Trend and change-point analyses of streamflow and sediment discharge in the Yellow River during

Transcription

1 Hydrological Sciences Journal Journal des Sciences Hydrologiques ISSN: (Print) (Online) Journal homepage: Trend and change-point analyses of streamflow and sediment discharge in the Yellow River during Peng Gao, Xunchang Zhang, Xinming Mu, Fei Wang, Rui Li & Xiaoping Zhang To cite this article: Peng Gao, Xunchang Zhang, Xinming Mu, Fei Wang, Rui Li & Xiaoping Zhang (21) Trend and change-point analyses of streamflow and sediment discharge in the Yellow River during , Hydrological Sciences Journal Journal des Sciences Hydrologiques, 55:2, , DOI: 1.18/ To link to this article: Published online: 24 Mar 21. Submit your article to this journal Article views: 18 Citing articles: 3 View citing articles Full Terms & Conditions of access and use can be found at

2 Hydrological Sciences Journal Journal des Sciences Hydrologiques, 55(2) THSJ Trend and change-point analyses of streamflow and sediment discharge in the Yellow River during Trend and change-point analyses of streamflow and sediment discharge in the Yellow River Peng Gao 1, Xunchang Zhang 2, Xinming Mu 1, Fei Wang 1, Rui Li 1 & Xiaoping Zhang 1 1 Institute of Soil and Water Conservation, Northwest A&F University, 26 Xinong Road, Yangling 7121, Shaanxi Province, China xmmu@ms.iswc.ac.cn 2 USDA-ARS Grazinglands Research Laboratory, 722 W Cheyenne St., El Reno, Oklahoma 7336, USA Received 15 May 28; accepted 7 August 29; open for discussion until 1 September 21 Citation Gao, P., Zhang, X.-C., Mu, X.-M., Wang, F., Li, R. & Zhang, X. (21) Trend and change-point analyses of streamflow and sediment discharge in the Yellow River during Hydrol. Sci. J. 55(2), Abstract The objectives of this work are: (a) to statistically test and quantify the decreasing trends of streamflow and sediment discharge of the Yellow River in China during , (b) to identify change points or transition years of the decreasing trends, and (c) to diagnose whether the decreasing trends were caused by precipitation changes or human intervention, or both. The results show that significant decreasing trends in annual streamflow and sediment discharge have existed since the late 195s at three stations located in the upper, middle, and lower reaches of the Yellow River (P =.1). Change-point analyses further revealed that transition years existed and that rapid decline in streamflow and sediment discharge began in 1985 in most parts of the basin (P =.5). Adoption of conservation measures in the 198s and 199s corroborates the identified transition years. Double-mass curves of precipitation vs streamflow (sediment) for the periods before and after the transition years show remarkable decreases in proportionality of streamflow (sediment) generation. All percentiles of streamflow and sediment discharge after the transition years showed rapid reduction. In the absence of significantly decreasing precipitation trends, it is concluded that the decreasing trends were very likely caused by human intervention. Relative to the period before the transition, human intervention during reduced cumulative streamflow by 13.5, 14.3 and 24.6% and cumulative sediment discharge by 29., 24.8 and 26.5%, at Toudaoguai, Huayuankou and Lijin, respectively, showing the quantitative conservation effect in the basin. Key words Yellow River; precipitation; streamflow; sediment discharge; human intervention; change-point analysis Analyses des tendances et des points de changement dans les débits d eau et de sédiments du Fleuve Jaune de 195 à 25 Résumé Les objectifs de ce travail sont: (a) de tester et de quantifier statistiquement les tendances décroissantes des débits d eau et de sédiments du Fleuve Jaune, Chine, de 195 à 25, (b) d identifier les points de changement ou les années de transition des tendances décroissantes, et (c) de diagnostiquer si les tendances décroissantes sont causées par des changements dans les précipitations, ou par des interventions humaines, ou par les deux. Les résultats montrent l existence depuis la fin des années 195 de tendances décroissantes des débits annuels d eau et de sédiments pour trois stations situées dans les tronçons amont, central et aval du Fleuve Jaune (P =.1). L analyse des points de changement révèle que des années de transition existent et que le déclin rapide des débits d eau et de sédiments a commencé en 1985 dans la majeure partie du bassin versant (P =.5). L adoption de mesures de conservation dans les années 198 et 199 corrobore les années de transition identifiées. Les courbes des doubles cumuls précipitation/débit liquide (sédiments) pour les périodes avant et après les années de transition présentent des baisses remarquables en proportion de la génération de l écoulement (sédiments). Tous les percentiles des débits liquides et de sédiments après la transition présentent une réduction rapide. En l absence de tendances de précipitations décroissantes significatives, il est conclu que les tendances décroissantes sont très vraisemblablement dues aux interventions humaines. Relativement à la période précédant la transition, les interventions humaines entre 1985 et 25 ont réduit les débits liquides cumulés de 13.5, 14.3 et 24.6%, et les débits sédimentaires cumulés de 29., 24.8 et 26.5%, respectivement à Toudaoguai, Huayuankou et Lijin, montrant l effet quantitatif de la conservation dans le bassin versant. Mots clefs Fleuve Jaune; précipitation; débit de rivière; débit sédimentaire; intervention humaine; analyse des points de changement ISSN print/issn online 21 IAHS Press doi:1.18/

