Environ. Process. DOI 10.1007/s40710-017-0258-3 O R I G I N A L A RT I C L E

Impact of Abandoned Opencast Mines on Hydrological Processes of the Olidih Watershed in Jharia Coalfield, India V. T. Shinde 1 & K. N. Tiwari 2 & M. Singh 1 & B. Uniyal 3

Received: 14 April 2017 / Accepted: 16 July 2017 # Springer International Publishing AG 2017

Abstract The Olidih watershed hydrology was affected by opencast mines for the past five decades. This study explores the potential hydrological effect of these mines using Soil and Water Assessment Tool (SWAT2012). The calibration and validation of the model was performed using daily streamflow and sediment yield data (2005–2008) at the outlet of the water shed. The model performed satisfactorily during simulation when tested with statistical indicators. The alternative scenario of no-mines was also modelled to assess the potential impact of abandoned opencast mines for the period 2005–2010. Results show that the abandoned opencast mines play a crucial role in altering hydrological processes of the watershed with 16% increase in the annual sediment yield and reduction of 51% and 6% in annual surface flow and water yield, respectively. This may be due to surface soil disturbance and accumulation of surface runoff in large depressions that resulted in less surface runoff and 13% more groundwater flow. The contribution of this analysis is the application of SWAT in modelling potential hydrological effect of abandoned opencast mines by defining large opencast mines as pothole during simulation. Keywords Opencast mines . Hydrological response . SWAT . Scenario analysis

1 Introduction Mining is one of the harmful anthropogenic activities performed in the past decade whose effect can be easily seen in the atmospheric variables at local or regional level (Wan et al. 2013). Mining basically involves the exploration and extraction of minerals from the earth

* V. T. Shinde [email protected]

1

Navsari Agricultural University, Navsari, Gujarat -396 450, India

2

Agricultural & Food Engineering Department, IIT Kharagpur, Kharagpur -721302, India

3

Institute for Water Management, Leibniz University Hannover, Hannover, Germany

Shinde V.T. et al.

crust. However, mining activities may be associated with certain permanent and irreparable damages to the geo-environment (Bhakdisongkhram et al. 2007). Mining changes the hydrology of a catchment by affecting its groundwater flow processes (Hawkins 2004) followed by influencing the groundwater levels, inflow and its recharge. Due to changes in groundwater flow components, streamflow are directly influenced (Siddle et al. 2000; Xiao et al. 2014). Therefore, it is really important to assess the effect of change of the landuse and land cover in the catchments on hydrological processes (Van Roosmalen et al. 2009). Usually the mining effects are long-term which can be usually seen in the form of floods, drought risks and water resources problems (Dvoracek et al. 2004). Thus, the behaviour of opencast mines (physical, chemical and biological) is inextricably linked with the hydrology of the catchment as initial behaviour of the catchment is completely changed by them (Choubey and Rawat 1991). Jharia is one of the largest coal fields in India covering about 460 km2 of area. Opencast and underground mining methods are used to extract coal from the mines which had adversely affected the surrounding environment (Saini et al. 2015). Apart from the normal water quality and quantity issues, this area is also prone to open as well as underground fires and suspended particulate matter which had worsened the air and soil quality of the respective area (Michalski 2004). Therefore, it is really essential to continuously monitor the connected ecosystems as far as the mining activities are concerned to favour a sustainable approach in managing the available natural resources. Many research studies have been reported to evaluate the influence of opencast mines using different assessment techniques, including remote sensing data (Jhanwar 1996) and continuous monitoring of flow and heavy metals or sediments in the river (Yellishetty et al. 2013). Also, the development of hydrological models are used to mimic the current and past conditions (Wan et al. 2013; Gálvan et al. 2009) for environmental impact assessment and monitoring of mining activities, as the rest of the two methods of monitoring, i.e., field investigations and the use of satellite imagery, are cumbersome, time consuming and expensive. Therefore, hydrological models are the best option for simulating the hydrology of abandoned opencast mines. Abandoned opencast mines show very similar behaviour to that of karst aquifers due to their highly heterogeneous nature (Burbey et al. 2000). SWAT has been used in this research in modelling the hydrology of the study area which comprises 30% agriculture and nearly 20% of mined area. SWAT justifies both conditions as it is a good model for simulating the hydrology of agricultural catchments (Neitsch et al. 2011) and also it has been applied to many studies concerning karst aquifers (Amatya et al. 2011). In the past years, very few researchers have used SWAT for simulating the effect of underground mines on hydrological parameters. For example, Gálvan et al. (2009) and Olias et al. (2011) used SWAT to assess the transportation of pollutant load from acid mine drainage coming into the river, whereas Wan et al. (2013) investigated hydrological processes in abandoned underground mines in Ohio, USA, using SWAT. Significant reduction in surface runoff due to underground coal mining in Kuye River basin of Northern China was observed by Li et al. (2016) using statistical approaches and SWAT model. For Indian conditions, no study on application of the SWAT model to assess the effect of mines on runoff and sediment yield could be found. Furthermore, no study in the past has reported the application of SWAT model in simulation of hydrological behaviour in opencast mine areas, as required for reclamation and sustainable development of watersheds with mines. Hence, the aim of present analysis is to model the hydrologic response of opencast mines by defining large mine pits as potholes during SWAT simulation, which has not been studied in the past. Modelling results corresponding to the mined and unmined scenarios for different water balance components have also been evaluated.

