Generation of runoff characteristics over three time periods for four typical forests in Jinyun Mountain, Chongqing City, southwest China

Abstract

In order to provide a basis for water conservation in the restoration of vegetation for an urban water resource area, we studied the generation of runoff characteristics in four typical forests over three time periods in Jinyun Mountain, Chongqing City, based on the observation data of rainfall and runoff processes during the period 2002-2005. The results show that: 1) Rainfall was distributed evenly during the years 2002-2005. Annual rainfall variability was 4.46% and coefficient of variation was 0.0618. Average monthly rainfall tended towards a normal distribution N (113.8, 4597²). 2) Both precipitation and runoff can be clearly divided into a dry and a wet season. The dry season was from October to March and the wet season from April to September. Most of annual runoff of the four forest stands occurred in the wet season. The surface runoff in the wet season accounted for more than 85% of the annual runoff, and more than 75% of underground runoff. 3) Both peak values of surface runoff and underground runoff occurred in June. The relation between monthly rainfall and surface/underground runoff was fitted by the model W=a{P}^{\mathrm{2}}+bP+c. The order of annual surface runoff was as follows: Phyllostachys pubescens forest > shrub forest > mixed Pinus massoniana-broadleaf forest > evergreen broad-leaved forest. The annual underground runoff was evergreen broad-leaved forest > mixed Pinus massoniana-broadleaf forest >Phyllostachys pubescens forest > shrub forest. 4) Under similar rainstorms events, the order of the surface runoff coefficient was: evergreen broad-leaved forest < mixed Pinus massoniana-broadleaf forest < shrub forest < Phyllostachys pubescens forest. The underground runoff coefficient was: evergreen broad-leaved forest > mixed Pinus massoniana-broadleaf forest >Phyllostachys pubescens forest > shrub forest. The relation between rainstorms and surface runoff was fitted by the linear relationship: Q=mp-n. Both mixed Pinus massoniana-broadleaf forest and evergreen broad-leaved forest have better flood regulation effects on an annual and monthly basis and per individual rainstorm. The function of Phyllostachys pubescens forest is the worst on all three bases.

Keywords

Jinyun Mountain / time periods / runoff characteristics

Introduction

Runoff, as an important hydrological event, is affected by watershed vegetation, soil, climate and other hydrological factors. It has become a basic indicator to evaluate the function of forests in water and soil conservation and in the reduction of flood peaks (Fang et al., 2001). The process of precipitation-runoff plays an important role in forest ecosystems and is a reflection of forest functions in soil and water conservation (Zhou et al., 2000). In recent years, studies on the runoff generation characteristics over different periods have become an urgent matter and an important issue in the area of hydrology. Hydrological phenomena change over space and time. Over long periods of time, hydrological events show considerable differences and changes in property characteristics (Li et al., 2005). At present, spatial studies on the generation of runoff characteristics are generally conducted on a small scale, such as a slope, or somewhat larger for a watershed (Phliip 1991; He et al., 1997; Swanson 1998; Li et al., 2006; Qin et al., 2006). The study on the generation of runoff characteristics over different time periods has mainly been carried out for changes in the river runoff (Zhao et al., 2002; Liu et al., 2003; You et al., 2005). Few studies are reported on surface and underground generation of runoff characteristics of forests.

Jinyun Mountain is located at the tail-end of the Three Gorges reservoir. It is covered by a complete, subtropical evergreen broad-leaved forest community. Already for a long time, it has been a natural laboratory for the study of water and forest conservation and management. We analyzed the changes in the generation of runoff characteristics of different stands on an annual or monthly basis and on individual rainstorm events and discuss the various hydrological effects of different stands. We hope to provide technical parameters and a scientific basis for afforestation, for the purpose of water conservation in urban water source areas.