3 276 P. Gao et al. INTRODUCTION Rivers are major pathways for delivering terrestrial materials to oceans (Walling & Fang, 23; Meybeck & Vörösmarty, 25). In the second half of the 2th century, environmental changes resulting from intensified human activity have altered many large river systems in the world (Steffen, 24). The Yellow River, which is extensively influenced by human activity (Mei & Dregne, 21; Xu, 23), is an excellent basin for studying the impacts of climate variability and intensified human activity over the last several decades (Walling & Fang, 23; Wang et al., 27). Streamflow and sediment discharge provide useful information on processes of soil erosion and sediment delivery occurring in a basin (Siakeu et al., 24). Due to improper land use and excessive exploitation, the upper and middle reaches of the Yellow River basin are counted among the most severely eroded areas in the world. Since the 195s, many soil conservation measures have been implemented in the basin, which include construction of terraces, dams and reservoirs, conversion of croplands to grasslands and woodlands, and vegetation restoration (Yu, 26; Zheng et al., 27). To date, it is believed that the Yellow River basin has become one of the basins with the most human interventions in China (Mei & Dregne, 21; Xu, 23). Recent studies have shown that streamflow and sediment discharge of the Yellow River decreased since the late 195s (Yu, 26). Fu et al. (27) have shown that climate variability had significant impacts on streamflow in the Yellow River and that streamflow was sensitive to both precipitation and temperature in the basin. Zheng et al. (27) reported that there were no significant changes in trend for annual streamflow in four headwater catchments of the Yellow River during , based on the non-parametric Mann-Kendall test. Li et al. (27) studied annual streamflow during in the Wuding River (a tributary in the middle reaches of the Yellow River) and reported that there was a significant downward trend in annual streamflow, beginning in The streamflow during was reduced by some 42% compared with the period. They further estimated that 87% of the total streamflow reduction was caused by implementation of soil conservation measures and the remaining 13% by changes in precipitation and potential evaporation. Although several studies have shown that total amounts of streamflow and sediment discharge have decreased in the Yellow River, the magnitudes of the decreases have not yet been fully quantified and statistically tested in a systematic manner for different physiographic regions (distinctive sub-basins) as well as the entire basin. The downward trends need to be statistically tested in order to discern whether they are random fluctuations or actual changes in trends. If a downward trend exists, it is important and useful to further know exactly when the change began and which factors drive or cause the change. The quantitative effect of each driving factor (nature vs human) has not yet reached scientific consensus (Huang & Zhang, 24; Liu & Zhang, 24). Understanding the impacts of climate variation and human activity on the hydrological regime and sediment dynamics is useful for developing effective conservation strategies in the Yellow River basin. The objectives of this study are: (a) to statistically detect trends and change-points in annual streamflow and sediment discharge at four key hydrological stations along the Yellow River mainstream; (b) to analyse possible impacts of precipitation factors and human activities on annual streamflow and sediment discharge dynamics in relation to change-points or transition years detected in this study; and (c) to further estimate the effects of the identified driving factor based on the two contrastive periods before and after the transition years. STUDY AREA AND DATA SETS The Yellow River originates in the Qinghai-Tibet Plateau and flows about 54 km eastwards towards the Bohai Sea. The basin lies between 96 E 119 E and 32 N 42 N with a drainage area of 795 km 2. A data set from 2 meteorological stations with long-term annual precipitation data ( ) in the Yellow River basin was analysed in this study. Data were obtained from precipitation stations of the Yellow River in: the sub-basins of the source water area, above the Tangnaihai hydrological station (two stations); the upper reaches, between Tangnaihai and Toudaoguai (seven stations); the middle reaches, between Toudaoguai and Huayuankou (1 stations); and the lower reaches, between Huayuankou and Lijin (one station) (see Fig. 1). The precipitation data were provided by The National Meteorological Information Centre (NMIC), China Meteorological Administration (CMA). Four key hydrological stations in the Yellow River mainstream (Tangnaihai, Toudaoguai, Huayuankou

