Contribution of mountainous flow to the lateral recharge of downstream sedimentary aquifer

Transcription

1 Contribution of mountainous flow to the lateral recharge of downstream sedimentary aquifer Yu-Hsuan Kao 1, Chen-Wuing Liu 1, *, Tz-Yun Lin 2, Kao-Hung Lin 3 1 Department of Bioenvironmental Systems Engineering, National Taiwan University, Taipei 106, Taiwan, ROC 2 Capital Engineering Corporation, Taipei 106, Taiwan, ROC 3 Research Center for Environment and Resources Management, National Chen Kung University, Tainan, 701, Taiwan, ROC Abstract Groundwater is one of the most important and indispensable water resources. Groundwater consists of 40% of the total water resources utilization in Taiwan. Among 10 groundwater regions, Choushui river alluvial fan is the largest region which supplies billion tons of water annually. The mountainous catchment located in the upstream of Choushui river alluvial fan, and Wuchieh river basin, plays an important role in conveying surface and subsurface runoff to the downstream sedimentary aquifers. The purpose of the study is to estimate the annual groundwater lateral recharge from upstream mountainous area using base-flow and rainfall infiltration methods. Geographic Information Systems (GIS) is applied to facilitate the estimation. The monthly base-flow data of the representative stations of two basins were collected and complied. The amount of groundwater recharge was obtained by multiplying the estimated annual base-flow (m/yr) with the corresponding catchment area (m 2 ). The total lateral recharge estimated from upper mountainous areas was m 3 /yr of which Wuchieh river and Choushui river basins comprised (25.2%) and (74.8%) billion ton/year, respectively. These results are similar to those estimated by O 18 isotopic analysis showing 22% and 78% were from the Wuchieh river and Choushui river basins, respectively. Moreover, the estimate amount using C 14 of the lateral flow in the upstream Choushui river alluvial fan was 0.83 billion ton/year, which is close to our estimation. Groundwater recharged from rainfall infiltration was estimated based on precipitation, evaporation, land use, soil types. The total groundwater recharge obtained by the rainfall infiltration method from upper mountainous area is billion ton/year, which is close to billion ton/year by the base-flow method. These results provide reliable data of the input of boundary conditions for the modeling of subsurface groundwater flow model. 1

2 Introduction Owing to the growth of population and the increasing demand for water resources, groundwater becomes an important and indispensable water resource in Taiwan. About 40% of the total water resources utilization in Taiwan depends on groundwater, which is extensively provided to meet domestic, irrigation, aquacultural and industrial needs. The primary groundwater usage includes agricultural (86.88%), industrial (7.13%) and domestic (5.99%) demands (Water Resources Agency, 2005). However, overuse of groundwater sequentially causes the subsidence, seawater intrusion and soil salinization problems in the downstream coastal area. Estimation of groundwater recharge hence is important for proper management of groundwater systems and for groundwater resources conservation. Based on the hydrogeological setting, groundwater basins can be divided into 10 regions in Taiwan. Among these regions, Choushui river alluvial fan is the largest one which supplies billion tons of water annually. The major recharge area of the Choushui river alluvial fan is eastern mountainous catchment. Groundwater recharge takes place by various processes, namely direct regional recharge by infiltration of rainfall, along watercourses, from lakes, return seepage from irrigation and lateral movement of groundwater through subsurface flow due to a natural hydraulic gradient (Lerner et al.,1990; Scanlon et al., 2002). Many methods can be used to estimate groundwater recharge, such as direct measurements, water-balance methods, Darcian approaches, and tracer techniques (Lerner et al. 1990). Recharge is usually estimated by multiplying the magnitude of water-level fluctuations in wells by the specific yield of the aquifer material or by applying the water budget model or using the water-balance method. Currently, common technique of estimating regional recharge is the application of water-balance model (Finch 1998; Simmons and Meyer 2000; Chen et al. 2005). This approach is generally less intensive computationally, and requires knowledge of the vegetation, soil types within the study area, and a number of basic meteorological variables such as air temperature and precipitation. Although parameters of the water-balance equation, such as precipitation and runoff, are relatively easy to measure, recharge remains an elusive process to quantify. It depends not only on precipitation but also on meteorological conditions, as well as on soil type, soil moisture status, vegetation cover and cultivation practices, and most of all, on evapotranspiration, which is a function of the previously noted factors. Processes of infiltrating recharge and fluxes on the water balance concept also have been discussed (Arnold and Allen 1999; Louie et al. 2000;Otto 2001; Yeh et al. 2004). For groundwater recharge inspecting, estimating the groundwater component of streamflow has been a research focus for more than a century. Following the work of 2

