Shi-jun Sun¹, Yao-yuan Ding², Bo Cao²  

1. Shenyang Agricultural University; Shenyang 110161，China

2. Beijing Hydraulic Research Institute, Beijing 100044，China

Abstract: Since the 1980s, more and more groundwater resources have been extracted in the northern plains of China due to the lack of rainfall. As a result, the farm-land groundwater table has decreased while the storage capacity of soil has increased in the well irrigation areas. It is necessary to adopt suitable measures to reduce the runoff and evaporation losses, and to use flood water and runoff to recharge the local groundwater. In this paper, four years data on rainfall, groundwater and soil moisture from the water resources experimental area in south-eastern suburb of Beijing was processed in order to identify the dynamic characteristics of soil water under deep groundwater table conditions. The data was divided into three phases for each year--the period of spring irrigation, the rainy season and the period from late autumn to early spring, in order to analyze the dynamic characteristics of soil moisture in different seasons. The results showed that during the dry season, a small quantity of irrigated water or precipitation (p<80 mm) is mainly stored in 0~1mdeep soil whereas during the rainy season, when the top of soil is relatively damp, the water will be mainly stored in 1~2.5 m deep soil layer without surface flow appearing if the rainstorm (p=100~150 mm) occurs. The research also shows that soil reservoir had very good regulation capacity and 85% of the normal annual precipitation was stored in the 3m deep soil (namely the soil reservoir depth). Therefore, if appropriate measures are undertaken on farmland before a heavy rainstorm occurs, there will be minimal runoff appearing and most of the precipitation will be stored in the soils.

Key words: well irrigation areas, rainfall and runoff, groundwater, soil water dynamic state

Introduction

Soil water plays a central role in the interaction between surface water and groundwater. It is a vital water resource especially for the agricultural sector, whereby they are absorbed by crops after transforming from rainfall and irrigation water. In the past 20 years, because of insufficient rainfall and uneven rainfall distribution in different seasons of each year, many research projects have been carried out to look into the soil-water movement laws and potential rainfall use in the arid and semi-arid areas in China,[1-5]. Nevertheless most focused on dry farming fields, where groundwater levels are usually very deep and not taken into account as an important influence on soil water.

Since the 1970’s, shallow groundwater has been utilized in most parts of the country. Particularly in the last 20 years, groundwater has been over-extracted in northern China due to long-term rainfall deficit, which resulted in water levels falling gradually[6].

At the same time, with the increase of unsaturated soil depth, the soil capacity of water storage increases as well. Some studies in the northern China plain proved that there was almost no runoff occurring in local agricultural fields during rainfall events (under 150mm)[7~10]. Most rainfall infiltrated into the unsaturated soil with some losses via evapo-transpiration.

In this paper, soil water content dynamic characteristics of well irrigation fields is discussed by examining observation data from the Tiantanghe Valley water resources experimental site in the south-eastern plain of Beijing, which would provide vital information for extracting groundwater rationally, making full use of rainfall and recharging soil water resources in the northern China plain.

The experimental site is located on the left bank of Yongding River in the south-eastern suburb plain of Beijing, whereby the total area is 422.5km2 with 243.5 km2 being arable land. The mean annual rainfall is556.3 mm, over 80% of which takes place during the flood season (from June to September). According to investigation conducted on the experimental site, field soils are mainly composed of silt, sand clay, and partly clay layer in different depths. Since 1981, owing to the fact that there is no surface water supply anymore, all agricultural fields irrigation depended much on groundwater extraction. A local survey found that till 1994 the common depth of groundwater approached 7 m to 13 m.

Soil Moisture Dynamic Characteristics in Different Depth

There is a designated soil moisture observation station in the water resources experimental site, where a 503DR neutral probe and soil moisture tension meter were set up to survey the 0-3m deep soil water content periodically and four years data has been attained from 1991 to 1995. Figure 1 shows the typical soil profile in the experimental fields.

            Fig.1 Representative soil profile in experimental region                   Fig.2 Soil moisture dynamic variations in different           soil depths(1 Nov 1993 to 1 Nov 1994)                                  

Fig.2 shows that soil moisture often fluctuates in the 0 m to 1 m depth of soil layer for one year, which is probably due to the effect of transpiration and infiltration, much more on the top soil layer than the deeper layer(1-2m or 2-3m). Generally speaking, soil moisture reaches its lowest value in the middle of February to March and the highest between July and September. During rainfall periods or irrigation and its interval, soil water responds more rapidly in the 0-1m layer than in the 1-2 m or 2-3 m layer, where the range of variation is between 4% and 12%(moisture weight percentage). Over four years of observation, soil moisture changing characteristics are more similar in the 1-2m and 2-3m layer than in the 0-1m layer, where water content increases in rainy period and decreases over the autumn and winter time of the year, but slower than that in the 0-1m soil layer.