4 Trend and change-point analyses of streamflow and sediment discharge in the Yellow River 277 Fig. 1 Location of the study region and stations in the Yellow River basin (circles: rain stations, and squares: Tangnaihai, Toudaoguai, Huayuankou and Lijin hydrological stations along the Yellow River mainstream). The shaded zone is the inland drainage area. Table 1 Drainage area and average annual streamflow and sediment discharge at the four hydrological stations. Hydrological station Drainage area (1 4 km 2 ) Streamflow (1 8 m 3 ) Sediment discharge (1 8 t) Tangnaihai (195 25).13 ( ) Toudaoguai (195 25) 1.8 (195 25) Huayuankou (195 25) 9.72 (195 25) Lijin ( ) 7.78 ( ) The record length used in this study is given in parentheses. and Lijin), which gauge streamflow and sediment discharge at the source water area, upper, middle and lower reaches of the Yellow River, respectively, were chosen for analysis (Fig. 1); the corresponding drainage areas are given in Table 1. Annual streamflow and sediment discharge data at the four stations from 195 to 25 were obtained from the Chinese River Streamflow and Sediment Communiqués, Ministry of Water Resources, China (MWR). All measured data used in this study are of good quality and were checked for quality control by the corresponding agencies. ANALYSIS METHODS Trend test The rank-based, non-parametric Mann-Kendall statistical test (Mann, 1945; Kendall, 1975) is commonly used for trend detection due to its robustness for nonnormally distributed and censored data, which are frequently encountered in hydroclimatic time series (e.g. Hirsch et al., 1982; Burn & Elnur, 22; Yue et al., 23; Yue & Pilon, 24). The results of the trend test can be used to determine whether or not the observed time series of hydrological variables exhibits a trend that is statistically significant from a trend that could occur by chance. However, to do this, it is necessary to test the serial correlation of the data (Jenkins & Watts, 1968). The presence of serial correlation can complicate the identification of trends, in that a positive serial correlation can increase the expected number of false positive outcomes for the Mann-Kendall test (von Storch & Navarra, 1995). Thus, any serial correlation should be removed before conducting the Mann-Kendall trend test. In this work the trend free pre-whitening (TFPW) method of Yue et al. (23) was used as follows. (1) Remove any significant linear trend from the raw time series using: Y = X β t (1) t t

5 278 P. Gao et al. where X t is the series value at time t; b is the linear regression slope of the trend in the raw time series; and Y t is the de-trended series. (2) Remove serial correlation if the lag-one serial correlation coefficient (r 1 ) of the de-trended series is statistically significant at the 5% level using the pre-whitening method of Kulkarni & von Storch (1995): where Y t is the de-trended and pre-whitened series, referred to as the residual series. (3) Add the linear trend that was removed at Step 1 back to the de-trended or residual series using: where Y t is the trend free pre-whitened series. A Z statistic was obtained from the Mann-Kendall test on the whitened series from Step 3. In addition, to confirm the results provided by the Mann-Kendall test, we also performed linear regression analysis. Change-point analysis A number of methods can be applied to determine change points of a time series (Buishand, 1982; Chen & Gupta, 2; Radziejewski et al., 2). In this study, we used the non-parametric approach developed by Pettitt (1979) to detect change-points in streamflow and sediment discharge time series. This method detects a significant change in the mean of a time series when the exact time of the change is unknown. The test uses a version of the Mann-Whitney statistic U t,n, that tests whether two sample sets x 1, x t and x t+1, x N are from the same population. The test statistic U t,n is given by: and Y t = Y t ry t 1 1 (2) Y = Y +β t (3) t t N UtN = Ut, N+ sgn( Xt Xj) for t = 2,..., N, 1 if if if j= 1 ( Xt X j) > sgn( Xt X j) = 1 ( Xt X j) = sgn( Xt X j) = ( X X ) < sgn( X X ) = 1 t j (4) The test statistic counts the number of times a member of the first sample exceeds a member of the t j (5) second sample. The null hypothesis of the Pettitt test is the absence of a change point. The test statistic K N and the associated probability (P) used in the test are given as: K N = t N t N Double-mass curve max U (6) 1, P 2exp { 6( K ) /( N + N )} The theory of the double-mass curve is based on the fact that a plot of the two cumulative quantities during the same period exhibits a straight line so long as the proportionality between the two remains unchanged, and the slope of the line represents the proportionality. This method can smooth a time series and suppress random elements in the series, and thus show the main trends of the time series. In this study, double-mass curves of precipitation vs streamflow and precipitation vs sediment are plotted for the two contrasting periods to estimate changes in regression slope (proportionality) to quantify the overall