Site description

The Jinyun Mountain Nature Reserve Region is located on the west bank of Wentangxia of the small Three Gorges of the Jialing River (29°45′N, 106°22′E). The highest elevation is 951.5 m with a 600 m height difference from its lowest point. Soils are mostly yellow and paddy soils (pH 4.0-4.5). The horizontal zone belongs to a typical mid-subtropical climate, with an average annual temperature of 13.6°C, an average annual precipitation of 1611.8 mm and an annual evaporation of 777.1 mm. There are rich plant species, including six major vegetation types of evergreen broad-leaved forests, warm conifer forests, Phyllostachys pubescens forests, evergreen broad-leaved shrub forests, subtropical shrubs, grasses and aquatic vegetation. The major dominant species are Gordonia acuminate, Neolitsea aurata, Pinus massoniana, Cunninghamia lanceolata, Cinnamomum camphora, Pinus armandi and Symplocos setchuanensis. The hydrological effects of four typical forests (broad-leaved forests, mixed forests, bamboo forests and shrub forest) were studied (Tables 1 and 2).

Methods

Rainfall observation

Climate data such as precipitation, wind speed, evaporation, air temperature, humidity, soil temperature and moisture were measured by a set of fully automated weather devices outside the forests, down to a resolution of 10 min. Precipitation data were recorded by a B-432-Z automatic hyetometer.

Runoff observation

A runoff plot (20 m × 5 m) was established in each typical stand. In all of the runoff plots, precast concrete wall panels were placed deep into the bedrock to prevent runoff entering from outside the runoff plots. Filter layers were established in grooves at the top of walls and connected to holes with pipes, in order to direct underground runoff to an observation room with runoff testing equipment. Drainage channels were established at the bottom of the observation room. Along at the top and two sides of the runoff plots, cut-off drains were built to intercept unobserved runoff around the slopes of the runoff plots. In the runoff plots, the slope runoff of rainfall collected in the groove sets and water-flows went into the observation room through the catheters of the groove. Every observation room was equipped with a T9801 microcomputer-based multi-channel automatic dump flow meter which monitored and recorded automatically the processes of surface runoff and underground runoff of the different runoff plots. Runoff from each plot was recorded at a resolution of 10 min.

Results and analysis

Rainfall characteristics

Rainfall data from 2002 to 2005 in Jinyun Mountain were used to calculate average monthly and annual rainfall. During this period, rainfall mainly occurred from May to September, i.e., 66.08% of total rainfall (Table 3). Annual rainfall characteristics were denoted by the annual rainfall variability (K) and coefficient of variation (CV) (Zhang et al., 1991). These reflect, respectively, the stability of relative changes and the rate of changes in rainfall. The larger the two indices were, the greater the changes in annual rainfall and the higher the probability of the occurrence of droughts and floods.

(1)K={\displaystyle \sum _{i=\mathrm{1}}^n\left|{\displaystyle \frac{H_i-\overline{H}}{\overline{H}}}\right|}\times \mathrm{100}\%/n

(2)\mathrm{C}\mathrm{V}=\sqrt{{\displaystyle \sum _{i=\mathrm{1}}^n{({\displaystyle \frac{H_i}{\overline{H}}}-\mathrm{1})}^{\mathrm{2}}}/(n-\mathrm{1})}

where H_i is the rainfall and \overline{H} the average annual rainfall. According to equations 1 and 2, the annual rainfall variability was 4.46% and the coefficient of variation 4.46%. Both of them were far less than 100%. The distribution of the annual rainfall is relatively homogeneous and the variation among inter-annual rainfalls is small. According to the division by Lu et al. (2006) between wet and dry seasons, there is a clear division between the two in our study area. The dry season is from October to March and the wet season from April to September. The proportion of annual rainfall in the wet season is 74.29% and only 25.71% in the dry season. The distribution of the average monthly rainfall conforms somewhat to a normal distribution N (113.8, 4597²) (Fig. 1).

Seasonal distribution of runoff

Surface runoff and underground runoff occur with the same frequency as the rainfall during the wet and dry seasons (Fig. 2). The wet season was from April to September during our research period. Amounts of runoff from each typical forest in the wet season were larger than those in the dry season. Surface runoff from each typical forest in the wet season is given in the following order: evergreen broad-leaved forest (92.6%) > mixed Pinus massoniana-broadleaf forest (92%) > shrub forest (91%) > Phyllostachys pubescens forest (89.9%). Underground runoff during the wet season is presented in the order of shrub forest (98.4%) > mixed Pinus massoniana-broadleaf forest (92.0%) > evergreen broad-leaved forest (86.4%) > Phyllostachys pubescens forest (77%).