Soil Moisture Dynamic Characteristics in Different Seasons for One Year

    During the experimental period from 1991 to 1995, dry periods were encountered in 1992 and 1993(annual rainfall of 400.8mm in 1992 and 359.4mm in 1993), and wet periods for 1994 and 1995 (annual rainfall of 751.2mm in 1994 and 624.5mm in 1995). Soil moisture changes over these years provided a good representation on current conditions. According to the variation of water content within one year, 3 phases could be deduced, which are the simple fluctuation period in the spring irrigation season, sharp change period in the rainy season and slow decrease of soil moisture from late Autumn to early Spring (in Fig.3, soil moisture in average soil depth of 0-3m, weight percentage).

Fig.３ 0～3 m depth Soil moisture、 groundwater and rainfall dynamic variations from 1991 to 1995

Soil Moisture Dynamic Characteristics during the Spring-irrigation Period

Irrigation time for wheat in North-China takes place during the spring period of the middle of March to May. Less precipitation is generally encountered at that time, and the weather is dry with the air temperature increases gradually. Evapo-transpiration of soil moisture is intense during spring time, resulting in the need for field irrigation. Affected by irrigation, the upper soil moisture of the 0-3m depth of soil fluctuates frequently, but its variation is relatively small at within 4%. After one or two days of wheat irrigation, the soil moisture in the 0-1m soil layer reaches a peak value, and at the start of the infiltration process, the deep soil moisture of the 1-2m and 2-3m layers increased steadily. While at the same time, soil moisture of the 0-1m soil layer decreases gradually due to the dual influence of transpiration and infiltration. The spring-irrigation period is identified as the period when soil water is lost in total. And because groundwater has been extracted for wheat irrigation, the groundwater level reaches its lowest level in the year at the experimental site before the rainy season arrives.    

The effect of spring-irrigation on soil moisture is analyzed using the following data from a case study conducted at the end of March 1994. The total precipitation was 2.2mm in the experimental site before and after the irrigation took place (from 14 February to 15 April).

According to the need for wheat growth, about 60mm of water was irrigated to the experimental field on 29 March, 1994 (there has been no rainfall before for 20days ). 5 days later, average soil moistures in different depths (0-1m, 1-2m and 2-3m) were 21.8%,16.67% and 20.71%, with an increase of 4.5%,2.0% and 0.6% respectively. The above analysis shows that, if there is no previous rainfall, irrigated water mainly supplied the upper soil layer (0-1m depth) , with the 1-2 m soil layer having a minor recharge and 2-3 m layer lesser.

                     （a）                                        (b)

Fig.4   Variations of soil water storage during the spring irrigation season , 1994

Soil Moisture Dynamic Characteristics during the Rainy Season

Over 80% of annual rainfall took place in June to September (rainy season) according to the 20 year historical data, with the soil moisture in 0-3m soil layer fluctuated dramatically. In every storm event, soil moisture increases immediately and then went down due to high temperature, which resulted in high soil and wheat evapo-transpiration. This phenomenon appeared frequently until the end of rainy season. Nevertheless, soil moisture increased overall in the rainy season and groundwater level went up because of lots of rainfall infiltration(see Fig.2. and Fig.3.).

（1）The following is based on a single rainfall event took place in the middle of June 1995, analyzing the soil moisture changes before and after the rainfall event. Table 1 shows all the rainfall events in the experimental site from 2 June to 25 June 1995.

Table1   Rainfall process in the experimental area on June 2-26, 1995

According to the recorded observation, from 9 to 14 June, soil moisture in the 0-0.5 m layer decreased gradually, but went up 0.28% in 1-2m layer and 0.02% in 2-3m depth soil. The main probable reason is that there was no rainfall event since the mid of May and the soil was so dry that irrigation was introduced for wheat.