3 Boussinesq (1877), numerous studies have investigated the recession of streamflow, particularly base-flow, and have estimated the contribution of groundwater to streamflow (Bevans 1986; Rutledge 1992; Rutledge and Daniel 1994; Mau and Winter 1997; Chen and Lee 2003). Generally, surface runoff flow and base-flow are assumed to be the two major constituents of streamflow. Moreover, the value of base-flow is assumed to be equal to groundwater recharge in some cases. The base-flow model approach is considered when water falls from the atmosphere to the ground surface. Most rainfall passes through the ground surface and infiltrates into the unsaturated zone, and the rainfall that does not infiltrate will form surface runoff and flow into river courses. The water in an unsaturated zone will vertically infiltrate into the deeper zone of the ground due to gravity. Recharge occurs when water flows past the groundwater level and infiltrates into the saturated zone. The water in the saturated zone will start to flow sideways due to the hydraulic gradient effect and form groundwater flow. Eventually, the groundwater flow will be discharged into river courses and form a base-flow. Base-flow displays spatial and temporal variability due to climate, land use, soils, frequency and amount of recharge, vegetation, topography, and geology (Stuckey, 2006; Delin et al., 2007). The most commonly base-flow method is the base-flow separation for estimating the groundwater discharge and recharge from streamflow records. This method aims at estimating a continuous or daily record of base-flow under the streamflow hydrograph and is expected to be most practical for regional-scale studies where the long-term mean annual value of the spatially variable recharge is of interest (Zektser 1977; Rutledge 1992; Mau and Winter 1997). The purpose of the study is to estimate the annual groundwater lateral recharge from upstream mountainous catchment of the Choushui river alluvial fan using base-flow method, which combines the water-balance model, base-flow-record estimation, and stable-base-flow analysis. Moreover, due to the rapid and direct recharge characteristic of mountainous catchment of the Choushui river alluvial fan, rainfall infiltration model, which is based on widely accepted concepts of soil physics is relatively easy to use. One-dimensional rainfall infiltration model is hence adopted as another recharge estimation method and the estimated result is used to compare with the amount obtained by the base-flow method. Geographic Information Systems (GIS) is applied to facilitate the estimation. The results are expected for providing not only reliable data of boundary conditions of subsurface groundwater flow model, but also helpful information of groundwater utilize management to local government. 3

4 Methodology The large mountainous catchment play an important role of groundwater recharge, which sequentially flows toward downstream of Choushui River alluvial fan. Water balance model include base-flow and rainfall infiltration methods are approaches to estimate the annual groundwater lateral recharge from upstream mountainous area. In this study, the spatial amount of rainfall, streamflow and evaporation of each year, which are linked with different land use and soil type are inputted to geographic information system for groundwater lateral recharge estimation. The results obtained from the base flow and rainfall infiltration methods are compared. 2.1 Study area The area of Choushui River basin is 4,693 km 2, and the length of the Choushui river is about km, that is the largest basin in Taiwan. This area is located within the central region of Taiwan includes Taichung basin and Choushui River plain alluvial. As shown in Fig. 1, the Choushui River alluvial fan is surrounded by the natural geographical boundaries of Taiwan Strait in the west, the Central Mountain Ridge in the east, Wuchieh River in the north, and Choushui River in the south. The annual rainfall in this area is around 2,115 mm, and 55% occurs from May to October (Central Weather Bureau of Taiwan ). The annual runoff in the river is about 6.08 billion tons (Hydrological Year Book of Taiwan 2000). Sediment loads composed of weathered rock fragments of different sizes gradually settled on the riverbed, flood plane, and seabed to form the alluvial fan. It is partitioned mainly into the proximal-fan, the mid-fan and the distal-fan areas (Central Geological Survey. 1999). Based on subsurface hydrogeological analysis to a depth of around 300 m, the formation is divided into six interlayered sequences, including three marine sequences and three non-marine sequences, in the distal-fan and the mid-fan areas. The non-marine sequences of the formation, with coarse sediment, ranging from medium sand to highly permeable gravel, can typically be regarded as aquifers, whereas the marine sequences of the formation with fine sediments, can be regarded as aquitards. The sediments in the Choushui River alluvial fan originated from the rock formation in the upstream watershed, including slate, metamorphic quartzite, shale, sandstone, and mudstone (Fig. 2). The hydrogeological formation of the proximal-fan, which consists entirely of gravel and sand, is considered to be an unconfined aquifer and an important groundwater recharge area where the mountainous catchment located in the upstream of Choushui river alluvial fan, is our study areas. 4