Monthly distribution of runoff

Based on the distribution of monthly runoff during the period 2002-2005, the change of monthly surface runoff and underground water runoff of the four stands is shown in Fig. 3, which indicates that the peak flow occurred in June. Monthly surface runoff followed in the order of: Phyllostachys pubescens forest (137.0 mm)>shrub forest (51.4 mm) > mixed Pinus massoniana-broadleaf forest (48.8 mm)>evergreen broad-leaved forest (36.6 mm). Monthly underground runoff is shown in the following order: evergreen broad-leaved forest (385.0 mm) > mixed Pinus massoniana-broadleaf forest (250.5 mm) > Phyllostachys pubescens forest (46.1 mm) > shrub forest (15.2 mm). Therefore, given the same annual rainfall, evergreen broad-leaved forests are efficient in converting surface runoff into underground runoff. Evergreen broad-leaved forest had the best flood regulation effects. Phyllostachys pubescens forests were the worst in both aspects.

The generation of runoff of the four types of forests in our study area is of the type that belongs to a stored-full runoff. Surface runoff and underground runoff are mainly correlated with rainfall (Lu et al., 2006). Monthly runoff was calculated using the following model:

(3)W=a{P}^{\mathrm{2}}+bP+c

where W is the monthly surface/underground runoff, P the monthly rainfall and a, b, c are coefficients to be estimated for each forest type. The calculated results are shown in Table 4. When the monthly rainfall breaches a certain level and if it continues to increase, both surface runoff and underground runoff of the four forest stands face substantial increases.

Runoff characteristics under rainstorm conditions

Observations show that if the rainfall, occurring in the four forest stands, was less than 10 mm, there would be no surface and underground runoff generated. Wang et al. (2005) reported that the response of runoff on rainstorms was obvious. Therefore, runoff characteristics are correlated with rainstorms. Most rainfalls in the study area lasted for less than 1440 min during an entire rainfall process or during specific periods. Fan’s model was introduced to calculate rainstorm standards (Fan et al., 2003)

(4)K=P^{\mathrm{2}}/t

where P is rainfall and t length of rainfall. When K is larger than 2, it is a rainstorm event.

Characteristics of surface runoff

Single rainfall events during 2002 to 2005 were divided by K (Eq. 4). Twenty-five surface runoffs and underground runoffs under conditions of typical rainstorms were selected and analyzed. Table 5 shows that the average rainfall lasted 14.53 h. Therefore, most rainstorms in the study area are of a long-time nature. Compared with the length of a rainfall, the surface runoff of shrub forests lasted 55.8 min, 174.6 min for Phyllostachys pubescens forests, 188.4 min for evergreen broad-leaved forests and 201.6 min for mixed Pinus massoniana-broadleaf forest. The average initial abstraction duration of surface runoff of the four forest stands are in the following order: evergreen broad-leaved forests (88.6 min) > mixed Pinus massoniana-broadleaf forests (66 min) > shrub forests (60.5 min) >Phyllostachys pubescens forests (60.4 min). Evergreen broad-leaved forests and mixed Pinus massoniana-broadleaf forest provide useful services in flood regulation, while Phyllostachys pubescens and shrub forests play a much smaller role.

Given the same rainstorm events, the average surface depth of runoff from evergreen broad-leaved forests was 3.4 mm, 17.1 mm from Phyllostachys pubescens forests, 7.0 mm from mixed Pinus massoniana-broadleaf forests and 8.2 mm from shrub forests. The order of surface runoff coefficients was as follows: evergreen broad-leaved forests (0.06) < mixed Pinus massoniana-broadleaf forests (0.13) < shrub forests (0.15) < Phyllostachys pubescens forests (0.31). Therefore, evergreen broad-leaved forests enhance interception of rainfall and decrease surface runoff. It has the best flood regulation effects. Mixed Pinus massoniana-broadleaf forests and shrub forests are poor in this regard and Phyllostachys pubescens forests are the worst.