Till 16 and 17 June, 72.3mm rainfall depth is obtained. 3 days later, the soil moisture of 0-1m went up significantly by 3.32% compared to that of 14 June. In addition, the soil moisture of 1-2m increased by 0.58% and 0.27% in 2-3m depths.

There was no rainfall event during 17 to 24 June, so soil moisture in 0-1m went down steadily, but the deep soil moisture still keep increasing progressively.

It is believed that , under the condition of dry field soil ,when a single rainfall event under 80mm takes place , the rainfall will mainly recharge the 0-1m soil depth.

Fig.5  Variations of soil water storage in the rainy season of 1995          Fig.6  Variations of soil water storage in the rainy season of 1994

(2)In order to analyze the soil moisture changes before and after certain large rainfall events, another case study based on several major rainfall events in the middle of July 1994 was put forward.

 There were 6 rainfall events from 29 June to 13 July (Table 2), among these , 2 rainfall events were above 100mm( 7 July, 12 July ).

Table 2   Rainfall process on July 2～12, 1994 in the experimental area

A total of 46.9mm depth of rainfall is obtained on 3 July 1994. According to the 2 section lines in Fig.6.(29 June and 4 July), it can be concluded that most rainfall was stored in the 0-1m soil layer. After that event, it rained for 20.7mm on 5 July and 116.6mm for 2 days during 7-8 July. Till 8 July, the soil moisture of 0-1m layer went up by 29.18%, with an increase of 8.04% compared to that of 4 July (21.4%). This fluctuation was the most considerable in four years from 1991 to 1995. In addition, the soil moisture of the 1-2m layer increased by 2%, which showed that when upper soil was recharged significantly the soil water would then infiltrate into the deeper soil.

There was no rainfall event on both 9 July and 10 July. The soil moisture in the 0-1m soil layer reduced gradually owing to surface soil transpiration, but increased in the 1-3m soil depth. Then it rained for 16.6mmon 11 July and 132.5mm on 12 July. The experiment data showed that the soil moisture in the 0-1m soil layer reached 30.8% on 13 July, hence the soil is almost saturated. In addition, analyzing the soil water data for 10 July and 13 July, it is found that the average soil moisture reaches 25.4% (the highest in the 4 years of experiment) and more rainfall infiltrate into 1-2.5m soil depth after the storm events.

The soil moisture in the 0-3m soil depth increased 7.99% from 29 June to 13 July. Another important fact is that there was almost no runoff occurring in the experimental fields after that rainstorm took place on 12 July according to the field investigation. A special borehole at the experimental site showed that its groundwater level raised 1.41m over the period from 29 June to 13 July.

From the above two cases, it can be concluded that, when there is no previous rainfall event or it rains little under dry soil conditions, if a single rainfall events below 80mm take place, rainfall would mainly recharge the 0-1m soil later. Then along with water transpiration and infiltration, soil moisture in 0-1m layer will decrease gradually. When the opposite happens, the upper soil layer will be recharged largely. If at this moment there is a single rainfall event over 100mm, most rainfall will be stored in the 1-2.5m soil layer. There would be no runoff appearing in the fields and at the same time groundwater level will rise a lot.

Soil Moisture Dynamic Characteristics from the End of Autumn to Early Spring

    Soil moisture changes slowly during the middle of November to early March. There are less rainfall events or field irrigation. In addition, because it is the cold season of the year, soil water transpiration is relatively lower than the usual season. The soil moisture fluctuates little, and the general trend is going down (Fig.2, Fig.3.).  On the other hand, to early March, along with the rise of air temperature and rapid growth of winter wheat , evapo-transpiration increase significantly. Soil moisture decreases to the lowest of the year, but groundwater level reaches its highest point of the year before the coming spring irrigation for wheat begins.

Effect of Different Rainfall Years on Soil Moisture

The changes of soil moistures are different before and after rainy seasons in different rainfall years. Table 3 shows the related data on soil moisture in different rainfall years.

Table 3   Comparison of soil water content of 0-3m soil depth in different rainfall years

   Note：Details of calculation process is in reference [11]。

Table 3 shows little rain in the rainy seasons of 1992 and 1993. The increments of soil moisture after the end of rain season were only 2.33% and 0.93% respectively, more than that of the beginning of the rainy season. But in 1994 and 1995 the increments of soil moisture reached 5.19% and 4.22% respectively, several times more than that of 1992 and 1993. So it is obvious that field soil is getting less water supply after the end of the rainy season in a dry year and vice versa. During the rainy year, the groundwater level increases significantly because of more rainfall infiltrates into the soil. As the groundwater levels are comparatively deep in the experimental site (generally beyond 7m), this results in the lagging effect of deep soil on the rainfall, so the surge in the groundwater level lasts till the beginning of the next spring irrigation season.   