5 2.2 Base-flow model A water balance model was presented by Kulandaiswamy et al. (1969). River flow Q SF is assumed to be sum of surface flow Q SR and base flow Q BF ( Q SF = Q SR + Q BF ). According to the known surface flow, groundwater recharges in a watershed can be analyzed by base flow. The base flow is considered with river flow and flow hydrograph which are based on the conditions of rivers in a watershed and recharges and discharges of groundwater. Cherkauer and Ansari (2005) pointed out that base-flow separation involves developing a simple groundwater budget equation, in general, be written as I + GW = Q + GW + ET + NP + ΔS t (1) in bf out / where I is infiltration to the system, GW in is the groundwater influx to the watershed through an aquifer, Q bf is the groundwater discharge to the streamflow base-flow, GW out is the groundwater efflux from the watershed through an aquifer, ET is the evapotranspiration losses from the watershed, NP is the net pump of groundwater by people into or out of the watershed, and ΔS/t is the rate of change of groundwater storage with respect to time. If watersheds can be selected where GW in = GW out = NP =ΔS/t = 0, and if recharge is defined as net groundwater recharge (I ET), then Eq. (1) can be reduced to Recharge = Net recharge = Q bf = Streamflow base-flow (2) The conceptual diagram of groundwater recharge from upper mountainous area to downstream plain is established according to Eq. (2) (Fig. 3). The groundwater recharge amount is estimated by the base-flow separation, which is associated with the interaction of river and groundwater. The concept of the base-flow model is to use the base-flow separation from the total streamflow discharge to obtain a measure of groundwater recharge. This will establish the relationship between the rate of recharge and the controlling physical properties of watersheds. The base-flow estimation explained here is a form of streamflow partitioning. Rutledge (1992) first developed this method that the principles of this method are as follows: (1) Daily data of streamflow is required, and (2) Linear interpolation is used to estimate groundwater discharge during the period of surface runoff. There are some basic assumptions of this model: (1) surface runoff and base flow are the two main components of streamflow; interflow is out of considered; (2) the groundwater table is invariable; natural factors such as evapotranspiration, precipitation, and human-induced factors do not affect the water level; (3) the evapotranspiration of saturated zone approaches zero; (4) the aquifer is underlain by impermeable material; its side boundaries are 5