Impact of Abandoned Opencast Mines on Hydrological Processes of ...

2 Study Area Considering the availability of quantity and quality of meteorological, hydrological, soil and other relevant data, the study was undertaken in a small watershed, namely Olidih (5512 ha), which falls in Jharia coalfield (JCF) region of India (Fig. 1). JCF is India’s most important storehouse of prime coking coal that feeds most of industrial demand. The mining activities in Jharia started in the end of the nineteenth century, were intensified in the 1920s, and thereafter, they have been growing extensively, almost exponentially. At present, there are around 35 large underground and opencast mines in the JCF (Saini et al. 2015). 20% of the area of the Olidih watershed is affected by opencast mines. JCF is situated between Latitudes 23°39′ to 23°48′ N and Longitudes 86°16′ to 86°27′ E along the north of Damodar river. The elevation in the watershed varies from 157 to 240 m above MSL. The climate of the study area is subtropical subhumid, with mean annual rainfall of 1250 mm during the years 2005 to 2009, comparable with 1256 mm reported by Tripathi et al. (2003). Southwest monsoon remains active from June to October which contributes 80% of the total rainfall. The streamflow and sediment yield data of Olidih gauging station, located near the watershed outlet (Fig. 1), for years 2005–2008 were collected from the Soil Conservation Department, Damodar Valley Corporation (DVC) Hazaribagh, India. The average maximum temperature recorded during April and May varies from 37 °C to 41 °C. The average minimum temperature ranges between 7 °C and 10 °C which is usually recorded during the months of December and January. Joriya river flowing through this area is affected by the surrounding mining activities. Paddy is the major crop grown in this region followed by maize. The satellite data of Landsat (Sensor- ETM+, Path/Row- 140/43) were used to map various coal mines and their spatial extent by visual interpretation technique using onscreen digitization in ArcGIS software. Water, forest, agriculture, barren, urban and mine areas were mapped for year 2005 for model development.

3 Methodology 3.1 Model Inputs and Set-up Specific data related to topography, soil properties, climate, land cover and vegetative information is required in SWAT model for obtaining parameters which regulate the hydrological processes of the catchment. Data used for the model development are well elaborated in Table 1. The Olidih watershed was divided into seven sub-basins based on spatial heterogeneity, which were further discretized into 84 hydrologic response units (HRUs) on the basis of unique landuse, soil and slope combinations. The SWAT model simulates various parameters initially at HRU level and then for subbasin by using weighted mean. Climatic data and other physical characteristics of the watershed are considered for each sub-basin. Twenty soil samples, equally and spatially distributed in mine and no-mine areas of the study watershed were collected and analyzed in the soil testing laboratory of Agricultural and Food Engineering Department of the IIT Kharagpur, India. Two new soil classes were defined in SWAT database using the analyzed soil properties and data collected from National Bureau of Soil Survey and Land Use Planning (NBSS & LUP), Kolkata, India.

Shinde V.T. et al.

Fig. 1 Location of the study area

3.2 Pothole Representation and Baseline Model Calibration/Validation In SWAT, impact of depressions on hydrologic simulation can be modelled using Potholes, Ponds, and Wetlands. Many researchers in the past have considered it as wetlands and ponds during simulation in SWAT (Almendinger et al. 2014). However, ponds and wetlands do not represent the characteristics of mine pits, as these pits are defined at sub-basin level and linked

Impact of Abandoned Opencast Mines on Hydrological Processes of ... Table 1 Data required for the model simulation Sl. No.

Data

Scale/Duration

Source

1 2

DEM Land use/land cover

30 m 30 m

3

Soil map

1: 2,50,000

4

Rainfall

2005–2010

5

Temperature, Wind velocity and Relative humidity Streamflow and sediment yield

2005–2010

ASTER Landsat (Sensor- ETM+, Path/Row- 140/43) Global Land Cover Facility website NBSS & LUP, Nagpur (Two new soil classes were defined into the SWAT soil database) Soil Conservation Department, DVC Hazaribagh, India Soil Conservation Department, DVC Hazaribagh, India For Olidih gauging station from Soil Conservation Department, DVC Hazaribagh, India

6

2005–2008

to the channel routing process of SWAT (Almendinger et al. 2014; Evenson et al. 2015). The Olidih watershed has a special topography due to active opencast mining activities from the past several years. As a result, the head catchment has many small depressions which must be included in the hydrological simulation and are referred to as potholes. Runoff generated within these areas is accumulated in potholes rather than flowing into the streams. Pothole needs to be defined at HRU level (Neitsch et al. 2011) where whole HRU or a fraction of a HRU can be defined as a pothole, i.e., in the model IPOT (.hru) (Number of HRU defined as a pothole) must be set to HRU number. In addition to the basic model inputs (topography, soils, land use, and weather), some other parameters (Table 2) pertaining to potholes are required to consider during pothole definition. Some of these parameters were estimated (SLOPE, POT_FR, POT_VOLX) and some were taken from predefined database in SWAT as well as from available literature (EVLAI, POT_TILE). In the Olidih watershed, large opencast mines (20–45 m depth) are present in sub-basins 3, 4 and 5. The effects of these mines were incorporated by defining HRUs as

Table 2 Parameters of pothole in SWAT model Variable name IPOT

Definition

Number of HRU that is impounding water (that contains the pothole) Variables in release/impound operation line: MONTH/DAY Timing of release/impound operation. MGT_OP Operation code. (13 for release/impound operation) IMP_TRIG Release/impound action code (0: impound, 1: release) SLOPE Slp: Slope of the HRU (m/m) POT_FR frpot: Fraction of the HRU area draining into the pothole EVLAI LAIevap: Leaf area index at which no evaporation occurs from the water surface POT_VOLX Vpot,mx: Maximum amount of water that can be stored in the pothole (mm). It is the maximum ponding water depth of the pothole. POT_TILE qtile: Average daily tile flow rate (mm)

File name .hru

.mgt .mgt .mgt .hru .hru .bsn .hru

.hru

Shinde V.T. et al.

potholes (Table 3) during SWAT model set up. The effect of other small mines was considered by land use and slope categories. After successful representation of the potholes, the developed SWAT was calibrated at the outlet of the Olidih watershed for streamflow and sediment yield using SWATCUP (SWAT Calibration Uncertainty Procedures). The actual condition, i.e., mine affected condition was considered as baseline condition and scenario of no-mines (unmined) was modelled to assess the potential impact of abandoned opencast mines for the period 2005–2010. The model calibration was performed for the year 2005 and 2006 and validation for 2007 and 2008. In SWATCUP, parameters can be manually adjusted iteratively between autocalibration runs. Sequential Uncertainty Fitting Version 2 programme (SUFI-2), a semiautomated procedure linked to SWATCUP, was used for model calibration (Narsimlu et al. 2015; Abbaspour et al. 2007). SUFI2 algorithm helps to identify the most sensitive model parameters before calibration, which would have significant effect on the model output (Abbaspour 2008). The SUFI-2 has been widely used because of its easy functioning and less iterations to attain good model simulation (Yang et al. 2008). The first step in calibration using SUFI2 is to define the method followed by defining upper and lower limit of parameters. All parameters are assumed to be uniformly distributed within the basin. Sensitivity analysis was performed by altering each parameter within a specific range.

