The relationship of Streamflow-Precipitation-Temperature in the Yellow River Basin of China during 1961-2000 Academic research paper on "Materials engineering"

Available online at www.sciencedirect.com

SciVerse ScienceDirect PfOCSCliQ

Environmental Sciences

Procedia Environmental Sciences 13 (2012) 2336 - 2345 ^^^^^^^^^^^^^^^^^^^

The 18th Biennial Conference of International Society for Ecological Modelling

The relationship of Streamflow-Precipitation-Temperature in the

Yellow River Basin of China during 1961-2000

Z.F. Yang , Y. Yan, Q. Liu

State Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, Beijing 100875, China

Abstract

Global climate change has tremendously altered the spatial and temporal distributions of precipitation. The Streamflow of the Yellow River decreased in the past decades, both because of human consumption and climate variations. This paper analyzed the trends of natural and observed streamflow of the Lijin hydrological station, as well as the precipitation and air temperature of the whole Yellow River Basin. There was an obvious change in the natural streamflow in 1991 when analyzed by the Mann-Kendall (M-K) method. Precipitation decreased in the whole Yellow River Basin. 22 Meteorological stations have slight increases, while the other 67 stations measured declines, in which 47 of the Meteorological stations' decrements were over 5%. There was a descent of temperature for only 2 stations, while the temperature rose for the other 87 stations. The average temperature increment was 0.82 °C and the highest was 1.91 °C in Tsinghai Center station. The quantitative Streamflow-precipitation-temperature relations plot was drawn by Geostatistical Analyst module of ArcGIS 9.3. The streamflow have various responses to the different extents of precipitation and air temperature. In the past four decades, precipitation elasticity to streamflow was 1.95, and the elasticity index may give a general relationship between precipitation and streamflow. However, it varies with different scenarios.

© 2011 Published by Elsevier B.V Selection and/or peer-review under responsibility of School of Environment, Beij ing Normal University.

Keywords: Mann-Kendall; precipitation elasticity; streamflow-precipitation-temperature; the Yellow River

1. Introduction

Global climate changes have tremendously altered the spatial and temporal distributions of precipitation. The streamflow of the Yellow River has decreased in the past decades, both because of human consumption and climate variations. With the population growth and economic development of

* Z.F. Yang. Tel.: +86-10-58807951; fax: +86-10-58803006.

E-mail address: zfyang@bnu.edu.cn

1878-0296 © 2011 Published by Elsevier B.V. Selection and/or peer-review under responsibility of School of Environment, Beijing Normal University. doi:10.1016/j.proenv.2012.01.222

the Yellow River Basin, the consumed runoff grew considerably, from about 7.4 billion m3 in 1949, to 27.7 billion m3 in 1980, and the average annual consumption was 30.8 billion m3 in 1988-1992 [1]. In addition, the natural streamflow decreased from 1961-2000. The average precipitation in the 1990s was about 8.8% less than that in the 1950s [2].

The streamflow was sensitive to the climate changes, such as precipitation and air temperature variations. Therefore, climate change is an influential factor to the streamflow. In order to explore the relationships between streamflow and climate change, Schaake [3] established elasticity to evaluate the sensitivity of streamflow response. Fu et al. [4] employed the Geostatistical Analyst module of ArcGIS, to interpolate the observed data of the above variables, so as to quantify the relationships of streamflow-precipitation-temperature.

Previous studies on the response of streamflow to climate change mainly focused on the whole basin or sub-basin, and did not reveal the accumulated impact of the lower reaches. The aim of this study was to explore the impact of climate change at the basin scale, such as precipitation and air temperature, to the streamflow of Lijin hydrological station, and furthermore, to identify the influential factors of hydrological alterations.

2. Data and methods

2.1. Data preparation

Data on the monthly precipitation and monthly means of the daily mean temperature were collected from 89 meteorological stations along the Yellow River, which were supplied by National Meteorological Information Center (Fig. 1.). The study period was from 1961 to 2000. The daily streamflow were used to investigate the variations of flow regime in the Lijin station (Fig. 1). The Lijin station is the last station before the Yellow River reaches the Bohai Sea, and is located in the Yellow River Delta, approximately 100 km from the sea. The observed flow data in Lijin collected from 1961 to 2000 was provided by the Institute of Geographic Sciences and Natural Resources Research, CAS.

The concept of 'natural runoff generally refers to the runoff produced under meteorological and geographical physical conditions, such as hydrological, geomorphologic, geologic, vegetation, soil, and agricultural characteristics, and is different from the runoff observed at the gauge stations. The difference between observed runoff and natural runoff generally results from three factors: (a) the amount of water directly extracted from the river channel for irrigation, industry, and domestic usages, and the amount returning to the downstream river channel after usages; (b) the amount of water controlled by dams, including extra water losses through evaporation and seepage due to dams; and (c) the amount of water transported into and out of the watershed [1]. In this study, the natural flow data were used for all analysis except for observed flow shown in Fig. 2.