6 vertical and have no flow crossing them. For most applications, it should be reasonable to assume that nearly all of the groundwater in the watershed discharges into the stream, except for that which is lost by evapotranspiration or human activity. The area of contribution in the groundwater system is equal to the surface drainage area for the purpose of expressing flow in units of specific discharge (Rutledge 1998). To increase the speed of analysis and reduce the subjectivity inherent in manual analysis, Rutledge (1993) proposes several computer programs: RECESS, RORA, and PART, and newer versions have been proposed (Rutledge, 1998; 2000). A computer program for base-flow estimation, which named PART, was designed by U.S. Geological Survey (USGS) (Rutledge, 1992). To avoid the estimation inaccuracies which are caused by the application of PART on mountains catchment, the most complete streamflow data of 4 stations are collected to determine the groundwater recharge, including (1)Wu-Chieh Bridge, (2)Nan-Kang Bridge, (3)Yen-Pong Bridge, and (4)Yu-Feng Bridge. The drainage areas and the time periods of these 4 stations are shown in Table 1. According to daily streamflow, the quantity of groundwater recharge is evaluated by PART with the estimation of the quantity of river discharge. The estimation procedures of PART are as follows (Fig. 4): 1. Ranking daily flow data from streamflow gauging stations and supplementing deficient data. 2. Calculation of monthly averaged base-flow of each streamflow gauging stations. 3. Accumulation of monthly averaged base-flow. Four minimum values are collected for regression to obtain annually stable base-flow. 4. Estimation of groundwater recharge of each stations by following equation: Groundwater recharge (tons/yr)=stable base-flow(m/yr) Drainage area(m 2 ) (3) The stable base-flow analysis using PART programs in this study to obtain a more reliable result, which results is determined from the complete and consecutive daily flow data. The stable base-flow estimation aims to separate the base flow from the streamflow hydrograph in order to evaluate the discharge drained from the groundwater to streams. The method is based on the assumption that base flow (groundwater discharge) is equal to streamflow on days that fit a requirement of antecedent recession. 2.3 Rainfall infiltration method The rainfall infiltration method considers the vertical infiltration processes of surface water or precipitation, which are effected by land use, soil type and vegetation 6

7 cover. In this study, the amount of rainfall and evaporation over each year, different land use and soil type are considered to evaluate the amount of groundwater recharge by GIS and software of Surfer and Matlab. The procedure of recharge estimation by this method is shown in Fig. 5. According to the result of Water Resources Agency (2006), the averaged daily evapotranspiration is 3.96 mm/day. When precipitation is less than 3.96 mm/day, the infiltration are assumed to be 0; When precipitation is greater than 3.96 mm/day, the precipitation can be regarded as effective rainfall. Under the condition of that the effective rainfall is less than saturated soil infiltration, infiltration can be estimated with following equation: Q=0.001 P A α (4) where Q is amount of infiltration; P is the effective rainfall (mm/day); A is the area of each cell (500m 500m); α is infiltration ratio of different land use. If the effective rainfall is greater than saturated soil infiltration, infiltration can be estimated with following equation: Q=0.001 φ A α (5) where φ is the rate of saturated soil infiltration (mm/day). 3. Results and discussion 3.1 Base-flow estimation Figure 6 shows relationships between streamflow and base-flow analyzing by 4 streamflow gauging stations. Figure 6, shows four hydrographs from 4 stations that indicate streamflow partitioning for the drainage area and the time base of surface runoff. In Fig. 7, groundwater discharge is equal to streamflow when the curves coincide. According to PART calculation and Eq. (3), the stable base-flow and estimated annual groundwater recharge of 4 streamflow gauging stations is listed in Table 2. The total lateral recharge estimated from upper mountainous areas was ton/yr where Wuchieh river basin, where is consisted of stations (1) and (2), and Choushui river basin, where is consisted of stations (3) and (4), comprised (25.2%) and (74.8%) billion ton/year, respectively. Liu (1996) indicated that the annually groundwater recharge in the upstream of Choushui river alluvial fan was about 0.83 billion ton, using C 14 tracer and tritium dating techniques. This result is close to our estimation. Chiang et al., (1996) estimated the groundwater recharge percentages of recharge source, including Wuchieh river, Choushui river, terrace deposit and precipitation, by mass balance analysis for the oxygen isotopic 7

