Wang Guoqing ^1，2 ,Shi Zhonghai³ ,Sun Zhengqin⁴ and Guo Baoqun⁴

 (1. Department of Water Resources and Environment, Hohai University, Nanjing, 200098)

(2. Yellow River Institute of Hydraulic Research, Zhengzhou, 450003)

(3. Zhengzhou Environmental Protection Bureau, Zhengzhou, 450007)

(4. Hydrology Bureau, Zhengzhou, 450004)

Abstract: Structure and principle of the SIMHYD rainfall runoff model were introduced, and the model was used to simulate discharge of Qingjianhe River basin. The results show that SIMHYD model has good simulation result for monthly discharge, and Nash-Sutcliffe efficiency coefficients are larger than65%, meanwhile, relative errors are less than 5%. But the model can not simulate daily discharge well. Therefore, the model can be used to analyze runoff variation in a larger time scale.

Key words: SIMHYD rainfall runoff model, Qingjianhe River Basin, Discharge simulation, Application

1. Introduction

Hydrologic model is mathematical description for catchments hydrology cycle based on cognition of hydrologic laws. At present, hydrologic models have been widely used in flow forecasting, runoff regulation, water resources allocation, etl; and have become an important tool in the modern basin administration. With the construction of digital Yellow River, hydrologic model will become an essential way to maintain health life of Yellow River scientifically. 

Development of hydrologic simulation technique has history of more than half century, and hundreds hydrologic models have been developed in worldwide. In recent years, the rapid development of application techniques of computer and 3S has promoted the progress of distributed hydrologic model. With the limitation of cognization for hydrologic laws and its demand for a large quantity of hydrologic and meteorological data, this kind of hydrologic model has not gotten to degree of ready-to-use, although it’s a promising model and has bright prospects in application. On the contrary, the lumped hydrologic model has merits of more flexible, less data requirement, and this kind of model still acts an important role in hydrologic analysis and environment effect assessment.

Yellow River is the most complex river in the world partly due to its complex laws of runoff generation and concentration, and easily changeable hydrologic phenomena. And hydrologic simulation of Yellow River Basin has been becoming a difficult problem facing before hydrologists for a long time. With data of typical tributaries of Yellow River, to verify some hydrologic models which have been successfully used in other river basins could supply some reference experiences for construction of digital Yellow River, and it’s also be of importance to utilize and develop water resources of Yellow River.

2. General Information of Qingjianhe River Basin

Qingjianhe River is first class tributary of middle Yellow River. It originates from Ansai county of Shanxi Province, and runs through 4 counties of Ansai, Zichang, Qingjian, and Yanchuan County from west to east, and empties itself into Yellow River at Tugang Village of Yanchuan County. The total drainage area of Qingjianhe River Basin is about 4080 km2 with total river length of 167.8 km. There are two hydrology gauging stations of Zichang and Yanchuan in the river basin, both of which were built in 1950s, and drainage areas above the two gauging stations are 913 km² and 3468 km² respectively.

Qingjianhe River Basin situates in Eastern-Asia monsoon climate district, its climatic condition is arid and low precipitation. Based on statistic for data from 1955-2002, Multi-year average annual precipitation is about 470mm, and maximum annual precipitation is as much as 2.9 time of minimum precipitation, rainfall in flood season from June to September is more than 70% of annual value. Seasonal distribution of runoff is also uneven, and its inter-annual variation is quite larger. Mean annual runoff is about 1.462 million cubic meters, and 67% of annual runoff occurred in the flood season, maximum annual runoff is 4.4 times of minimum value.

To prevent losing of water and sediment, large-scale soil and water conservation measures have been taken in the basin since 1970s. Till to 19999, 5 reservoirs have been built, and 17970hm² of terrace, 83853hm² of reforestation area and 11206hm² of grassland have been constructed, and the formed check dam land was up to2731 hm². These human activities have impacted hydrological process greatly.

3. SIMHYD Rainfall Runoff Model

SIMHYD is a simplified conceptual rainfall-runoff model that estimates daily or monthly stream flow from daily rainfall and areal potential evapotranspiration data (Chiew, et al, 2002). The model has 7 parameters, and has already been applied in some semiarid or humid basins which are located in America, Australia, and other countries. The structure of SIMHYD model is shown in Figure 1. And calculation principle was introduced below.

Fig.1  the structure of SIMHYD model

In SIMHYD, rainfall first fills the interception store, which is depleted by evaporation subject to potential evapotranspiration rate. The excess rainfall is then subjected to an infiltration function that determines the infiltration capacity. The excess rainfall that exceeds the infiltration capacity becomes infiltration excess runoff. Moisture that infiltrates is subjected to a soil moisture function that diverts the water to the interflow, groundwater recharge and soil moisture store recharge.

Evapotranspiration Loss includes two parts which are from interception of vegetation cover and soil moisture storage. The evapotranspiration from interception is calculated by potential evaporation rate, and the second part is estimated as a linear function of the soil wetness, but cannot exceed the atmospherically controlled rate of areal potential evapotranspiration. The evapotranspiration calculation formulas were given below:

Where, ET1 is evapotranspiration from interception of vegetation cover, ET is evapotranspiration from soil moisture store, INS is interception capacity, PET is potential evapotranspiration rate which is measured by E601 evaporimeter, SMS is soil moisture store, SMSC is soil moisture capacity, and POT is remaining potential evaporation capacity.

