Tang Jie  Bian Jianmin  Lin Nianfeng  Yang Jianqiang

(College of Environment and Resources, Jilin University, Changchun 130026, China)

Abstract: Groundwater table changes might exert on the environment, the coupling technology with GIS and PModflow had been used to define the warning threshold of the groundwater environment early warning in the paper. By the means, the phreatic water table with no-salinization had been defined to be the maximum limit, and the pumping limit of the phreatic water been defined as the minimum limit. On the basis of system model generalization, water table had been simulated and predicted. The spatial analysis of GIS was used to overlay the grid and attributes judgment, and the water environment early warning had been finished through the comparison of the predicted water and the water table warning limit. Results showed that the early warning state in 2015a was heavier than 1999a, and the secondary salinization was the main case due to the phreatic water table uplift.

Keywords:GIS; PModflow; phreatic water table; water environment early warning; numerical simulation; integrated system

1 Foreword

The concept of early warning was initially applied to radar technique and missile defense system in military affairs, and was widely used in flood forecast, economy system, weather and environment engineering disaster prevention and so on recently. But the research and application in the field of resources and environmental science was relatively less, and it was seldom found in the study literature on water resource environment early warning at present^[1-4].

The western Songnen plain was wide, and belonged to semi-arid and semi-humid continental monsoonal climate. The average annual rainfall was 400-500mm, of which the rain from June to September occupied the 74-84 percent. The average annual evaporation increased from the east to the west, which was from 1400 to 2100mm. There were abundant resources of water, soil, living things and oil in the area, and it was important base of agriculture, herd and energy sources. There were few rivers in the area which distributed unevenly, and the noncontributing area was large. There was lack of surface water in the area and relatively rich in the groundwater. In recent years, reservoirs built in the upstream intercepted the streamflow. The capacity of Tao’er River, Huolin River and Jiaoliu River decreased even resulted in the dry strand, and the groundwater recharge reduced as well. Under the double influence of global warming and intense exploitation, the eco-environment deteriorated rapidly, and it was serious that the short of water resources, soil salinization and degradation. Drought, soil salinization and regional water level depression had been the main problems in the area. In the recent years, Chinese Academy of Sciences, colleges and manufacture branches have carried out a great deal of scientific researches and productive practices, which offered to the reasonable exploitation of water resource and the integrate remedy of the eco-environment. However, it was the first time in studying the water resources early warning from the point of view of water level amplitude.

It was adopted that the GIS-Modflow integrated system in the paper^[5,6]. By the generalization of the hydro-geology condition, the numerical model about the phreatic water flow was built up. As to the P-Modflow software and Method of Finite Difference (MFD), numerical simulation and prediction was done on phreatic water flow to open out the transport mechanism and forecast it. On the GIS flat roof, through the comparing operation of water level status in quo to guard line, based on which, early warning was done to the water environment.

2 Early warning threshold determination

(1)

There is high in east, south and west, and low in north in the area, which made it a water-bearing basin like a dustpan. The ground water had a wide gathering area, and flush recharge sources, which include precipitation and Huolin River backward dispersion, and the lateral recharge of mountain groundwater and river. There were groundwater lateral recharges in the northwest mountainous area, Changling-Songliao watershed, and high plain. The groundwater was recharged along the run off of Tao’er River and Jiaoliu River in the middle reaches and upstream area, and was supplied from partial surface water of Songhua River and Nen River at flood season as well.

(2)

 

Warning limit confirmation   Key to water environment early warning was to determine the warning limit. Whereas the phreatic water was nearly contacted to environment, the phreatic water level amplitude was chosen to be the main study object. The phreatic water table with no-salinization had been defined to be the maximum limit, and the pumping limit of the phreatic water been defined as the minimum limit. By the long term observation and integrated study, the deepest exploitation of the phreatic water was half of the thickness of it from the surface, and any exploitation exceed would result in the depletion of the water resources.

The western of Songnen plain was wide, with complex and varied land form styles, soil structure and hydro-geology conditions, it was quite difficult to obtain the phreatic critical depth in the whole area. Therefore, the concept of reasonable phreatic depth was brought forward for protecting salt accumulation in soil. That is, the phreatic water level should be within the changing range to control the soil salinization in a relatively low value. Based on the phreatic water level in 1996, geodynamics model for inversion^[9] was used to solve the reasonable phreatic water level.

(1)

Based on the geodynamics theory, to confirm the influence of the nature and human being to the groundwater system, it should be considered as followings, the increase of the output factors of the groundwater dynamics was produced by the increase of factors input and the artificial influence to the groundwater system, where the increase may be negative.

