Long-term Hydrological Variations of Discharge, Soil Loss and Recession Coefficient in Three Small Forested Catchments, Kyonggi, Korea

Kim, Kyongha

Forestry Research Institute, Seoul, Republic of Korea

ABSTRACT This study aims to clarify the hydrological variations of the amounts of discharge and soil loss and the recession coefficient in three small forested catchments using the long-term hydro-data from 1983 to 1992. The experimental catchments are the natural-matured deciduous, artificial-planted coniferous and erosion-control worked mixed forest. The amount of discharge and soil loss changed with the amount of rainfall and forest type. Especially it was concerned on the variation of the recession coefficient related closely to soil storage capacity in three forested watersheds to clarify the effects of forest development on water resource augmentation. In erosion-control worked mixed forest, the recession coefficients affected by interflow and groundwater had gradually reduced since erosion control work finished. This may be interpreted to be caused by the increase of the soil storage capacity caused by an improvement of the soil physical properties e.g. soil porosity and bulk density in the catchment after erosion control work. In artificial-planted coniferous forest, the recession coefficients of surface runoff and interflow slowly decreased as the forest grew. However, all recession coefficients in natural matured-deciduous forest did not show any evident trends because of little change of the forest stand and soil structure.

Key Words: Discharge, Soil loss, Recession coefficient, Long-term variation, Forest type.


In Korea, the amount of rainfall shows high seasonal and annual variation. Most of rainfall precipitates intensely in July to August as typical monsoon type. The discharge in the wet season runs fast off to ocean. On the other hands, water resource usually is short of supply in the dry season. Also the mountainous area is very steep and soil depth is shallow. Disadvantageous situation for water resource management results in low percentage of water utility. The proportion of the water usage is no more than 24 percent of total amount of water resources, 126.7 billion tons year-1.

More than 65 percent of the land in Korea is forest covered, and forested drainage basins serve as water supplies. Generally it is well known that forest can modulate flows, moderate hydrologic extreme such as floods and droughts. Many devastated areas in Korea have been rehabilitated for soil and water conservation since 1970s. Some coniferous species such as Pinus koraiensis, Larix leptolepis were planted to establish artificial forest. Nowadays, forest in the head water catchment needs to be managed positively for a storm flow control and augmentation of low flow. The changes of a storage volume in a forested catchment after reforestation and erosion control work have to be clarified for the water conservation technique through the forest management.

In this study, the recession curve analyses uses to find how forest change has influence on the storage capacity of a forested catchment. The recession curve tells in a general way about the natural storage feeding the stream. Accordingly, it contains valuable information concerning storage properties and aquifer characteristics (Tallaksen, 1995). If the recession curve is plotted on semi-logarithmic paper, the result is usually not a straight line but a curve with gradually decreasing slope (Linsley et al., 1982). By plotting the logarithm values, the three types of the recession coefficient: surface runoff, interflow, and groundwater can be determined by finding an inflection point. This work aims to understand the influence of forest change on the variation of the recession coefficient with the lapse of time.


2.1 Study area

The experimental catchments locate at Kyonggi near Seoul metropolitan. Table 1 shows the condition of the topography and vegetation in the three experimental sites.

Table 1. The conditions of topography and vegetation in experimental catchments.

Forest type



Elevation (m)

Parent material

Tree Height


DBH (cm)








Recovered in 74







Planted in 76







Natural forest


The natural deciduous forest is mature about 60-year old and covered predominantly with Quercus serrata and Carpinus laxiflora. The coniferous forest that consists of Pinus koraiensis and Abies holophylla was planted at a stocking rate of 3,000-stem ha-1 in 1976. The mixed forest has been devastated since an erosion control work was established in 1974. As shown in Fig. 1(a)-(c) deciduous and coniferous forest catchments shape laterally wide while the mixed forest is longitudinally long.

( a ) ( b )

( c )

Fig. 1. Topographic views of deciduous(a), coniferous(b) and mixed(c) forest catchment.

2.2 Instrumentation

Rainfall was recorded continuously by tipping bucket recorder at every 0.5-mm tip-1 on the chart and snowfall also is measured as an equivalent depth. Water level has recorded by long-term recorder since 1979 at 90 sharp crested V-notch weir.

2.3 Data analysis

Of the rainfall over 50 mm, hyeto-hydrographs were plotted in order to select an individual event showed clear recession curve during the period of 1983 to 1992. The recession curves selected are plotted on a semi-logarithmic paper to analyse 3 components of the hydrograph on the basis of the inflection point. The recession curve has traditionally been separated into the linear components of surface runoff, interflow and groundwater( Barnes, 1939 ).

Qt = Q0e-t ( 1 )

where Qt is the discharge t time units after Q0, e is the napierian base, and is recession coefficient.

Here, 1 and 2 represented the slope of first and second inflection point on the recession curve will be defined as the recession coefficient related to surface runoff and interflow. 3, the slope after second inflection point is the recession coefficient related to groundwater. The components represent different flow paths in the catchments; each characterized by different residence time, storage volumes and drainage functions.