Infiltration is the core part of model calculation, which assumes the infiltration rate is negative exponential function of soil wetness, and actual infiltration calculation formulas were given below:

Where, INF is infiltration rate, COEFF is maximum infiltration loss, SQ is infiltration loss exponent, INR is net rainfall, RAIN is rainfall in the calculation interval, INSC is interception store capacity, INS is interception amount which calculated by water balance principle, RMO is actual infiltration amount.

The model divided runoff into three components of surface flow, interflow, and base flow. The surface flow is infiltration excess runoff. And Interflow is first estimated as a linear function of the soil wetness. The equation used to simulate interflow therefore attempts to mimic both the interflow and saturation excess runoff processes. Base flow from the groundwater store is simulated as a linear recession from the store. And calculation formulas were given in the following:

Where, IRUN, SRUN AND BAS are surface flow, interflow, and base flow respectively. RUNOFF is simulated runoff. SUB is constant of proportionality in interflow calculation, K is base flow linear recession parameter, and GW is groundwater store.

There are three kinds of state variables of surface water store, soil moisture store, and ground water store. And soil moisture store is an important one which is close related to calculation for interflow and groundwater recharge. The three state variables are calculated by water balance. And recharges for soil moisture store and groundwater are calculated by the following formulas:

Where, REC is recharge of groundwater store, SMF is recharge of soil moisture, and CRAK is constant of proportionality in groundwater recharge calculation.

3. Application of SIMHYD Rainfall Runoff Model in Qingjianhe River Basin

Two study basins which are controlled by gauging stations of Zichang and Yangchuan were selected. The characteristic values of the two study basins were given in table1. As Qingjianhe River Basin had less human activities which have main function of water and soil conservation before 1970s, 1970 was chose as a critical year in the past studies. For conveniently comparing with the past study results, data in period before 1970 were used to calibrate model in the study.

Choose Nash efficiency coefficient R², average relative error Re as objective functions to calibrate model parameters. Better simulation result will have R² close to 1 while Re close to 0.

Initial values of state variables were usually given artificially, to eliminate the artificial influence to simulation result, the first one year or two year was taken as warm-up period, and the left data was used to calibrate model. Discharge simulation results were given in table1.

The result in table1 show that, (1) SIMHYD model has good simulation result for monthly discharge, Nash-Sutcliffe efficiency coefficients of two gauging stations are above 65%, and relative errors are less than 5%. (2) The model has poor simulation result for daily discharge, and Nash-Sutcliffe efficiency coefficients of two gauging station are less than 55%, and that of Yanchuan station is just 22.4%, meantime, relative errors are larger than 10%. (3)Daily discharge simulation result for Zichang station is much better than that of Yanchuan station, while monthly discharge simulation results of the two gauging stations are similar.

Table1 discharge simulation result for two gauging stations of Qingjianhe River Basin

For virtually, comparison result of monthly simulated and recorded discharge of Qingjianhe River Basin were plotted in the figure2. The figure indicates that simulated discharge is close to the recorded, and correlation coefficients are above 0.7. All points distributed uniformly about the 1:1 line. Simulation errors for high discharge of Yanchuan station are quite large. Although the model can simulate low flow of Zichang station well, middle and high flows of the station are underestimated.

Fig.2 comparison between monthly recorded and simulated discharge

of Qingjianhe River Basin

Due to basin hydrology was influenced by some uncertain factors, long-interval (month) data can eliminate the stochastic effect to some extent, and this leads to better simulation result for monthly discharge, while poorer result for daily charge simulation. In addition, as lumped hydrologic model didn’t consider the influence of uneven spatial distribution of precipitation on flow yielding and concentration, that’s why hydrologic model can get better discharge simulation result for small-scale watershed, and poor result for large-scale basin.

4. Conclusion and Discussion

Hydrologic model is one of impotant content of digital basin construction. And application result of SIMMHYD model in Qingjianhe River basin indicate, (1) SIMHYD model has good simulation result for monthly discharge, meanwhile, poor simulation result for daily discharge. Therefore, the model can be applied to analyze variation of hydrology in larger time scale. (2) SIMHYD model has characteristics of simple structure of fewer parameters. And mechanisms of infiltration excess runoff and saturation excess runoff were considered in the model, most river basins can meet data demands of the model, therefore, the model has wide prospects of application. (3) As the model hasn’t considered effect of human activities on hydrology, therefore, consistent of hydrologic data should be analyzed firstly before model application, otherwise, it’s difficulty to get good simulation result.

Yellow River is main water source of North-western China and Northern China. As lower precipitation and higher evapotranspiration in the basin, water resources are in shortage relatively. To enhance application of hydrologic models in Yellow River Basin can supply scientific support for water resources planning, and it’s significant for development and sustainable utilization of water resources.

References

C.Loumagne, M.Normand, M.Riffard, 2001. Integration of remote sensing data into hydrological models for reservoir management [J], Hydrological Sciences Journal, v46, 89-101

Chen Xiande. Yellow River Hydrology [M]. Yellow River Press, Zhengzhou, 1996.

Chiew, F.H.S., Peel, M.C. and Western, A.W. 2002. Application and testing of the simple rainfall-runoff model SIMHYD. In: Mathematical Models of Watershed Hydrology, Water Resources Publication, Littleton. Colorado.

Nash J.E. and Sutcliffe J. 1970. River flow forecasting through conceptual models, Part 1, A discussion of principles, Journal of Hydrology, 10, 282    

Source:  www.yellowriver.gov.cn   Editor:HuangFeng