Characteristics of underground runoff

Underground runoff has a more obvious hysteresis effect than surface runoff. Average initial abstraction duration of underground runoff of evergreen broad-leaved forest was the shortest at 113.3 min. It was 124.3 min for the mixed Pinus massoniana-broadleaf forest, 230.7 min for the Phyllostachys pubescens forest and 372.2 min for the shrub forest, i.e., the longest (Table 5). Evergreen broad-leaved and mixed forests could quickly convert surface runoff into underground runoff, decrease surface runoff and weaken the cumulative effect of flood peaks.

Table 5 indicates that the underground runoff lasted longer than the time of precipitation. The underground runoff of mixed Pinus massoniana-broadleaf forest lasted 53.06 h. The underground runoff of evergreen broad-leaved forest lasted 57.78 h while it lasted for 8.92 h for the Phyllostachys pubescens forest and 5.24 h for the shrub forest. Extending the underground runoff can effectively slow down the conflux time of slope runoff and underground runoff and weaken the effect of channel floods. Therefore, evergreen broad-leaved forests and mixed forests have the better flood regulation function.

The order of the depth of underground runoff was: evergreen broad-leaved forests (30.9 mm) > mixed Pinus massoniana-broadleaf forests (24.6 mm) > Phyllostachys pubescens forests (3.5 mm) > shrub forests (1.4 mm). The underground runoff coefficient was ranked in the following order: evergreen broad-leaved forest (0.56) > mixed Pinus massoniana-broadleaf forest (0.45) > Phyllostachys pubescens forest (0.06) > shrub forest (0.03). Evergreen broad-leaved forests and mixed forests can effectively intercept precipitation and convert it into underground runoff. They have a good hydrological function, while Phyllostachys pubescens forests and shrub forests are not very effective.

A good linear relationship could be established between rainfall and surface/underground runoff of a rainstorm event (Table 6). The model is as follows:

(5)Q=mp-n

where Q is surface runoff (or underground runoff), p is rainfall of a single rainstorm event and m and n are coefficients.

Under the condition of a single rainstorm event, both the surface and underground runoffs were positively related with an increase in rainfall. The correlation of underground runoff to rainfall was poorer than that of surface runoff. This was mainly due to the effect of rainfall intensity. Because of a thinner litter layer in Phyllostachys pubescens forests and shrub forests and lower soil infiltration capacity of short-lived rainstorms, most of the rainfall came too late to be converted into underground runoff. Therefore, underground runoff of both Phyllostachys pubescens and shrub forests were mostly affected by rainfall, especially rainfall intensity.

Conclusions

Rainfalls were distributed evenly during the years 2002 to 2005. Annual rainfall variability was 4.46% and the coefficient of variation 0.0618. Average monthly rainfall came close to a normal distribution N (113.8, 4597²).

Both precipitation and runoff can be clearly divided into a dry and a wet season. The dry season is from October to March and wet season from April to September. Most of the annual runoff of our four forest stands occurred during the wet season. The surface runoff in the wet season accounted for more than 85% of the annual runoff and for more than 75% of underground runoff.

Both peak amounts of surface runoff and underground runoff occurred in June. The relation between monthly rainfall and surface/underground runoff was fitted by the model

(6)W=a{P}^{\mathrm{2}}+bP+c

. The order of annual surface runoff was as follows: Phyllostachys pubescens forest (137.0 mm) > shrub forest (51.4 mm) > mixed Pinus massoniana-broadleaf forest (48.8 mm) > evergreen broad-leaved forest (36.6 mm). The order of annual underground runoff was evergreen broad-leaved forest (385.0 mm) > mixed Pinus massoniana-broadleaf forest (250.5 mm) > Phyllostachys pubescens forest (46.1 mm) > shrub forest (15.2 mm). Evergreen broad-leaved forests have the best flood regulation effects. Phyllostachys pubescens forests are the worst.