efficiency of soil conservation measures before and after transition years. RESULTS AND DISCUSSION N Trend analysis of observed annual precipitation, streamflow and sediment discharge (7) The observed average annual streamflow and sediment discharge at the four hydrological stations along the Yellow River mainstream are given in Table 1. The long-term averaged annual streamflow and sediment discharge increased downstream at the Tangnaihai, Toudaoguai and Huayuankou stations, but decreased at Lijin station. The decreases in the lower reaches of the Yellow River (between Huayuankou and Lijing) were because of water extraction, seepage loss and sediment siltation due to the fact that the river bed was several metres above the flood plain and its longitudinal gradient was extremely gentle (Yu, 26). It should be pointed out that about 9% of sediment discharge at Huayuankou came from the middle reaches between Toudaoguai and Huayuankou, where the most severely eroded Loess Plateau is situated. The observed annual precipitation, streamflow and sediment discharge during at the four

6 Trend and change-point analyses of streamflow and sediment discharge in the Yellow River 279 stations are shown in Fig. 2(a) (d), and their corresponding Mann-Kendall test results in Table 2. The streamflow, sediment discharge and precipitation did not show significant change at Tangnaihai (Fig. 2(a), Table 2). This sub-basin is considered to represent natural conditions of runoff and streamflow generation because of minimal human intervention in this headwater region. However, at the Toudaoguai, Huayuankou and Lijin stations, the streamflow and sediment discharge decreased significantly, while average annual precipitation depths in their corresponding drainage areas did not show a significant downward trend (Fig. 2, Table 2). Change-point analysis Since the Mann-Kendall tests showed significant downward trends for streamflow and sediment discharge, the Pettitt test was further used to detect the change points or transition years when the significant changes began (Table 3, Fig. 2). The Pettitt test was also applied to annual precipitation for the four stations, and results showed that no transition years could be detected at P =.5. These results corroborated the Mann-Kendall test results of lack of significant trends in precipitation. For annual streamflow, the change-point year of 1985 was detected for the Toudaoguai, Huayuankou, and Lijin stations (P =.5), Fig. 2 Observed annual precipitation, streamflow and sediment discharge during at: (a) Tangnaihai, (b) Toudaoguai, (c) Huayuankou, (d) Lijin hydrological stations in the Yellow River basin. The horizontal solid lines are the mean values and the vertical dashed lines indicate the transition years.

7 28 P. Gao et al. Table 2 Results of trend analysis for annual precipitation, streamflow and sediment discharge at four stations. Station Variable Mann-Kendall Linear regression Z statistic Sig. level t statistic Sig. level Tangnaihai Precipitation ns ns Streamflow ns ns Sediment discharge ns ns Toudaoguai Precipitation ns ns Streamflow 4.7 ** ** Sediment discharge 5.11 ** ** Huayuankou Precipitation.755 ns.925 ns Streamflow ** ** Sediment discharge ** ** Lijin Precipitation.392 ns.265 ns Streamflow 5.99 ** ** Sediment discharge ** ** A positive sign indicates an increasing trend and a negative sign a decreasing trend. ** significant at P =.1; ns: not significant at P.1. Table 3 Change-point years detected by the Pettitt method for streamflow and sediment discharge. Station Variable Change-point year Tangnaihai Precipitation - Streamflow Sediment discharge - Toudaoguai Precipitation - Streamflow 1985* Sediment discharge 1985* Huayuankou Precipitation - Streamflow 1985* Sediment discharge 1979* Lijin Precipitation - Streamflow 1985* Sediment discharge 1985* *Significant at P =.5. + Significant at P =.1. and the year of 1989 for Tangnaihai (P =.1). However, considering the test result of an insignificant trend (P =.5) at Tangnaihai, it can be safely assumed that a change point does not exist in this source-water sub-basin (P =.5). For annual sediment discharge for the four stations, the change-point year of 1985 was detected for the Toudaoguai and Lijin stations, and the year of 1979 for Huayuankou with a significance level of 5% (Table 3, Fig. 2). Frequency exceedence curves before and after the change-point year Based on the results of the change-point analyses, streamflow and sediment discharge were divided into two periods by the transition years for the Toudaoguai, Huayuankou and Lijin stations, and further analysed for shifts in the exceedence probability curves for the two periods. All percentile streamflow and sediment discharge after the transition years were shifted to much smaller values relative to those of the period before the transition years (plots not shown). The shift indicated a significant decrease in streamflow or sediment discharge at any given probability. Streamflow and sediment discharge of the 1th, 5th and 9th percentiles were chosen to represent wet-, normal-, and dry-year conditions for further comparisons for periods both before and after the transition years for the Toudaoguai, Huayuankou and Lijin stations (Tables 4 and 5). The percentage changes, relative to the values before the transition years, are also Table 4 Annual streamflow of the 1th, 5th and 9th percentiles (P 1, P 5 and P 9 ) for the two periods before and after the change-point year at Toudaoguai, Huayuankou and Lijin stations. Station Before change-point year (1 8 m 3 ) After change-point year (1 8 m 3 ) Change (%) Wet year, P 1 Normal year, P 5 Dry year, P 9 Wet year, P 1 Normal year, P 5 Dry year, P 9 Wet year, P 1 Normal year, P 5 Dry year, P 9 Toudaoguai Huayuankou Lijin

