THE IMPACT OF METEOROLOGICAL DROUGHT ON STREAMFLOW IN THE JINGHE RIVER BASIN OF CHINA

Transcription

1 Proceedings of the 13th International Conference on Environmental Science and Technology Athens, Greece, 5-7 September 2013 THE IMPACT OF METEOROLOGICAL DROUGHT ON STREAMFLOW IN THE JINGHE RIVER BASIN OF CHINA LIN ZHAO 1, AIFENG LÜ 2, JIANJUN WU 3i, MICHAEL HAYES 4, ZHENGHONG TANG 5, and BIN HE 6 1. School of Resource and Environmental Sciences, Wuhan University, Wuhan , China; 2. Institute of Geographical Sciences and Natural Resources Research, CAS, Beijing , China; 3. Academy of Disaster Reduction and Emergency Management, Beijing Normal University, Beijing , China; 4. National Drought Mitigation Center, School of Natural resources, University of Nebraska-Lincoln, Lincoln, NE, USA, 68588; 5. College of Architecture, University of Nebraska-Lincoln, Lincoln, NE, USA, 68588; 6. Academy of College of Global Change and Earth System, Beijing Normal University, Beijing , China; EXTENDED ABSTRACT Under global climate change, drought has become one of the most serious natural hazards, affecting the ecological environment and human life. Drought can be categorized as meteorological, agricultural, hydrological or socio-economic drought. Among the different categories of drought, hydrological drought, especially streamflow drought, has been given more attention by governments, researchers and the public in recent years. Identifying the occurrence of streamflow drought and issuing early warning can provide timely information for effective water resources management. Hydrological drought has some relationships with meteorological drought in different areas. In this study, meteorological drought is detected by using the Standardized Precipitation Index (SPI), whereas streamflow drought is detected by the Standardized Runoff Index (SRI). Comparative analyses of frequency, magnitude, onset and duration are conducted to identify the impact of meteorological drought on streamflow drought. This study focuses on the Jinghe River Basin in Northwest China. It mainly provided the following findings. (1) 11 meteorological droughts and 6 streamflow droughts were indicated during 1970 to 1990 after pooling using IC (Inter-event time and volume Criterion) method. (2) Streamflow drought in Jinghe River Basin lagged meteorological drought for about 127 days. (3) The frequency of streamflow drought in Jinghe River Basin was less than meteorological drought. However, the average duration of streamflow drought is longer. (4) The magnitude of streamflow drought is greater than meteorological drought. These results not only play an important theoretical role in understanding relationships between different drought categories, but also have practical implications for streamflow drought prevention and regional water resources management Keywords: Streamflow drought; Meteorological drought; Time lag; Jinghe River Basin 1 INTRODUCTION Drought is a multi-faceted phenomenon that occurs across a range of temporal and spatial scales and is experienced across a range of societal sectors that are dependent on climate and water resources [Wilhite, 2000]. It can cause significant damages to natural environment and human activities [Austin et al., 1998; Nicholson et al., 1998; De Gaetano, i Corresponding Author: Jianjun Wu jjwu@bnu.edu.cn