8 compositions. The results show that 4.4% of groundwater is recharged from Wuchi river, 5.2% from northeastern terrace deposit, 34.0% from Chouchui river. In the another word, the recharge percentage of Wuchi river and Chouchui river were 22% and 78%, respectively, which were similar to our estimation. Therefore, the base-flow method is a simple and rapid approach for groundwater recharge estimation. Moreover, the drainage areas of Wuchieh river basin, where is consisted of stations (1) and (2), and Choushui river alluvial basin, where is consisted of stations (3) and (4), are and km 2, respectively. The total area is km 2, which covered 85% of the upstream of Choushui river alluvial fan ( km 2 ). The estimation of groundwater recharge from mountainous catchment is representative for groundwater recharge of Choushui river alluvial fan. 3.2 Rainfall infiltration method The data of annually precipitations, evapotranspiration, land use, and soil type is integrated by GIS, Surfer and Matlab. According to the rainfall records of Water Resources Agency of Taiwan, data from 16 precipitation gauging stations in upstream of Choushui River alluvial fan was collected between 2000 to The extent of each station was divided by Thiessen's Polygon Method (Fig. 8). The monthly evapotranspiration data from 2000 to 2005 was collected from Central Weather Bureau of Taiwan (Table 3). Annually and daily averaged evapotranspiration in upstream of Choushui River alluvial fan is mm and 2.18 mm, respectively. Figure 9 shows that the agricultural operation is the main land use of Choushui River alluvial fan. In general, land use can be particularly classified into five types, including paddy field, dry farmland, forest, grassland and non-infiltrated land. Chow et al. (1988) indicated that runoff coefficient was affected by land use type. The runoff coefficients used in this study are 0.45 for farm land and forest, and 0.44 for grassland. Moreover, the grids were set up with 500m 500m in 16 districts for mapping by geostatistic function of GIS. According to Eqs.(4) and (5), annually groundwater recharge estimated by rainfall infiltration method (Table 4), is an average of billion ton that results of the rainfall infiltration method from 2000 to 2005 show Fig 10. This result is close to billion ton/year by the base-flow method. However, the considered region of base-flow method is different with that of rainfall infiltration method. Hence, the estimated groundwater recharge of upstream mountainous catchment of the Choushui River alluvial fan can be regarded as that of whole region. Conclusions Mountainous areas, the upstream of Choushui river alluvial fan, commonly have frequent precipitation; therefore, streams draining basins do not have frequent or long 8

9 groundwater recessions. The estimated groundwater recharge of Wuchieh river by the streamflow records of Wuchieh Bridge and Nan-Kang Bridge stations is billion ton per year. It is the primary lateral recharge source of the downstream Taichung basin. Moreover, the estimated groundwater recharge of Choushui river by the streamflow records of Yan-Ping Bridge and Yu-feng Bridge stations is billion ton per year, and can be regarded as main lateral recharge source of the downstream Choushui River alluvial plain. The covered area of four streamflow gauging stations is about km 2, which is 85% of upper mountainous areas ( Km 2 ), and is representative for groundwater recharge estimation. The results are similar to other groundwater recharge estimations by O 18 isotopic analysis and C 14 and tritium dating/trace. The rainfall infiltration is also used in this study to estimate groundwater recharge from precipitations, which resulted billion ton per year of recharge. Moreover, the result of rainfall infiltration method is close to that of base-flow estimation. The result of groundwater recharge estimation can be used in simulation of boundary in the groundwater lateral recharge of Taichung basin and Choushui river plain alluvial. References Arnold, J.G., Allen, P.M., Automated methods for estimating baseflow and ground water recharge from streamflow records. Journal of the American Water Resources Association 35, Bevans, H.E., Estimating stream aquifer interactions in coal areas of eastern Kansas by using streamflow records. In: Subitzky, Seymour (ed.) Selected papers in the hydrologic sciences. U.S. geological survey water-supply paper 2290, Boussinesq, J., Essa sur latheories des eaux courantes. Memoires presentes par divers savants a l Academic des Sciences de l Institut National de France. Tome XXIII, no. 1. Central Geological Survey, Project of Groundwater Monitoring Network in Taiwan during First Stage-Research Report of Chou-Shui River Alluvial Fan, Taiwan. Taiwan Water Resources Bureau. Central Weather Bureau, Year book of climate. Central Weather Bureau, Taiwan. Chen, J.F., Lee. C.H., Yeh, T.C.J., Yu, J.L., A water budget model for the Yun-Lin plain, Taiwan. Water Resources Management 19, Cherkauer, D.S., Ansari, S.A., Estimating ground water recharge from topography, hydrogeology, and land cover. Ground Water 43, Chiang,C.J., Chen J.E., Lai, T.H., Huang, C.C., A study of groundwater charge 9