Conclusions

With the experimental area representing typical well-irrigation fields (without surface water supply for agricultural purposes), the soil moisture variations within a year are composed of 3 phases, firstly the simple fluctuation period in spring irrigation season (from March to May), then the sharp change period in rainy season (from June to September ), and subsequently the slow reduction of soil moisture from late Autumn to early Spring (from October to March). Among these periods, soil moisture decreases notably during the spring irrigation season, increases evidently during the rainy season and goes down again as a whole from late Autumn to early Spring.

Under dry soil conditions (non rainy season), when there is agricultural irrigation or the occurrence of a rainfall event (less than 80mm), the water would mainly recharge the first meter of soil layer. When wet antecedent conditions are experienced, the upper soil layer will be recharged largely. If at this moment there is a single rainfall event of over 100mm, most of the runoff will be stored within the 1-2.5m layer of the soil. There would be no runoff occurring in the fields and at the same time groundwater level will rise.

Currently the groundwater depth extends beyond 7m in the Beijing plain, and almost no runoff occur in the agricultural fields after large rainfall storm events ( under 100mm). The research shows that the 0-3m soil layer has considerable storage capacity. 85% of mean annual precipitation could be stored in the 0-3m soil layer during the rainy season from 1991 to 1995. Hence if available measures were put into place to impound natural precipitation, such as leveling the ground, constructing small ridges or introducing deep plowing techniques in the fields, then when there is an extreme storm event taking place, there would not be much of runoff occurring. And most of the rainfall will recharge the field soil, which would be of significance towards the deteriorating situation of water shortage in northern China.

Acknowledgement  

The author wishes to thank Professor Tian,Yuan and Professor Xue, Song  for their constructive contributions.

References

[1]         Zhang, Li, Zhang, Caiying, Peng, Chunxiang(1988). Developing water storage scheme in the Changzhou district. Transaction of Agricultural Engineering 19, 22-29.

[2]         Wang, Jiantian, Xu, Zhuling(1993). Soil moisture dynamic changing characteristics and relative measurements on making use of rainfall in arid zones. Journal of Beijing Agricultural University 19,31-41.

[3]         Gong, Xuechen, Yang, Liting, Niu, Ruiming(1998). Agricultural field soil moisture movement trends analysis in the semi-arid plain of north-eastern Heibei privince. Journal of soil and water conservation 4,88-91.

[4]         Yang, Xingguo, Ke, Xiaoxin, Zhang, Xudong(2000).Study on soil water content movement characteristics in rain-fed agricultural fields of Gansu province. Journal of applied meteorology 11,205-212.

[5]         Yan, Changrong,Ju, Hui, Peng, Shiqi(2002).Agricultural field soil water content dynamic characteristics in the arid northern China. Transaction of Agricultural Engineering 18,11-14.

[6]         National Information Association of Groundwater(2000).Summary of 7^th national conference of groundwater in China. Ground Water 1,1-4.

[7]         Ding, Yueyuan,Xue,Song(1994). Experimental research of making use of surface and groundwater jointly in the typical plain of Beijing.Beijing Hydraulic Research Institute Report No12.

[8]         Fang,Sheng,Chen,Xiuling(1999).Multi-objective and comprehensive management measures of irrigation fields under the condition of joint control using surface water and groundwater. Groundwater 21,79-83.

[9]         [9] Xue, Song,Shi, Fengcai(1999). Flood water controlling and usage. Research and practice of  National groundwater development and utilization,24-29.   

[10]     Sun, Shijun, Ding, Yueyuan, Tian, Yuan(2001).Study on the use of precipitation and runoff during the flooding period to regulate and control groundwater in the Beijing plain. Advancement in water science 12,190-195.

[11]     Sun, Shijun, Ding, Yueyuan, Cao, Bo(2002). Regulating capacity of soil reservoir in the well irrigation fields of Beijing plain. Journal of Natural Resources 17,42-47.

[12]    Lei, Zhidong,Yang,Sixiu,Ni, Guangheng(1992). Soil water content dynamic characteristics under different groundwater levels.Journal of Hydraulics 2,1-6.