2 1999; Pausas, 2004].The very dry areas over global land have increased from 12% to 30% since the 1970s [Dai, 2004]. Many heavily populated countries face a growing threat of severe and prolonged drought in coming decades [Dai, 2011]. So, it is very important to monitor drought effectively and predict its occurrence. Coping with droughts requires a deep understanding of different categories of drought, as well as linkages that may exist between the associated categories [Mishra et al., 2010; Mishra and Cherkauer., 2010]. Wilhite and Glantz [1985] identified four categories of drought: meteorological drought, agricultural drought, hydrological drought, and socio-economic drought. Among these four categories, hydrological drought received great amounts of attention because of its severe direct impacts to surface and underground water resources, along with indirect impacts to ecology, environment, navigation, tourism, water power, and so on. Identification of the relationship and possible time lag between hydrological drought and meteorological drought would be helpful for hydrological drought monitoring and early warning. Over the years, previous studies have developed a variety of indices to assess meteorological drought and hydrological drought. For the former, The standardized precipitation index (SPI) [McKee et al., 1993] is the most commonly used because of its robustness in flexible timescale, comparability in time and space, and sensitivity to meteorological drought [Hayes et al., 1999; Pai et al.,2010; Stricevic et al.,2011]. Research on streamflow drought as a form of hydrological drought has been established on the Runs Approach, which is also known as threshold method and was first applied to analyze drought events by Yevjevich [1967]. In recent years, some new indices were developed to analyze hydrological drought, in which the standardized runoff index (SRI) [Shukla and Wood, 2008; Vasiliades et al., 2010] has been widely used given its comparability in time and space, simplicity in computation and limited data requirements. A lot of studies have been focusing on individual drought categories whereas only a few studies have analyzed the relationships between meteorological drought and hydrological drought because of its complexity in underlying characteristics such as landcover, vegetation, topography, and other associated hydrologic/climatic variables [Mo, 2008; Mishra et al., 2010; Mishra and Cherkauer, 2010]. The time lag between them will provide strong support in forecasting and preparing for hydrological drought from views of meteorological drought, of which data is much easier to access. This issue has been addressed in some drought-prone regions in Europe and Africa [Hisdal and Tallaksen, 2003; Vicente-Serrano and López-Moreno, 2005; Edossa et al., 2010;]. However, it has not been carried out in northwestern China yet, which is an semi-arid region and also suffers drought a lot. Considering these issues above, the main objective of this research is to investigate associations between streamflow drought and meteorological drought of Jinghe River Basin in northwestern China. In this study, more efforts will be devoted to compare drought characteristics of these two categories and find out the possible time lag between them. 2 STUDY AREA As one of the top ten drainages in the Yellow River Basin, the Jinghe River Basin is located in the central part of the Loess Plateau in Northwestern China (Fig.1). This area is a typical inland river basin in a semi-arid and semi-humid region, together with a vulnerable climate and distinct inter-seasonal variations in climate and vegetation cover [Wei and Lin, 1996; Chen et al., 2007]. Droughts are very frequent in this area [Liang et al., 2005; Guo and Zha, 2009]. To carry out our analysis, precipitation and streamflow data were used. The daily precipitation data were obtained from 4 meteorological stations within the basin (Fig. 1), which were provided by the China Meteorological Administration (CMA, covering the period between 1960 and The daily streamflow data were obtained from the Zhangjiashan hydrometric station, which was located at the outlet of this drainage (Fig. 1). It was provided by the Institute of Soil and Water

3 Conservation (ISWC, Chinese Academy of Sciences & Ministry of Water Resources (CAS & MWR), covering the period between 1970 and Fig. 1 Map of study area with meteorological stations and hydrometric station 3 METHODOLOGY Evaluation of Meteorological Drought In this study, meteorological drought events were evaluated based on the SPI developed by McKee et al. [1993]. More detailed calculation procedures could be found in related references [McKee et al., 1993; Tabrizi et al., 2010]. 3.2 Evaluation of Streamflow Drought The SRI was utilized to indentify streamflow drought in this study [Shukla and Wood, 2008; Nalbantis and Tsakiris, 2009]. To match the monthly timescale of meteorological drought based on the SPI, a 30-day timescale was selected for characterizing streamflow drought based on the SRI. Detailed calculation procedures could be found in related references [Nalbantis and Tsakiris, 2009]. 3.3 Pooling Drought Events The selection of daily-recorded precipitation and streamflow data may result in the separation of adjacent drought events. During a prolonged dry period it is often observed that the flow exceeds the threshold level for a short period of time and thereby a large drought is divided into a number of minor droughts that are mutually dependent. Thus, a consistent definition of drought events should include some kind of pooling in order to define an independent sequence of droughts [Tallaksen et al., 1997]. In the study, the IC method was used which proved to be widely used [Tallaksen et al., 1997]. In this method, two adjacent drought events can be pooled if the number of days between the two droughts was less than the defined critical duration t c, and the ratio between the inter-event excess volume and the preceding deficit volume was less than a critical ratio. Detailed calculation procedures could be found in related references [Zelenhasic and Salvai, 1987]. 5 RESULTS 5.1 Identification of Meteorological Droughts To select the most suitable time scale of SPI, Pearson correlations between SRI and different timescales of SPI (1-12 months) have been conducted (figure omitted). It can be found that the increasing of time scale from 1-month to 4-month raises the correlation coefficient significantly. After 4-month, it goes almost steady till 8-month and then decreases. Thus, the 4-month is selected as the most suitable time scale of SPI to indentify the meteorological drought regarding of its relation with streamflow drought.