10 sources in the Choshuichi groundwater basin. Proceeding of 2nd Conference on Resources Engineering in Taiwan Chow, V.T., Maidment, D.R., Mays, L.W., Applied Hydrology, McGraw-Hill, New York. Delin, G.N., Healy, R.W., Lorenz, D.L., Nimmo, J.R., Comparison of local to regional scale estimates of ground water recharge in Minnesota, USA. Journal of Hydrology 334, Finch, J.W., Direct groundwater recharge using a simple water-balance model-sensitivity to land surface parameters. Journal of Hydrology 211, Hydrological Year Book of Taiwan, Water resources (in Chinese). Bureau Ministry of Economic Affairs, Republic of China Ineson J, Downing RA, The ground-water component of river discharge and its relationship to hydrogeology. Journal of Institutional Water Engineering 18, Kulandaiswamy, V.C., Seetharaman, S., A note on Barnes method of hydrograph separation. Journal of Hydrology 9, Lerner, D.N., Issar, A.S., Simmers, I., Groundwater recharge: a guide to understanding and estimating natural recharge. International Association of Hydrogeologists 8, Heinz Heise, Hannover. Liu, T.K., Tyan, C.L., Chiu, T.H., Chang, Y.M., Groundwater resource in Choushuichi Fan-Delta revealed by radiocarbon and tritium dating/tracer studies, Proceedings of Symposium on Groundwater and Hydrogeology of the Choushuichi Alluvial Fan, Taipei, Taiwan Louie, M.J., Shelby, P.M., Smesrud, J.S., Gatchell, L.O., Selker, J.S., Filed evaluation of passive capillary samplers for estimating ground water recharge. Water Resources Research 36, Mau, D.P., Winter, T.C., Estimating ground-water recharge from streamflow hydrographs for a small mountain watershed in a temperate humid climate, New Hampshire, USA. Ground Water 35, Otto, R., Estimating ground water recharge rates in the southeast Holstein region, northern Germany. Hydrogeology Journal 9, Rutledge, A.T., Methods of using streamflow records for estimating total and effective recharge in the Appalachian Valley and Ridge, Piedmont, and Blue Ridge physiographic provinces. In: Hotchkiss WR, Johnson AI (eds) Regional aquifer systems of the United States, aquifers of the southern and eastern states. American Water Resources Association Monograph Series 17, Rutledge, A.T., Computer programs for describing the recession of groundwater discharge and for estimating mean ground-water recharge and discharge from streamflow records. United States geological survey. Water 10

11 resources investigations report , 45. Rutledge, A.T., Daniel, C.C., Testing an automated method to estimate ground-water recharge from streamflow records. Ground Water 32, Rutledge, A.T., Computer programs for describing the recession of groundwater discharge and for estimating mean ground-water recharge and discharge from streamflow records update. United States geological survey water resources investigations report , 43. Rutledge, A.T., Considerations for use of the RORA program to estimate ground-water recharge from streamflow records. United States geological survey open-file report , 44. Scanlon, B.R., Healy, R.W., Cook, P.G., Choosing appropriate techniques for quantifying groundwater recharge. Hydrogeology Journal 10, Simmons, C.S., Meyer, P.D., A simplified model for the transient water budget of a shallow unsaturated zone. Water Resources Research 32, Stuckey, M.H., Low flow, base flow, and mean flow regression equations for Pennsylvania streams. United States Geological Survey Scientific Investigations Report Water Resources Agency, Groundwater master plan. Environmental & Infrastructural Technologies, Incorporation. Water Resources Agency, Reviewing and Revising Survey of Reasonable Amount to Agricultural Irrigation Water. Center for Water Resources Management and Policy Research. Yeh, H.F., Chen, J.F., Lee, C.H., Application of a water budget model to evaluate rainfall recharge and slope stability. Journal of Chinese Institution Environmental Engineering 14, Zektser, I. S., Laws of formation of groundwater runoff study (in Russian). Nedra, Moscow,