Where: Y-value of the phreatic water level, -changes of the phreatic water level;

M-the influence degree of human activities, soil salinization degree in the paper, -contribution degree of the human activities;

K₁ - influence intensity coefficient.

                                       

(2)

Based on the data of phreatic water level in 1983 and 1996, and the RS interpreted data of soil salinization in 1983 and 1995, using the overlay and analysis functions in GIS, the grid maps were formed that the influnce intensity of phreatic water level to soil salinization by the overlay of the two maps under nature and human activities (omited). By the overlay analysis and the calculation of the maps of soil styles, soil salinization, and groundwater phreatic water level, it was discovered that as to the influence coefficient of phreatic water level to soil salinization, the area of more than 0.5 was 10.28% of the whole calculated area, and 0.3-0.5 was 12.77%, where there was the most serious salinization. Supposed that the increase of soil salinization was controlled at 1%, that is, the soil salinization of each unit was increased by 1%, the relation between phreatic water level and soil salinization was as following formula.

That is to say, the units reached 50% or 30% of all that the influence intensity coefficient with more than 0.5-0.3.

Where:

H₁：phreatic water level in 1996

H₂：phreatic water level in 2015

M₁：soil salinization in 1996

M₂ ：soil salinization in 2015

If H₁, M₂ and M₁were known, H₂ could be calculated from the formula (2).

According to the analysis of the RS interpret data of the area, the soil salinization increased at rate of 2.61%/a in the twelve years of 1989-2001. To protect the eco-environment and hold down the development of soil salinization, the aggravated speed should be kept at rate of 1% in the area, and the H₂ was the controlled phreatic water level in 2015 under the condition. Based on H₂, it was drew that the early warning maximum limit map using GIS based on the H₂ value in the units, shown as fig. 1.

3 Water table forecast model foundation

3.1 Generalization of the system model

 

Fig.1 The upper water table threshold of the early warning in west of Songnen Plain

 (1)Spatial distribution of the aquifer 

The phreatic aquifer in the area were divided into: ①Porous phreatic aquifer of sloping plain in front of western mountains, which comprised of sandy pebble gravel and cobble of mid Pleistocene series and upper Pleistocene series, with coarse grain and pole-strength conductivity (K＝140-300m/d); ② Porous phreatic aquifer in valley plain, eastern high plain and in the middle of the Huolin River, which comprised sand and sandy gravel, with relatively strong conductivity (K＝10-100m/d); ③ Porous phreatic aquifer in the area between Songhua and Lalin River, and in the western low plain, which comprised fine sand, middle and powdery fine sand, with relatively weak conductivity (K＝10-20m/d); ④ Porous phreatic aquifer in the middle of the low plain as Qian’an and Da’an district, which mostly comprised loess sub-sandy soil, powdery fine sand and sub-sandy soil, with quite fine grain and very bad conductivity (K＝3-10m/d). The conceptual model of the hydrogeology in the area was shown in Fig. 2, where illustrated the distribution of the boundary, the character, the zone of the hydrogeology conductivity and the distribution of the observation wells.

(2) Generalization of the lateral boundary 

The Tao’er river alluvial-proluvial fan in the northwest of the calculated area was adjourn to the bedrock mountain area which was mainly lava and with quite low conductivity, so it was generalized to be the confining boundary. The Tao’er and Jiaoliu River valley with perennial water were considered to be runoff boundary. Using the runoff information of the upstream and lower river and the water level contour, after integrated analysis, the transform from surface water to groundwater was calculated based on the water balance relation. As to the south and the east, the boundary was generalized to be confined for the watershed of the groundwater. In the northeast, there was the Nen, Songhua, lalin and secondly Songhua River, where were all drainage boundary, and could be generalized to be first boundary, that is, known water head boundary. In addition to the above boundaries, there were lots of water level observation wells at other boundary areas, where could be generalized to be first boundary with known water head.

(3)Generalization of the vertical boundary   

Based on the dynamic changing information of the phreatic water level, the phreatic changes was influenced by the manual exploitation, phreatic evaporation, and the infiltration of precipitation, irrigation, and of the river bed, so the upper boundary was the water exchange boundary. The lower boundary was divided into two kinds of conditions, one was at single phreatic aquifer as the porous-phreatic area in alluvial-proluvial fan and the valley plain, where the lower boundary was relatively confined and was generalized as confining boundary, the other was double layer aquifer, which was upper pore phreatic water and lower pore confined water, and the vertical boundary was belong to the discharge boundary of Daqinggou group with a feeble leakage aquifer.

Fig.2 Groundwater hydrogeological conceptual model in west of Songnen

3.2 Numerical model foundation

Based on the hydrogeology conceptual model above, numerical model of the phreatic flow was built up as following.