4 IC method is used to pool these minor meteorological drought events. Critical pooling parameters (t c and ) are selected through different frequency curves of meteorological drought events under different parameters combination (figure omitted). Minor meteorological drought events have been pooled under definite t c and (50 days and 0.4). The duration and the average intensity of every drought events identified by SPI 4 (4-month timescale of SPI) are illustrated in Figure 2. Fig. 2 Meteorological Drought events identified by SPI 4 during 1970 to 1990, (Left: raw; Right: after pooling under t c=50 days and =0.4) During 1970 to 1990, 123 raw meteorological droughts are pooled into 12 new meteorological droughts which are independent with each other. Statistic results of meteorological droughts after pooling are listed in Table 1, except the first one, of which the beginning is not clear within the limited period. In Table 1, magnitude is the sum of SPI 4 between beginning date and ending date. Intensity is the ratio between magnitude and duration. Table 1 Meteorological drought events indicated by SPI 4 Beginning Date Ending Date Duration (days) Magnitude Intensity ⅱ ⅲ ⅱ ⅲ ⅲ ⅰ ⅰ ⅰ ⅰ Average: (Top three of Duration, Magnitude and Intensity are in bold and labeled superscript ⅰ, ⅱ and ⅲ) 11 meteorological droughts are indicated by SPI 4 after pooling during 1970 to Its duration is between 76 and 713 days, with the average of days. The top three longest meteorological droughts lasted for 713, 522 and 514 days respectively. Its intensity is between and -0.74, with the average of -0.5 which indicates Mild drought according to Table 1. However, from the view of raw meteorological droughts in Figure 2 (left), the intensity of some drought events are on the grade of Moderate drought, Severe drought or Extreme drought. This is due to intensity is weakened by averaging the SPI value in pooling procedure. For the magnitude, it retains the information of both duration and intensity. The magnitude of meteorological droughts is between and , with the average of Although the magnitude can not indicate the grade of drought events absolutely, it can represent the relative severity of drought events. The top three most

5 severe meteorological droughts happened in 1986/2/10 to 1988/1/23, 1979/1/31 to 1980/6/27 and 1981/12/29 to 1983/4/ Identification of Streamflow Droughts Daily SRI of monthly time scale for the Zhangjiashan hydro-metric station is utilized to indicate streamflow drought events in the Jinghe River Basin. Being similar to meteorological drought, it need to pool the streamflow drought events by IC method. Minor streamflow drought events have been pooled under the definite parameter t c and (Figure 3). Fig. 3 Streamflow Drought events identified by SRI during 1970 to 1990 (Left: raw; Right: after pooling under t c=90 days and =0.5) From 1970 to 1990, there were 7 streamflow drought events indicated by SRI. However, the beginning time of the first drought event, which ended in the beginning of February 1970, was not clear. So, only 6 main streamflow drought events were finally detected from Figure 3. Statistic results of streamflow droughts after pooling are listed in Table 2. Table 2 Streamflow Drought events indicated by SRI Beginning Date Ending Date Duration (days) Magnitude Intensity i ⅰ ⅰ ⅱ ⅱ ⅱ ⅲ ⅲ ⅲ Average: (Top three of Duration, Magnitude and Intensity are in bold and labeled superscript ⅰ, ⅱ and ⅲ) After pooling, the duration of streamflow drought was between 102 and 1406 days with the average of days (Table 2). There were three dominant drought spells: January 1971 to November 1974, September 1977 to March 1983 and August 1986 to December The longest drought event, occurring from the beginning of January 1971 to the middle of November 1974, lasted for 1406 days. This event was also the most severe event with the biggest absolute value of magnitude (sum of the SRI is ) and greatest intensity (average of the SRI is -0.78). The second longest drought event happened from the middle of March 1979 to August 1981, lasting for 887 days with the average SRI of The third longest event occurred from the beginning of August 1986 to December 1988, lasting for 877 days with the average SRI of Streamflow drought s Response to Meteorological Drought

6 The response of streamflow drought to meteorological drought can be generally found comparing the respective graphs (Figure 2 and Figure 3). They have showed that the dominant meteorological drought events can lead to corresponding streamflow droughts. In this section, quantitative analyses of the responses will be undertaken by comparing drought characteristics in terms of occurring time, duration, magnitude and intensity (Figure 4). Fig. 4 Comparison of meteorological and streamflow drought events during 1970 to 1990, based on SPI 4 and SRI From Figure 4, it can be found that streamflow drought in this basin has coupled with meteorological drought quite well. All of the 6 streamflow droughts responded to the corresponding meteorological droughts. Three streamflow droughts were related with the only one meteorological drought. The other three streamflow droughts are related with two or three meteorological droughts. And the occurrence of streamflow droughts lagged that of meteorological drought. For a further coupling analysis, the relation between streamflow droughts and meteorological droughts are named Drought Event Pair in this study (Table 3). In Table 3, time lag of beginning is the lag days between streamflow drought and meteorological drought regarding of occurrence. Table 3 Drought Event Pairs in Jinghe River Basin during 1970 to 1990 Streamflow droughts Meteorological droughts Time Lag of Beginning Beginning Date Ending Date Beginning Ending No. Date Date (days) / / / / All 6 streamflow drought events corresponded to meteorological drought events. The top-three most severe streamflow drought events corresponded to two or three meteorological drought events. It also indicates that consecutive meteorological drought events probably have an essential impact on the lasting of streamflow drought. From Table 3, it can be found that there is an obvious time lag between the beginning of streamflow drought and meteorological drought. Every streamflow drought event happened behind the