8 Trend and change-point analyses of streamflow and sediment discharge in the Yellow River 281 Table 5 Annual sediment discharge of the 1th, 5th and 9th percentiles (P 1, P 5 and P 9 ) for the two periods before and after the change-point year at Toudaoguai, Huayuankou and Lijin stations. Station Before change-point year (1 8 t) After change-point year (1 8 t) Change (%) Wet year, P 1 Normal year, P 5 Dry year, P 9 Wet year, P 1 Normal year, P 5 Dry year, P 9 Wet year, P 1 Normal year, P 5 Dry year, P 9 Toudaoguai Huayuankou Lijin presented. Percentage decrease in streamflow after the transition year ranged from 34.6% for dry years at Toudaoguai to 8.5% for dry years at Lijin, relative to the period before the transition year (Table 4). The percentage reductions at Huayuankou and Lijin were greater in dry years than in wet years because water extraction for agricultural use was greater in dry years. Decrease of sediment discharge after the transition year ranged from 5.% for wet years at Huayuankou to 95.2% for dry years at Lijin (Table 5). In general, percentage reduction in sediment was greater than its counterpart in streamflow, especially in dry years. This is because: (a) water diversion/ extraction reduces not only streamflow but also sediment loads, (b) reduced flow has much smaller sediment transport capacity and therefore increases chances for sediment deposition in the riverbed, and (c) most conservation measures, such as check dams, reservoirs, conversion of croplands to grasslands, are more efficient in trapping sediment than water. All these factors seem more efficient in reducing sediment in drier years. Double-mass curve of precipitation-streamflow and precipitation-sediment To further quantify the streamflow and sediment discharge changes before and after the transition years, double-mass curves, along with the linear regression lines, are plotted in Fig. 3 for the Toudaoguai, Huayuankou and Lijin stations. There existed clear breakpoints between the two regression lines for both streamflow and sediment discharge for the three stations, suggesting that the transition years identified by Pettitt s method are correct and meaningful. The slopes of the regression lines were lower after the breakpoints or transition years (i.e. at higher cumulative precipitation values) than before the breakpoints for both streamflow and sediment discharge at the three stations. Since there was no discernible change of precipitation in each sub-basin, the lower regression slope demonstrated that the proportionalities between streamflow and precipitation, as well as between sediment discharge and precipitation, were considerably reduced after the transition years. To estimate the relative reduction of total streamflow and sediment discharge for the period after the transition years, equations fitted to the double-mass curves before the transition years (Tables 6 and 7) were extrapolated to the cumulative precipitation amounts of 25. The extrapolated cumulative streamflow (R c in Table 6) and sediment discharge (S c in Table 7) were based on the assumption that environmental conditions, including human impacts in the basin in the first period (before the transition years), remained somewhat unchanged in the second period (after the transition years). Compared with the extrapolated cumulative streamflow (R c ), observed cumulative streamflow (R o ) was reduced by 13.5, 14.3, and 24.6% at the Toudaoguai, Huayuankou, and Lijin stations, respectively (Table 6). The corresponding reduction for sediment discharge was 29., 24.8, and 26.5% at the three stations (Table 7). It should be noted that the percent reductions in cumulative sediment discharge were greater than those of cumulative streamflow. The reasons for this are similar to those given at the end of the section above. Given that streamflow and sediment discharge have decreased significantly since the late 195s, a relevant question is about factors that have caused those changes. Based on the Yellow River literature (Rao & Huo, 21), precipitation variation and human activity are the two most likely factors affecting basin hydrology and sediment delivery. In this study, precipitation did not show any significant change in trend in the basin since the 195s according to the Mann-Kendall and Pettitt tests. Thus, the most plausible factor responsible for the streamflow and sediment discharge decreases is the human factor. This conclusion is in agreement with those reported in previous studies (van den Elsen et al., 23; Xu, 23; Huang & Zhang, 24; Mu et al., 27).