Where:-hydraulic conductivity (m/d), -specific yield of the phreatic aquifer;

H,H₀-groundwater level, phreatic water (m); B-bottom of the aquifer (m);

Q_r-infiltration intensity (m/d); Q_d –drainage intensity

(include pumpage) (m/d);

Q_i-well exploitation (m³/d); h₀-initial head (m); h₁-water level of first boundary (m); q-unit wide flux of secondary boundary; x,y-coordinate (m); D-calculated area; - first and second boundary;

-inner normal line in the boundary; n-total wells number.

3.3  Solution to the model

Traditional quadrilateral dividing was selected to divide the calculated area into one layer, 38 rows, 70 columns, and 2660 units. There were 149 confirmed water head units, 753 invalidating calculated units, and 1408 water head changing units (efficient calculated units). The unit size should be consistent to the size of grid unit of the guard line. The calculated area was 35201.17km², which was 74.77% of the western Songnen plain, and the average unit area was 25km². Spaces between rows and columns were all 5000m, and the thickness was fixed by the specific condition.

3.4 Identification and validation of the model

Compared the water head of calculated to that of the observed, the relevent hydrogeology parameters were obtained in reversion. The identification time was from Oct. 15^th, 1996 to March 30^th, 1997, and the time step was 15d, with 11 periods of time. There were few recharges and drainages, and the groundwater was in normal year, so it was easy to calculate.

The parameter zone of calculated model on phreatic aquifer was basically according to the hydrogeology ranges of the area. The initial value of the parameters was confirmed by the pumping tests results in the perambulation and study periods. To identify the model in winter, there was no recharge from precipitation and irrigation, and the phreatic water evaporation could be omitted. The sink and source from leakages, rivers and lakes could be added into corresponding calculated units, and the domestic water for cities was distributed to the units by intensity.

At last, the sink/source and initial heads were input into the numerical model, and matched the hydrogeology parameter until the fit error between the calculated and observed water level reached the demand at the end of the run time. The validation indicated that the units with error of fit less than 0.5m were over 70% of the whole, which showed that the hydrogeology conceptual model and the numerical model were all reliable.

To validate the reliability, it was tested that the regional numerical model built. The lower water level period (Mar. 11 to May 11, 1997, that is, 61d) and water table ascending period (May 11 to Aug. 11, 1997, that is, 92) were chosen as the validated period of time. The lower water level was taken to draw the initial head and the sink/source of the different period of time, input and calculated the water level at the end time. The validated error of fit the calculated and the observation water level of the two period of time indicated that the units with error of fit less than 0.5m were over 73% of the whole, and the numerical model could be used to the groundwater forecast.

3.5  Phreatic water level forecast

According to the future program project of the study area, the water level time was confirmed as 15a, that’s 1999-2015. As the dynamic observation information of the groundwater was relatively complete in May, 1999, and there were many observation wells then, it was made to be the basic value of the forecast that the phreatic water level at that time.

(1) Boundary and sink/source   

The precipitation was the main recharge resource of the groundwater. The Monte-Carlo method was selected to the forecast in the research work. Ten lasting observation wells of the first boundary were chosen to set up the head-time relation formula, which showed that the validating correlation results were all obvious. Other boundary wells were given by interpolation method. The sink/source forecast of the phreatic water model was designed according to the water with program of the different industry and the water conservancy program, at the same time, the forecast was ran based on the correlation equation between the source-sink and the easy confirmed variance as time or else. For example, the correlation between river and groundwater was confirmed by correlated model of river leakage with precipitation and runoff.

(2)Forecast results and analysis   

The forecast results (Fig. 3) showed that the phreatic water flow changed little till 2015, no matter from the whole shape or the direction of the groundwater flow. The water table draw down relatively little, the maximum value was less than 5.5m. During the forecast period, the phreatic water head changed slightly along with the precipitation. At the alluvial-proluvial fan in Baicheng, where had strong pumping rate, the water level descended continually with small amplitue, the average draw down was 0.25m/a. At paddy field area in Qianguo county and Zhenlai county, where took surface water as the source, phreatic water level ascended a little for the infiltration recharge of the irrigation water. For example, the phreatic water head of 143# and 210# had ascended for 12 years in the 15 years. The result indicated that the exploitation of the paddy field could increase the phreatic water recharge, and the salinization risk would emerge along with the ascending of the phreatic water head.

Fig.3  The forecast of water table of the early warning in the west of Songnen

4 Phreatic Water Level Early Warning

Based on the water table forecast, calculated the water level of status in quo and of the forecast, and compared them to the warning limit to fulfill the early warning. GIS spatial analysis was used to do the grid map overlay and the attribute judgment, in which multilayer grid data composition was adopted, such as arithmetic operation and vector polygon overlay analysis. MapInfo was used to deal with the layers into grid map as threshold, phreatic water flow of status in present and of forecast, and do the early warning judgement. The data beyond the upper and lower warning limit were divided into different warning value and the warning condition. The results were shown in Tab. 1 and Fig. 4.