7 corresponding meteorological droughts. The beginning time of 6 streamflow drought events lagged that of meteorological drought by 127 days on average. This result is in accordance with the field investigation in the basin that when a storm or heavy rainfall lasted, the river level rose quickly, however it dropped down slowly when rainfall ended. In other words, the streamflow drought happened much later, which may be due to the buffering function of the soil layers and groundwater system to the meteorological drought. However, it can end quickly after rainfall has mitigated meteorological drought situation. It also implies that different time scale factors should be considered carefully when making local drought and flood plans. 6 CONCLUSIONS This study used the SPI and SRI to detect meteorological and streamflow drought respectively, and conducted a comparative analysis of frequency, severity, beginning, ending and duration to indentify streamflow drought s response to meteorological drought. The main conclusions from this research are listed below: (1) IC method was well used in pooling drought events. After pooling under the critical parameter, there were 11 meteorological droughts indicated by SPI 4 and 6 streamflow droughts indicated by SRI in the Jinghe River Basin during 1970 to The most severe meteorological drought happened during the end of August 1971 to September 1973, which led to the most severe streamflow drought occurring from January 1971 to November (2) Streamflow drought indicated by SRI can be well related to meteorological drought indicated by SPI 4. As a comparison, the frequency of streamflow drought was less than meteorological drought. However, it lasted for a longer time on average. Streamflow drought was also more severe in drought magnitude. These findings agreed with the results carried out in the UK [Peter et al., 2005] and Denmark [Hisdal and Tallaksen, 2003]. (3) Streamflow drought in the Jinghe River Basin lagged meteorological drought by about 127 days. This conclusion in terms of time lag was consistent with the previous studies in other regions such as Spain [Vicente-Serrano et al., 2005], Africa [Edossa et al., 2010], and Iran [Tabrizi et al., 2010]. The time lag in this area was comparatively long, indicating that the Jinghe River Basin was a catchment mainly fed by groundwater. This paper was the first one to investigate the relationships between streamflow drought and meteorological drought in the Jinghe River Basin. Although the short time series of streamflow data had limited up-to-date information, the significant time lag would not only play an important role in understanding of relationships between different drought categories in theory, but also in streamflow drought early warning, prevention and regional water resources management in practice. Furthermore, how is the time lag affected by rainfall intensity, temperature, vegetation cover, underlying surface and anthropogenic activity? What is this impact s spatial pattern within the basin? These issues should be the next steps for the following research. ACKNOWLEDGEMENTS This research is supported by the National Natural Science Foundation of China (Grant No. NSFC ) and Program for New Century Excellent Talents in University (Grant No. NCET ). The authors specially thank the Group of Drought & Risk Governance from Beijing Normal University for their technical supports on this paper. We also would like to thank the editor and anonymous reviewers for their constructive comments and improvement of the manuscript. REFERENCES