9 282 P. Gao et al. Cumulative Precipitation (mm) Cumulative Precipitation-Streamflow Cumulative Precipitation-Sediment 2 Cumulative Streamflow (1 8 m 3 ) Cumulative Sediment (1 8 T) Cumulative Precipitation (mm) Cumulative Precipitation (mm) (a) 12 Cumulative Precipitation-Streamflow Cumulative Precipitation-Sediment 6 Cumulative Streamflow (1 8 m 3 ) Cumulative Sediment (1 8 T) Cumulative Precipitation (mm) (b) Cumulative Precipitation (mm) Cumulative Precipitation-Streamflow Cumulative Precipitation-Sediment 6 Cumulative Streamflow (1 8 m 3 ) Cumulative Sediment (1 8 T) Cumulative Precipitation (mm) Fig. 3 Double-mass curves of precipitation streamflow and precipitation sediment during at: (a) Toudaoguai, (b) Huayuankou and (c) Lijin stations in the Yellow River mainstream. The straight lines are the regression lines for the cumulative data before and after change-point years. (c)

10 Trend and change-point analyses of streamflow and sediment discharge in the Yellow River 283 Table 6 Linear regression equations between cumulative streamflow and cumulative precipitation for the period before the transition years for Toudaoguai, Huayuankou and Lijin stations. Station Regression equation R c (1 8 m 3 ) R o (1 8 m 3 ) R c R o (1 8 m 3 ) 1(R c R o )/R c (%) Toudaoguai R =.675P (R 2 =.999) Huayuankou R =.982P (R 2 =.998) Lijin R =.8693P (R 2 =.99) R: cumulative streamflow; P: cumulative precipitation; R c : extrapolated cumulative streamflow by 25; R o : observed cumulative streamflow by 25. Table 7 Linear regression equations between cumulative sediment discharge and cumulative precipitation for the period before the transition years for Toudaoguai, Huayuankou Lijin stations. Station Regression equation S c (1 8 t) S o (1 8 t) S c S o (1 8 t) 1(S c S o )/S c (%) Toudaoguai S =.4P (R 2 =.991) Huayuankou S =.282P (R 2 =.995) Lijin S =.221P (R 2 =.993) S: cumulative sediment discharge; P: cumulative precipitation; S c : extrapolated cumulative sediment discharge by 25; S o : observed cumulative sediment discharge by 25. Human activity, such as the implementation of a broad range of soil and water conservation measures (e.g. constructing check dams and reservoirs, building water diversion irrigation systems, building bench and reverse slope terraces, converting croplands to grasslands, afforestation and other land-use changes), would most likely be responsible for the decreases in streamflow and sediment discharge in the Yellow River basin since the late 195s, and especially since Role of human activities in decreasing streamflow and sediment discharge Many studies have documented human and economic activities (especially after the 198s) that might have played an important role in reducing streamflow and sediment discharge in the Yellow River basin (van den Elsen et al., 23; Xu, 23; Huang & Zhang, 24; Mu et al., 27). Those activities can be summarized as follows: (a) Increased demand for water resources of the Yellow River for national economic development. With the rapid development of China s national economy, the demand for water resources from the Yellow River has increased steadily, especially after the 198s. Water extraction and diversion for agricultural irrigation and urban and industrial use has dramatically increased during the rapid development of the national economy, especially after the (b) 198s (Liu & Zhang, 24). This partially explains why the identified transition year of 1985, i.e. the inception of the downward trends in streamflow and sediment discharge, coincides with the fast-growing economy in the 198s and 199s in the Yellow River basin. Impact of soil and water conservation programmes and ecological environment rehabilitation campaign in the Loess Plateau. The Loess Plateau, situated in the middle reaches of the Yellow River, is the major sediment source area for the river. To reduce water and soil erosion, some water and soil conservation measures were implemented between 195 and 1978, cf. Table 8 (Mu et al., 27). However, a broad range of extensive conservation measures were installed between 1979 and 1997 due to various government-sponsored conservation programmes and environmental rehabilitation campaigns in the Yellow River basin. About 13.2% of the basin had some type of conservation measures by 1978, and the percentage increased to 41.1% by Based on the average area that received conservation measures per year, the rates of terrace building, check dam construction, afforestation and grass planting during were 2.1, 1.3, 4.1 and 5.1 times the rates during (Table 8). By 26, about 49% of eroded land was under protection, with some sort of soil and water conservation measures in the Loess Plateau (including km 2 of prime