12 Table 1 List of analyzed streamflow gauging stations. Upper mountainous area Drainage Time period analyzed No. Station River basin area (km 2 ) (year) 1 Wu-Chieh Bridge Wu River Nan-Kang Bridge Wu River Yen-Pong Bridge Choushui River Yu-Feng Bridge Choushui River Table 2 Result determined from the method of base-flow estimation. Station Station name Stable Annual map base-flow groundwater number (cm/year) recharge (billion tons) 1 Wu-Chieh Bridge Nan-Kang Bridge Yen-Pong Bridge Yu-Feng Bridge Sum Table 3 The monthly evapotranspiration data from 2000 to 2005 was collected from Central Weather Bureau of Taiwan Time period Average Jan Feb Mar Apr May Jun Jul Aus Sep Oct Nov Dec Sum (Unit : millimeter) 12

13 Fig. 1. Study area. Wuchieh River and Choushui River are the major river of Taichung basin and Choushui River alluvial fan, respectively. Solid circles represent streamflow gauging stations. Data from stations 1-4 is collected for recharge estimation in 2.2. Fig. 2. Regional geological map of the Choushui River alluvial fan (modified from CGS 1999) 13

14 Fig. 3. Conceptual diagram of groundwater recharge from upper mountainous area to downstream alluvial plain. (Not to scale) Collection of the hydrological data Streamflow, precipitation and evapotranspiration Base-flow separation Separating streamflow into surface runoff and base-flow Stable base-flow calculation Calculating base-flow index Application of water balance model Calculation of groundwater recharge amount of each cells by GIS Estimation of groundwater recharge of the Choshui River alluvial fan Fig. 4. Flowchart of groundwater recharge estimation by base-flow separation method. 14

15 Land use/land cover Soil type Rainfall record Establishment of cells (each of cell is 500meter 500 meter) Calculation of infiltration of each cell Daily total rainfall <3.96 mm (including no rainfall) Daily effective rainfall < soil saturated infiltration Daily effective rainfall > soil saturated infiltration Infiltration =0 Calculation of infiltration using effective rainfall (Eq. (4)) Calculation of infiltration using soil saturated infiltration (Eq. (5)) Calculation of annually surface infiltration of each cell Estimating groundwater recharge of surface infiltration Fig. 5. Flowchart of groundwater recharge estimation using rainfall infiltration method. Depths of annual infiltration (cm) Wu-Chieh Bridge Station Depths of annual infiltration (cm) Nan-Kang Bridge Station Month Month Depths of annual infiltration (cm) Yen-Pong Bridge Station Month Depths of annual infiltration (cm) Yu-Feng Bridge Station Month Fig. 6. Results of stable base-flow analysis of 4 streamflow gauging stations. 15

16 Discharge (cm) Wu-Chieh Bridge Station Daily mean streamflow Estimated daily mean groundwater discharge Jan Feb Mar Apr May Jun Jul Aus Sep Oct Nov Dec (yr) Discharge (cm) Nan-Kang Bridge Station Daily mean streamflow Estimated daily mean groundwater discharge Jan Feb Mar Apr May Jun Jul Aus Sep Oct Nov Dec (yr) Discharge (cm) Yen-Pong Bridge Station Daily mean streamflow Estimated daily mean groundwater discharge Jan Feb Mar Apr May Jun Jul Aus Sep Oct Nov Dec (yr) Discharge (cm) Yu-Feng Bridge Station Daily mean streamflow Estimated daily mean groundwater discharge Jan Feb Mar Apr May Jun Jul Aus Sep Oct Nov Dec (yr) Fig. 7. Results of streamflow partitioning of 4 streamflow gauging stations. Fig. 8. Mapping of 16 precipitation data analyzed by Thiessen's Polygon Method. 16

17 Land use type Industrial land Irrigation&Drainage land Transportation land Other land Construction land Protection &conservation land Agricultural land Recreation land Mining industry land Fig. 9. Land use type map of the upstream mountainous catchment of Choushui River alluvial fan. 17

18 The amount of rainfall infiltration (m 3 /yr) The amount of rainfall infiltration (m 3 /yr) The amount of rainfall infiltration (m 3 /yr) Fig. 10. Result of rainfall infiltration of the upstream mountainous catchment of Choushui River alluvial fan from 2000 to