8 1. Austin, R. B., C. Cantero-Martinez, J. L. Arrue, E. Playan, and P. Cano-Marcellan (1998), Yield-rainfall relationships in cereal cropping systems in the Ebro river valley of Spain, Eur. J. Agron., 8, Chen, C. C., G. D. Xie, and L. Zhen (2007), Characters of precipitation variation in Jinghe watershed (in Chinese), Resour. Sci., 29(2), Dai, A. G., K. E. Trenberth, and T. T. Qian (2004), A global dataset of Palmer Drought Severity Index for : relationship with soil moisture and effects of surface warming, J. Hydrometeorol., 5(6), Dai, A. G. (2011), Drought under global warming: a review, WIREs: Clim. Change, 2, 45-65, DOI: /wcc De Gaetano, A. T. (1999), A temporal comparison of drought impacts and responses in the New York city metropolitan area, Climatic Change, 42, Edossa D. C., M. S. Babel, and A. D. Gupta (2010), Drought analysis in the Awash river basin, Ethiopia, Water Resour. Manag., 24, Guo R., and X. C. Zha (2009), Analysis on the change law of drought and flood disaster in Jinghe basin from 1470 to 1979 (in Chinese), J. Shaanxi. Normal University (Natural Science Edition), 37(3), Hayes M. J., M. D. Svoboda, D. A. Wilhite et al. (1999), Monitoring the 1996 Drought Using the Standardized Precipitation Index, Bull. Amer. Meteor. Soc., 80(3), Hisdal H., and L. M. Tallaksen (2003), Estimation of regional meteorological and hydrological drought characteristics: a case study for Denmark, J. Hydrol. 281, Liang Y., Min Q. W., and Chen S. K. (2005), Concerted development in the eco-environment and economy of the region covered by the headwaters of the Jinghe river (in Chinese), Agr. Res. Arid Area, 3(2), McKee T. B., N. J. Doesken, and J. Kleist (1993), The relationship of drought frequency and duration to time scales, Preprints, 8th Conference on Applied Climatology, January, Anaheim, CA, USA, McKee T. B., N. J. Doesken, and J. Kleist (1995), Drought monitoring with multiple time scales, Preprints, 9th Conference on Applied Climatology, January, Dallas, Texas, USA, Mishra V. and K. A. Cherkauer (2010), Retrospective droughts in the crop growing season: Implications to corn and soybean yield in the Midwestern United States, Agric. For. Meteor., 150, Mishra V., K. A. Cherkauer, and S. Shukla (2010), Assessment of Drought due to Historic Climate Variability and Projected Future Climate Change in the Midwestern United States, J. Hydrometeor., 11(1), 46-68, doi: /2009JHM Mo K. C. (2008), Model-based drought indices over the United States. J. Hydrometeor., 9, Nalbantis I., and G. Tsakiris (2009), Assessment of hydrological drought revisited, Water Resour. Manag., 23, Nicholson S. E., C. J. Tucker, and M. B. Ba (1998), Desertification, drought and surface vegetation: an example from the west African Sahel, Bull. Amer. Meteor. Soc. 79: Pai D. S., L. Sridhar, P. Guhathakurta, and H. R. Hatwar (2010), District-wide drought climatology of the southwest monsoon season over India based on standardized precipitation index (SPI), Nat. Hazards 59(3), , DOI: /s Pausas J. G. (2004), Changes in fire and climate in the eastern Iberian Peninsula (Mediterranean basin), Climatic Change, 63, Peter E., H. A. J. van Lanen, P. J. J. F. Torfs, et al. (2005), Drought in groundwater drought distribution and performance indicators, J. Hydrol., 306, Shukla S., A. W. Wood (2008), Use of a standardized runoff index for characterizing hydrologic drought, Geophys. Res. Lett., 35, L02405, DOI: /2007GL

9 22. Stricevic R., N. Djurovic, and Z. Djurovic (2011), Drought classification in Northern Serbia based on SPI and statistical pattern recognition, Meteorol. Appl., 18(1), 60-69, DOI: /met Tabrizi A. A., D. Khalili, and A. A. K. Haghighi (2010), Utilization of time-based meteorological droughts to investigate occurrence of streamflow droughts, Water Resour. Manag., 24, Tallaksen L. M., H. Madsen, and B. Clauson (1997), On the definition and modeling of streamflow drought duration and deficit volume, Hydrol. Sci. J., 42(1), Vasiliades L., A. Loukas, and N. Liberis (2010), A water balance derived drought index for Pinios River Basin, Greece, Water Resour. Manag., 25, Vicente-Serrano S. M., and J. I. López-Moreno (2005), Hydrological response to different time scales of climatological drought: an evaluation of the Standardized Precipitation Index in a mountainous Mediterranean basin, Hydrol. Earth Syst. Sci., 9, Wei Z. X., and S. R. Lin (1996), Analysis on the hydrological characteristics in Jinghe basin (in Chinese), Hydrol., 2, Wilhite D. A., and M. H. Glantz (1985), Understanding the drought phenomenon: the role of definitions, Water Int., 10(3), Wilhite D. A. (2000), Drought as a natural hazard: concepts and definitions, In: Wilhite D A, ed. Drought: A Global Assessment, London &. New York: Routledge. 30. Yevjevich V. (1967), An objective approach to definition and investigations of continental hydrologic droughts, Hydrology Papers 23, Colorado State University, Fort Collins, USA. 31. Zelenhasic E., and A. Salvai (1987), A method of streamflow drought analysis, Water Resour. Res., 23(1),