11 284 P. Gao et al. Table 8 Land areas of major conservation practices and percentage area under control measures in different periods in the Loess Plateau in the middle reaches of the Yellow River basin. Control measures Increment in controlled area in different periods: Bench terrace (km 2 ) Farmland in check dam (km 2 ) Afforestation (km 2 ) Grass planting (km 2 ) Total conservation area (km 2 ) Cumulative conservation area (km 2 ) Cumulative area under conservation (%) Data source: (c) farmlands, km 2 of soil and water conservation forest and km 2 of grass planting), and more than 27 structures of key projects for gully erosion control and more than 4 3 structures of assisted small-scale projects were completed. Undoubtedly, the rapid adoption of soil and water conservation measures and engineering structures in the 198s and 199s have played a significant role in reducing streamflow and sediment discharge from the Loess Plateau in the middle reaches of the Yellow River. This rapid adoption period is in good agreement with the transition year of 1985 identified by the change-point analysis, considering that the conservation effects on sediment reduction may have a time lag in such a large basin. Impact of construction of water control projects. Construction of large/medium-sized multi-purpose water control projects has played a role in reducing sediment discharge in the Yellow River (Wang et al., 25). Reservoir siltation, though undesirable, has reduced sediment discharge in the mainstream. There were only a few large/medium-sized multi-purpose water control projects in the 195s in the mainstream. However, 18 large-sized and 118 medium-sized reservoirs were constructed to meet the needs of economic development and ecological rehabilitation by 22. Those reservoirs, facing various degrees of siltation problems, have reduced sediment discharge downstream. Apart from reservoirs, the rainwater collection project implemented in the upper and middle reaches of the Yellow River has also contributed to streamflow and sediment discharge reduction since the 198s. In this project, more than 2 5 small cistern structures were built by 2 to collect storm water for later use in the dry season. It was reported that the rainwater collection project stored nearly 1 million m 3 of precipitation water each year, which included the rainfall of precipitation events that did not produce runoff (Liu & Zhang, 24). SUMMARY AND CONCLUSIONS Precipitation, streamflow, and sediment discharge during were analysed by the Mann-Kendall trend test and the Pettitt change-point analysis for the Tangnaihai, Toudaoguai, Huayuankou, and Lijin stations along the Yellow River mainstream. No significant changes in trends were detected for the observed annual precipitation in all four sub-basins. However, significant downward trends in annual streamflow and annual sediment discharge were detected for Toudaoguai, Huayuankou and Lijin, except in the source water area at Tangnaihai. The change point, or transition year, for streamflow and sediment discharge was 1985 for Toudaoguai and Lijin. The transition year for Huayuankou was 1979 for sediment discharge and 1985 for streamflow. As indicated by the transition years, the downward trend began in 1985 for streamflow and sediment discharge in most parts of the Yellow River basin. Human intervention was largely responsible for the downward trends after the transition years. The effects of human intervention on reducing streamflow and sediment discharge could be quantified by comparing the two periods with the double-mass curves. Compared to the period before the transition years, by 25 (i.e. after the transition years), the measured cumulative streamflow at Toudaoguai, Huayuankou and Lijin, respectively, had decreased by 13.5%, 14.3% and 24.6%, and sediment discharge by 29.%, 24.8% and 26.5%. Since there was no changing trend in precipitation in the region, these

12 Trend and change-point analyses of streamflow and sediment discharge in the Yellow River 285 percentage reductions are very likely to have been caused by human intervention. Soil and water conservation in the Yellow River basin began in the late 195s, and the pace was more than tripled after 198, based on areas receiving conservation measures per year. The more extensive adoption of a broad range of conservation measures in the 198s and 199s, which will have altered the natural regimes of streamflow and sediment discharge, was in good agreement with the transition year of 1985, when a significant downward trend began according to the Pettitt test. The rates of decreases in streamflow and sediment discharge coincided well with the intensity and extent of human intervention and activities. The overall results showed that human activities, such as soil and water conservation programmes, an eco-environmental rehabilitation campaign, and construction of key water control projects, appear to be the major factors causing annual streamflow and sediment discharge to decrease significantly in the Yellow River basin, except in the source water sub-basin. Acknowledgements This paper is financially supported by The National Basic Research Programme (also called 973 Programme) (27CB4723) and Programme KZCX2-XB The authors would also like to thank Dr Ninghu Su of James Cook University and Dr John Gates for their insightful comments and suggestions. REFERENCES Buishand, T. A. (1982) Some methods for testing the homogeneity of rainfall records. J. Hydrol. 