Fig.4  Grid map of early warning on water environment in west of Songnen Plain in 2015a

 

Tab.1  Early warning Result of the phreatic water table

        Other warning areas were in district of Qian’an, Tongyu and the southwest of Da’an. The topography in the area was relatively low, and there were many dishing closed billabong, marsh and low plain, so the drainage was hintered after irrigation, and resulted in the phreatic water level ascended and secondary salinization. The area with warning value beyond the lower warning limit were mainly distributed in the plot between Songhua River and Lalin River, and in western sloping plain in front of the mountains, where the water resources were relatively rich and high exploited. The phreatic water level was higher than the confined water table in most of the area between Songhua and Lalin River. Under the influence of the manual exploitation on pore confined water, the phreatic water supplied the confined one, and it would be much more evident when the confined water was exploited largely. Together with the water table dropped down beyond the lower warning limit, the phreatic water were also supplied the confined aquifer in western sloping plain in front of the mountains. In the recent years, the climate in the area was gradually drying, precipitation was reduced, and the groundwater recharge resources reduced obviously. At the same time, the exploitation on the groundwater increased year by year, which resulted in the continually lowering of the water head even beyond the lower warning limit.

5  Conclusions

Water resources early warning was a new research subject, which involved many new theories and methods. It was important to define the early warning object and to confirm the warning limits on the groundwater environment early warning, which was a new search by GIS-PModflow system.

The results showed that there were warnings of both secondary soil salinization and local water table descending. The former was caused by water level ascended from irrigation, and the latter was caused by the heavy exploitation. So it was essential to exploit groundwater resource rationally, which was the important base to ensure the sustainable development of the economy and environment in the area.

The water environment system was rather complex and involved many influenced factors. Under the control of physical geography, hydrology and artificial factors, the early warning objects were different obviously in specific areas. So it was essential to have deeply studies in such problems as environment character, early warning object, foundation of warning index system, technique and method of early warning and so on. GIS-PModflow integrated system was adopted in the paper and useful experiments was done on water environment early warning in western Songnen plain. The method was predominant in numerical simulating, forecasting and early warning analysis judging of the phreatic water level, still with the warning limit confirming, and it would be a new research direction in the water environment study field.

References

[1] Fu Bojie. Study on the Theory and Application of Regional Eco-Environment Early Warning. Chinese Jounal of Applied Ecology. 1990, 4(4): 436-439.

[2] Xu Xuegong. Study on Evaluation and Pre-Warning of Ecological Environment in the Yellow River Delta. Acta Ecologica Sinica. 1996, 16(5): 461-468.

[3] Chen Guojie. An Approach on Environmental Warning. Chongqing Environmental Science. 1996, 18(5): 1-4.

[4] Chen Zhijian, Chen Guojie. Study on Early Warning System on Environment Impress estimation. Environmental Science.1992,13(4): 20-24.

[5] Lin Nianfeng, Tang Jie. The Functions of Combined System EIS-GIS and Its Significance in Quaternary Studies. Quaternary Sciences. 2001, 21(2): 100-107.

[6] Lin Nianfeng Tang Jie. The Effect and Signification of GIS and EIS on the Research of Environmental Geology. Soil and Environmental Sciences. 2000, 9(4): 259-262.

[7] He Yanbo, Yangkun. Remote Sensing and Geographic Information System Apllied to Hydrologic Modeling. Geology geochemistry. 1999, 27(2): 99-103.

[8] Yin Kuihao, Weng Lida. Application of GIS in Water Resource Development, Utilization, and Protection. Water Resources Protection. 1999, (1): 10-14.

[9] Fan Shengyue, Xu Jianhua. Normal Numerical Model on Human Activity Effect Analysis to Water and Soil Loss and Desertification System. Scientia Giographica Sinica. 1992, 12(4): 305-307.

[10] Wang Zhanxing, Xiu Qingshan, Lin Shaozhi, et al. Groundwater and Quaternary Geology at Baicheng Area. GeologyExpert Report(6) Hydrogeology and Engineering Geology(No. 3). Beijing: Geology Publishing House. 1992

[11] WeiWenqiu, YuJianying. Application of GIS to Hydrology and Water Resources Management. Advances in Water Science. 1997, (3): 296-300.

[12] LiuYong,JingWenyong. GIS Technology and Its Applications in Environmental Sciences and Engineering. Environmental Science. 1997, 18(2): 62-65.