3.1 Variability of the discharge and soil loss

Annual discharge in three catchments responded in accordance with the variation of rainfall ( Fig. 2 ). The amount of discharge varied depending on the forest type. The highest annual discharge was recorded in the mixed forest (71.5 %) while the lowest was in the coniferous forest (51.0 %), with deciduous forest (65.6 %) between the two forests. Main factors caused to show difference of discharge are thought to an interception loss and transpiration. Annual interception loss of the mixed forest is 18% and 29%, 32% in deciduous and coniferous forest respectively.

Fig. 2. Variations of annual rainfall and discharge in the study catchments for 10 years

The amount of monthly rainfall shows big difference in Fig. 3. Almost rainfall bursts in July to August. In the mixed forest, monthly discharge showed large discrepancy between the dry and rainy season. The mixed forest in the forest structure is still so poor that the amounts of interception loss and transpiration are less than other forest type. Also the soil physical properties in the mixed forest are worse in others. On the other hands, the variation of monthly runoff in the deciduous forest was less than that in the mixed forest. The coniferous forest resulted in least runoff ratio of three forest types for the whole year. The amount of soil loss from the catchment tended to increase in proportion to the amount of rainfall ( Fig. 4 ).

Fig. 3. Variations of monthly average rainfall and discharge in the study catchments for 10 years.

Fig. 4. Relationship between rainfall of wet season and soil loss in the study catchments.

The variation of the amount of soil loss showed wider range because of influences of the amount of soil loss in just previous year. The threshold annual rainfall occurred soil loss estimated about 500 mm. The mixed forest is more vulnerable to soil loss compared to others.

3.2 Lapsed variation of the recession coefficient

Physically based variation in the recession rate is cased by difference in climate during the time of recession, but is also determined by the conditions prevailing prior to the start of the recession ( Tallaksen, 1995). Several workers have recognized a seasonal variation in the recession behavior that follows the seasonal change in evapotranspiration ( Ando et al., 1986; Ambroise, 1988; Brandesten, 1988 ). All kinds of the recession coefficients in this study showed very large the coefficient of variability ranged from 20 % to 50 % during the study period.

Fig. 5 shows the variation of the recession coefficient of surface runoff( 1 ) for 10 years. 1 tended to gradually decrease in the coniferous forest whilst didnt show any tendency in others. That may be caused by the change of forest structure in the coniferous forest after planting. The amount of the initial loss by an interception and transpiration has increased greatly since 1976 as trees grow. Whilst the forest structure in others has rarely changed since 1983.

Fig. 5. Variation of the recession coefficient of surface runoff in the study catchments for 10 years.

The recession coefficient of interflow( 2 ) reduced in the coniferous and mixed forests as time passed ( Fig. 6 ). This can be interpreted by an increase of soil storage capacity after the planting and erosion control work. As the amount of evapotranspiration augments, the storage opportunity of rainfall in the soil increases. Increment of the storage capacity results in delaying the releasing time of interflow from the soil.

Fig. 6. Variation of the recession coefficient of interflow in the study catchments for 10 years.

Fig. 7. Variation of the recession coefficient of groundwater in the study catchments for 10 years.

In case of the recession coefficient of groundwater( 3 ), only mixed forest showed to reduce gradually during 10 years ( Fig. 7 ). The mixed forest has been devastated land until erosion control work finished in 1974. After work, the soil layer formed rapidly and the soil physical properties improved.


The variation of hydrological properties included the discharge, soil loss and recession coefficient showed very wide range depending on the amount of rainfall during the study period. The ratio of monthly discharge varied more in the mixed forest recovered by erosion control work than in the matured deciduous forest. The lapsed variations of the recession coefficient were different depending on the forest conditions. In case of the coniferous forest, a reforestation resulted in the reduction of the recession coefficient for surface runoff and interflow. In the mixed forest recovered by erosion control work, the recession coefficients for interflow and groundwater tended to decrease with the lapse of time.


Ando.Y., Y. Takahasi, T. Ito and K. Ito. 1986. Regionalization of parameters by basin geology for use in a groundwater runoff recession equation. IAHS Publ., 156: 151-159.

Barnes, B.S. 1939. The structure of discharge-recession curves. Trans. Am. Geophys. Union. 20:721-725.

Brandesten, C.O. 1988. Seasonal variation in streamflow recession in the mire complex Komisse, southern central Sweden. In: Hydrol. Of Wetlands and Mans Influence in It, Proc. Int. Symp., June 1988. Publ. Academy of Finland, Helsinki, pp. 84-91.

Linsley, R.K., M.A. Kohler and J. L. Paulhus. 1982. Hydrology for engineers. McGraw-Hill. pp. 206-207.

Tallaksen, L.M. 1995. A review of baseflow recession analysis. J. Hydrol. 165:349-370.