58(1-2), Burn, D. H. & Elnur, M. A. (22) Detection of hydrologic trends and variability. J. Hydrol. 255(1-4), Chen, J. & Gupta, A. K. (2) Parametric Statistical Change Point Analysis. Birkhauser, Boston, Massachusetts, USA. Fu, G. B., Charles, S. P., Viney, N. R., Chen, S. & Wu, J. Q. (27) Impacts of climate variability on stream-flow in the Yellow River. Hydrol. Processes 21(25), Hirsch, R. M., Slack, J. R. & Smith, R. A. (1982) Techniques of trend analysis for monthly water quality data. Water Resour. Res. 18, Huang, M. B. & Zhang, L. (24) Hydrological responses to conservation practices in a catchment of the Loess Plateau, China. Hydrol. Processes 18(1), Jenkins, G. M. & Watts, D. G. (1968) Spectral Analysis and its Applications. Holden-Day, San Francisco, California, USA. Kendall, M. G. (1975) Rank Correlation Measures. Charles Griffin, London, UK. Kulkarni, A. & von Storch, H. (1995) Monte Carlo experiments on the effect of serial correlation on the Mann-Kendall test of trend. Meteorol. Zeitschrift 4(2), Li, L. J., Zhang, L., Wang, H. Wang, J. Yang, J. W., Jiang, D. J., Li, J. Y. & Qin, D. Y. (27) Assessing the impact of climate variability and human activities on streamflow from the Wuding River basin in China. Hydrol. Processes 21(25), Liu, C. M. & Zhang, X. C. (24) Causal analysis on actual water flow reduction in the mainstream of the Yellow River. Acta Geographica Sinica 59(3), (in Chinese). Mann, H. B. (1945) Nonparametric tests against trend. Econometrica 13(3), Mei, C. R. & Dregne, H. E. (21) Review article: Silt and the future development of China s Yellow River. The Geogr. J. 167(1), Meybeck, M. & Vörösmarty, C. (25) Fluvial filtering of land-to ocean fluxes: from natural Holocene variations to Anthropocene. C. R. Geosci. 337, Mu, X. M., Zhang, L., McVicar, T. R., Chille, B. & Gau, P. (27) Analysis of the impact of conservation measures on stream flow regime in catchments of the Loess Plateau, China. Hydrol. Processes 21(16), Pettitt, A. N. (1979) A non-parametric approach to the change-point problem. Appl. Statist. 28(2), Radziejewski, M., Bardossy, A. & Kundzewicz, Z. W. (2) Detection of change in river flow using phase randomization. Hydrol. Sci. J. 45(4), Rao, S. Q. & Huo, S. Q. (21) Analysis on the characteristics of runoff and sediment and the outlook of their tendency in the future in the upper and middle Yellow River. J. Sediment Res. 2, (in Chinese). Siakeu, J., Oguchi, T., Aoki, T., Esaki, Y. & Jarvie, H. P. (24) Change in riverine suspended sediment concentration in central Japan in response to late 2th century human activities. Catena 55(2), Steffen, W. (24) Global Change and the Earth System: A Planet under Pressure. Springer, New York, USA. van den Elsen, E., Hessel, R., Liu, B. Y., Trouwborst, K. O., Stolte, J., Ritsema, C. J. & Blijenberg, H. (23) Discharge and sediment measurements at the outlet of a watershed on the Loess plateau of China. Catena 54(1 2), von Storch, H. & Navarra, A. E. (1995) Analysis of Climate Variability. Springer, New York, USA. Walling, D. E. & Fang, D. (23) Recent trends in the suspended sediment loads of the world s rivers. Global & Planetary Change 39, Wang, G. Q., Wu, B. S. & Wang, Z. Y. (25) Sedimentation problems and management strategies of Sanmenxia Reservoir, Yellow River, China. Water Resour. Res. 41, W9417, doi:1.129/24wr3919. Wang, H. J., Yang, Z. S., Saito, Y., Liu, J. P., Sun, X. X. & Wang, Y. (27) Stepwise decreases of the Huanghe (Yellow River) sediment load (195-25): Impacts of climate change and human activities. Global & Planetary Change 57(3 4), Xu, J. X. (23) Sediment flux to the sea as influenced by changing human activities and precipitation: example of the Yellow River, China. Environ. Manage. 31(3), Yu, L. S. (26) The Huanghe (Yellow) River: recent changes and its countermeasures. Continental Shelf Res. 26(17-18), Yue, S. & Pilon, P. (24) A comparison of the power of the t test, Mann-Kendall and bootstrap tests for trend detection. Hydrol. Sci. J. 49(1), Yue, S., Pilon, P. & Phinney, B. (23) Canadian streamflow trend detection: impacts of serial and cross-correlation. Hydrol. Sci. J. 49(1), Zheng, H. X., Zhang, L., Liu, C. M., Shao, Q. X. & Fukushima, Y. (27) Changes in stream flow regime in headwater catchments of the Yellow River basin since the 195s. Hydrol. Processes 21(7),