Under the same rainstorm conditions, the order in terms of the surface runoff coefficient was evergreen broad-leaved forest (0.06) < mixed Pinus massoniana-broadleaf forest (0.13) < shrub forest (0.15) < Phyllostachys pubescens forest (0.31) and the order of underground runoff coefficient was evergreen broad-leaved forest (0.56) > mixed Pinus massoniana-broadleaf forest (0.45) > Phyllostachys pubescens forest (0.06) > shrub forest (0.03). The relation between rainstorms and surface runoff was fitted quite well by a linear relationship Q=mp-n.

Both mixed Pinus massoniana-broadleaf forests and evergreen broad-leaved forests have excellent monthly and annual flood regulation effects. In the medium and long-term, these forests perform their role in water conservation and flood regulation very well. During rainstorms, they can lessen flood peaks to reduce flood hazards. Over longer periods of time, the function of shrub forests is not very constructive and those of Phyllostachys pubescens forest are the worst.

References

[1] Fan X K, Wu P T, Feng H (2003). Decision method and index of rainstorm. Sci Soil Water Conserv, 1(3): 72-75

[2] Fang X J, Meng G T, Lang N J, Li G Y, Yuan C M, Wen S L (2001). Law of runoff yield of man-made plant community in Mountain of Central Yunnan Plateau. J Soil Water Conserv, 15(1): 66-68 (in Chinese)

[3] He K N, Zhang J J, Zhu G Z (1997). The hillslope surface flow of the forests for soil and water Conservation in Western Shanxi, Loess Plateau Region. J Beijing For Univ, 19(4): 1-6 (in Chinese)

[4] Li F D, Song X F, Liu C M, Yu J J, Yang C, Liu X C, Hu K, Tang C Y (2006). Discharge recession from runoff plots in representative mountain area in North China. J Beijing For Univ, 28(2): 79-84 (in Chinese)

[5] Li X G, Bi H X, Kou X, Ma Y G (2005). The influencing factors of runoff and sediment and scale analysis. For Invent Plan, 30(6): 102–105 (in Chinese)

[6] Liu J P, Tian F W, Huang Q (2003). A multiple timescales analysis of the runoff variation along the upper reaches of the Yellow River. J Appl Sci, 21(2): 117–121 (in Chinese)

[7] Lu S Y, Zhang P Y, Yu G, Zhu W P, Xiang C S (2006). Wet-dry seasonal regularity of phosphorus removal from agricultural runoff by constructed wetland in Wangjiazhuang Lakeside Area. Dep Environ Sci Engin, 25(5): 1313-1317 (in Chinese)

[8] Phliip J R (1991). Hill slope in filtration: Divergent and convergent slopes. Water Resour Res, 27(6): 1035-1040

[9] Qin Y S, Yu X X, Chen L H, Liu S (2006). Spatial scale effects of runoff on the water resource conservation forest watershed with in the Miyun reservoir basin. Act Ecol Sin, 21(6): 913-918 (in Chinese)

[10] Swanson R H (1998). Forest hydrology issues for the 21st Century: A consultant viewpoint. Am Water Resour Assoc, 34(4): 755-763

[11] Wang Y Q, Wang Y J, Zhu J Z (2005). Responses of slope runoff generation to rainstorm with different land use types in Jinyun mountain in Chongqing city. Sci Soil Water Conserv, 3(4): 19-26

[12] You W H, He D M (2005). Correlation and multi-timescale characteristics of the monthly variations for Lancang River flow in the region of Yunnan. J Yunnan Univ, 27(4): 314-322 (in Chinese)

[13] Zhang Z G, Zeng B Q (1991). Preliminary analysis of rainfall data from Wangjianou Watershed in Western Shanxi. Soil Water Conserv Sci Technol Shanxi, 2: 6-9 (in Chinese)

[14] Zhao Y T, Yu X X, Zhang Z Q, Deng Y (2002). Nonlinear characteristics of precipitation and runoff in sub-alpine of the upper reaches of Yangtze River. J Beijing For Univ, 24(2): 72–74 (in Chinese)

[15] Zhou G Y, Yan J H, Shen W J, Hou A M, Yu Z Y, Lin Y B (2000). Surface flow in an acacia plantation and an orchard in Heshan, Guangdong province, China. Acta Phytoecol Sin, 24(4): 451–458 (in Chinese)

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg