend.Population growth,social and economic development,population mobility and policy factors are the most important driving forces of land use/cover change and the underlying surface conditions of the basin.(2)The climate evolution of the WRB were studied by linear regression method,MMK trend test method and heuristic segmentation method.Generally,there is a warming-drying tendency of the climate.Annual average temperature of the WRB shows a significant upward trend,and the spatial distribution of annual average temperature is uneven,gradually increasing from the northwest to the southeast.With change points dectected mostly in 1986 and 1996,the assumption of stationarity of the annual average temperature time series of the WRB is invaild.Human activities,including the use of fossil fuels,deforestation and changes in the underlying surface conditions,have a positive impact on the annual average temperature variation in the basin.Annual precipitation of the WRB shows a non-signficant decreasing trend.The spatial distribution of the annual precipitation is uneven,increasing from the northwest to the southeast.No change point was found in the annual precipitation series,which means annual precipitation series of the WRB is stationary.Annual runoff of the WRB is unevenly distributed during the year with a significant decreasing trend.Change points were found in the annual runoff series which concentrated in the late 1960s and early 1990s.Water withdrawl,changes in underlying surface conditions and climate change all lead to variations in annual runoff series.(3)Based on cloud model theory,trend test method and heuristic segmentation method,the uncertainty and nonstationarity of extreme temperature in the WRB were analyzed.Results show that the Tmin usually has higher dispersion,less stability and uniformity than the Tmax.Due to different underlying surface conditions,the extreme temperatures in the middle and lower reaches and Jinghe River Basin have higher dispersion,less uniformity and stability.The WRB is characterized by increasing trends of Tmax and Tmin,with more evident increasing trend in Tmin.There are change points found in Tmax series,which concentrated around 1996.Overall,the response of the Tmax to watershed warming is more significant than that of the Tmin.(4)A set of precipitation indices conerning the magnitude,frequecny and duration were adopted to study the changing patterns of precipitation extremes and the stationarity of extreme precipitation events.Furthermore,the correlations between large-scale ocean-atmospheric circulation patterns and precipitation extremes were explored.Results show that there is no significant response to regional climate warming-drying tendency excepet the upstream.With change points detected,the assumption of stationarity of extreme precipitation in the upstream of the WRB is invalid,thereby introducing large uncertainty to the design and management of water conservancy engineering.Although the underlying surface of the basin will affect the spatial and temporal distribution of extreme precipitation to some extent,it is not the direct reason to the nonstationarity of extreme precipitation series.Both Pacific Decadal Oscillation(PDO)and El Nino-Southern Oscillation(ENSO)events have a strong influence on precipitation extremes in the WRB,but their influence on the magnitude,frequecny and duration of precipitation extremes are different.(5)Nonstationary flood behavior in the WRB was examined comprehensively in terms of trends and the mean and variance change point with co-usage of different methods.Results show that the occurrence time of seasonal floods in the WRB are delayed.Magnitude of floods across the basin have high fluctuations.With detected significantly decreasing trends and change points,floods in the WRB exhibit nonstationary behaviors.Influenced by the regulation and storage of water conservancy projects and soil and water conservation measures in basin,the mean variation points often appear in the flood series of the main stream of the river basin,and the variance variation points often appear in the flood series of the Jinghe River basin.Bias arising from the variance change point is much more significant than that of the mean change point in estimating floods.Invalidity of stationarity in the flood series can be attributed to changes of underlying surface,forest vegetation,construction and storage of water conservancy projects.(6)Based on the Pearson correlation coefficient and Discrete wavelet transform,the stationarity and which time-scale is responsible for the presences of trend and change point in low flow series were determined.Moreover,the scale-dependent relationships between climatic variables and low flow series were studied.Results show that the assumption of stationarity in the low flow series over the study area is invalid with significantly decreasing trends and change points detected.The most important time-scales contributing to the trend production in the original low flow data are the 2-year and 4-year event and the most influential frequency for change points generation is 8-year periodic scale.The scale-dependent relationships between low flow series and climatic variables(precipitation,potential evaporation and soil moisture)demonstrate that the 2-year and 4-year scales generally present the strongest correlations,which suggests the importance of climatic factors for trend production in low flow series.Human activities,including water withdrawal and water and soil conservation,have direct influence on the nonstationarity of low flows. |
PDF Full Text Request