Fig. 1. Location of Meteorological stations and Lijin gauge station of Yellow River basin. 2.2. Methodology

2.2.1. The Mann-Kendall (M-K method)

The Mann-Kendall method is a useful tool to identify the obvious hydrological change over a period of time [5]. In this study, the M-K method with a nominal rejection rate of 5% was used to reveal the temporal trends for more accurate results of annual streamflow. The year in which the abrupt change

occurred was considered to be the separating year for the IHA calculation. The test statistics a'3 and dk are defined as:

(1 < dj < dr )

dk (k=2,3,4,...,n)

i=i j=i

where the time points are xl5 x2, ..., xn, and n is the total number of data in the time range. The expected value E ( dk ) and variance Var (dk ) are calculated as follows:

k ( k -1)

E d ) = -

Var(dk ) =

k ( k -1)( 2k + 5 )

The null hypothesis of no trend is tested by UF(dk) , and if the standard normal probability \UF (dk)| > ua is true, the null hypothesis of no trend will be rejected. The UF(dk) constitute curve Ci.

UF(dk) = d\ ~ E(dk) (5)

k Vrd)

The corresponding rank series for retrograde rows are similarly obtained for the retrograde sample (x„, x„_i, . ., Xj). Following the same procedure as shown in Eqs. (1) - (5), the statistic vector UF \dk)

can be calculated for the retrograde sample. The statistic variable UB(dk) is defined as the following, which constitute curve C2.

¡UB(dk) = -UF K.) (k 12 ) (6)

Ik' = n + 1-k (k=1,2,...,n) (6)

If the intersection point of the C1 and C2 is between the two confidence lines (P = a ), we can consider that the abrupt change took place at that point.

2.2.2. Sensitivity of streamflow to climate changes

Climate elasticity of streamflow may be defined as the proportional change in streamflow, Q, to the change in a climatic variable such as precipitation, P. Thus the precipitation elasticity of streamflow is defined as [3]:

, n ^ dQ/Q dQ P eCP,Q) = ,^ — (7)

dP/P dPQ

However, this elasticity is often estimated from a hydrological model and, of course, the form of the hydrological model is always unknown and validation of such a model remains a fundamental challenge [6]. Differences in model calibrations lead to differences in estimates of the model parameter which result in differences in the models' sensitivity to climatic variations. Sankarasubramanian and Vogel further verified that formula [7]:

e„= median

fQt - QP^

_ _ V P ~PQ,

where t represent the time, P and Q mean the average precipitation and streamflow, respectively. In this study, this formula was employed.

3. Results and analysis

3.1. The trends of streamflow in Lijin gauge

This paper analyzed the impact of climate on the natural streamflow, using the monthly precipitation, monthly means of the daily mean temperature and the natural streamflow of adjacent hydrological gauge. The observed streamflow decreased in the past decades (Fig. 2.), and the differences between the natural and observed streamflow expanded. The decreasing rate of natural streamflow was slower than the observed streamflow. This indicates the consumption of freshwater was increasing by human beings. Meanwhile, the decline of the natural streamflow shows that climate changes, such as the spatial and temporal distribution of precipitation variation and climate warming, were influential factors to the natural streamflow of the Yellow River.

Fig. 2. The natural and observed annual streamflow of the Yellow River at Lijin station (1961-2000).

Annual streamflow, revealed by the M-K method (Eqs. 1 to 6), presented an abrupt downward change in 1991 (a =0.05) (Fig. 3). According to the time of abrupt change, the streamflow were divided into two periods during 1958-2006 (1961-1990 and 1991-2000). Before 1991, the annual average natural streamflow was 61.3 billion m3, while after 1991, the annual average natural streamflow was only 41.6 billion m3 (Fig. 2).

Fig. 3. The abrupt analysis of Lijin natural streamflow (1961-2000).

3.2. The impact of precipitation and temperature on the streamflow

3.2.1. The variations of precipitation and temperature

Since there was an abrupt change in streamflow, 1991 was selected as the separating year to study the variations of precipitation and temperature. The precipitation and the temperature changes were evaluated by the following formula:

AP = Ppost ~ Ppre x100%

AT = T - T

post pre

where Ppre and Ppost represent the average annual precipitation before and after 1991, respectively; Tpre and Tpost represent the annual means of monthly temperature. The spatial distributions of

precipitation and temperature variations were shown in Fig. 4 and Fig. 5, employing the software ArcGIS 9.3.

Fig. 4. The precipitation changes from the baseline period to the period of change in the YRB.

Fig. 5. The air temperature changes in the YRB from the baseline period to the period of change.

For precipitation, the overall trend was decreasing; 22 Meteorological stations have slight increases, while the other 67 stations declined in precipitation. Of these, 47 Meteorological stations' decrements were over 5%, in which 19 Meteorological stations' decrements were over 10% and 28 stations decrements were between 5%-10%. This means 53% of 89 stations (more than a half) have obviously declined precipitations.

For air temperature, there was a descent of only two stations, while the temperature rose for the other 87 stations, with an average increment of 0.82 °C. There were 26 stations with a rise in temperature of over 1 °C, accounting for 29% of the 89 observed stations. Linhe station in Inner Mongolia and Tsinghai Center station had increases in temperature of 1.75 °C and 1.91 °C, respectively. There were 14 stations with the temperature increments between 0.85-1.0 °C. These are stations mainly distributed in the Middle Yellow River Basin.

3.2.2. Streamflow-precipitation-temperature relations

The quantitative Streamflow-precipitation-temperature relations plot was drawn by Geostatistical Analyst module of ArcGIS 9.3. The general relationship of the streamflow response to changes in precipitation and temperature is clear from Fig. 6. Streamflow was positively related to precipitation but negatively related to temperature, with the streamflow-streamflow sensitivity greater than the precipitation-temperature sensitivity. For example, a 10% precipitation increase resulted in an 11.2% increase in streamflow, and a 10% precipitation decrease resulted in an 11% decrease in streamflow. When temperature increased 1.0 °C, the streamflow would decrease 18% (Fig. 6).

Fig. 6. Contour plot of percentage annual streamflow change as a function of annual percentage precipitation and temperature change for the Yellow River in the Lijin station, China.

Precipitation changc(%)

The responses of the streamflow to the changes in precipitation were not symmetric. This difference is highlighted in Fig. 7. If the increment of precipitation exceeded 10%, the value of 'streamflow increment minus precipitation increment' would enlarge, indicating that the rate of the streamflow increase was larger than the rate of precipitation. When the precipitation change was between 0-10%, the increment of precipitation and streamflow were almost the same. And when precipitation variation was between -10%-0, the decrement rate of streamflow was faster than that of precipitation.

-30 -20 -10 0 10 20 30 40 50 Precipitation change(%)

Fig. 7. Average streamflow change minus precipitation change as a function of precipitation change in YRB.

3.3. Elasticity of streamflow response to precipitation

The quantified streamflow-precipitation-temperature relationship (Fig. 6) is used as an input to equation (8) to calculate the climate elasticity of streamflow index in different precipitation-temperature scenarios. The streamflow has various responses to the different extents of precipitation and air temperature. In the past four decades, precipitation elasticity to streamflow was 1.95, and the elasticity index may give a general relationship between precipitation and streamflow. However, the elasticity could not reflect the complex non-linear relationships of streamflow-precipitation-temperature.

4. Conclusions

Comparing the time period trend of natural and observed streamflow, it can be concluded that human consumption of freshwater was increasing, and the climate change, precipitation decrease and temperature increase are all influential factors to the streamflow decline of the Yellow River. There was an obvious change in the natural streamflow in 1991 as analyzed by Mann-Kendall (M-K) method. We studied the variations of precipitation and air temperature before and after 1991, and the following conclusions can be drawn.

Precipitation is decreasing in the whole Yellow River Basin. 22 Meteorological stations have slight increases, while the other 67 stations showed decline, in which 47 Meteorological stations' decrements were over 5%. There was a descent of only 2 stations, while the temperature rose for the other 87 stations. The average temperature increment was 0.82 °C and the highest was 1.91 °C in Tsinghai Center station.

The quantitative Streamflow-precipitation-temperature relations plot were drawn by Geostatistical Analyst module of ArcGIS 9.3, using precipitation and air temperature data of the Yellow River Basin and the natural streamflow of the Lijin hydrological gauge station. The streamflow have various responses to the different extents of precipitation and air temperature. In the past 40 years, the elasticity index for precipitation elasticity to stream-flow was 1.95, which indicates the relationship between precipitation and streamflow. However, elasticity varies with different scenarios.

Acknowledgements

This study was funded by the National Natural Science Foundation of China (No.50939001), National Basic Research Program of China (No. 2010CB951104), and the Fundamental Research Funds for the Central Universities.

References

[1] Fu, G.B., Charles, S.P., Viney, N.R., Chen, S.L., Wu, J.Q.. Impacts of climate variability on stream-flow in the yellow river. Hydrol Process 2007; 21: 3431-39.

[2] Wang, H., Yang, Z., Saito, Y., Liu, J.P., Sun, X.. Interannual and seasonal variation of the huanghe (yellow river) water discharge over the past 50 years: Connections to impacts from enso events and dams. Global Planet Change 2006; 50(3-4): 212-25.

[3] Schaake, J.. From climate to flow. Climate change and us water resources. P. Waggoner. New York, John Wiley 1990; 177206.

[4] Fu, G.B., Chen, S.L.. Geostatistical analysis of observed streamflow and its response to precipitation and temperature changes in the yellow river. IAHS Publication 2005; 296: 238-45.

[5] Chen, H., Guo, S., Xu, C., Singh, V.. Historical temporal trends of hydro-climatic variables and runoff response to climate variability and their relevance in water resource management in the hanjiang basin. JHydrol 2007; 344(3-4): 171-84.

[6] Fu, G.B., Charles, S.P., Chiew, F.H.S.. A two-parameter climate elasticity of streamflow index to assess climate change effects on annual streamflow. Water Resour Res 2007; 43, W11419,doi:11410.11029/12007WR005890.

[7] Sankarasubramanian, A., Vogel, R.. Climate elasticity of streamflow in the United States. Water Resour Res 2001: 37: 177181.