3.3 Alternative Scenario Assuming No-mines Scenario analysis is an effective technique to predict the response of hydrological process due to variation in actual field conditions such as change in land use/cover or climatic conditions of an area (Menzel et al. 2009). Scenarios are not predictions or forecasts; rather, they are Bplausible and often simplified descriptions of how the future may develop based on a coherent and internally consistent set of assumptions about driving forces and key relationships^ (Houghton et al. 2001). Therefore, in this study the potential hydrological effect of mines was accessed by simulating the calibrated and validated SWAT model from 2005 to 2010 using mined and unmined scenarios. The unmined scenario was established using three criteria for sub-basins with mine area. a) Potholes from mine affected HRUs were removed from the model setup b) Soil properties of mined HRUs were replaced by soil properties of unmined area, and c) Mined landuse of the HRUs was replaced by forest, barren land and agriculture based on slope conditions and other ground truth factors. Table 3 Description of defined potholes Sub- watershed

HRU No.

Description

Area (ha)

Sub- watershed area (%)

Pothole fraction (POT_FR)

SW3 SW3 SW4 SW4 SW4 SW4 SW5 SW5

25 28 41 42 43 44 53 54

SWRN/SL 168/4–9999 SWRN/SL 80/4–9999 SWRN/SL 168/4–9999 SWRN/SL 168/0–4 SWRN/SL 80/4–9999 SWRN/SL 80/0–4 SWRN/SL 80/0–4 SWRN/SL 80/4–9999

220.40 88.99 55.21 63.22 184.25 165.55 103.92 57.83

16.54 6.68 4.18 4.78 13.94 12.52 15.21 8.47

0.62 0.38 0.13 0.60 0.75 0.28 0.20 0.12

Impact of Abandoned Opencast Mines on Hydrological Processes of ...

During the modelling of the baseline scenario (mined), the soil properties of mine area were used and for no-mines scenario (unmined) soil properties from no-mine areas were used in SWAT. After scenario development, the soil/land use/slope map was reclassified in SWAT model setup for generation of new HRUs by considering changes as per the unmined scenario.

4 Results and Discussion 4.1 Sensitivity Analysis, Model Calibration and Validation Parameterization is essential for the SWAT model because it plays a crucial role during simulation of hydrological parameters and soil erosion (Gong et al. 2010). In this study, the SWAT2012 model was first run from 2005 to 2006 to identify the sensitive parameters for streamflow and sediment transport on a daily time step. The sensitivity analysis was done separately for streamflow and sediment transport, since some parameters are sensitive to both streamflow and sediment transport, whereas some can be sensitive to just one (Abbaspour 2007). The sensitivity rank and mean sensitivity index of parameters are presented in Table 4. For initial model run, three major problems were observed in the water balance of the shallow aquifer: a) high surface runoff; b) low lateral flow; and c) high base flow (inter flow or return flow). To overcome these problems and to get better simulation results, the sensitive parameters were adjusted during model calibration, as shown in Table 5. The calibrated value of ALPHA_BF (0.5) indicates a delayed response of catchment towards groundwater flow to recharge. The existing value of CN2 was reduced by 15% in all HRUs, which decreased the surface runoff, thereby increasing the infiltration rate. For Ch_N2 (0.05), its calibrated value is the typical value of Manning’s roughness coefficient for a natural stream (Neitsch et al. 2005). Also, the SOL_AWC was increased by adding 0.03 to its default value to decrease the movement of water in the soil profile. However, the groundwater parameter ESCO (0.37) shows that the model extracts most of the water from the upper soil layers. GWrevap coefficient (0.10) is a critical parameter which affects the movement of water from the shallow aquifer to the root zone. When the GWrevap coefficient increased, slight decreases in simulated base flow and total streamflow were observed. The relatively low calibrated value of GW_DELAY (8 days) indicated a faster movement of water, exiting the soil profile to reach the shallow aquifer. Apart from it, the calibrated value of SURLAG (3 days) suggests that much of the generated surface runoff reaches the outlet in relatively short time. The systematic and dynamic behaviour of the model was visualized by plotting the observed and simulated data on a daily time step. The model performance was fairly satisfactory during calibration, with coefficient of determination (R2), Nash-Sutcliffe efficiency (NSE) and root mean square error (RMSE) calculated between the best simulation and the observed data of 0.74, 0.61 and 0.50, respectively, for streamflow, and 0.69, 0.56 and 2.56, respectively, for sediment yield (Fig. 2). During the calibration and validation of SWAT model, an NSE value more than 0.50 indicates satisfactory performance of the model, an NSE value from 0.54 to 0.65 indicates adequate performance of the model and a value of NSE more than 0.65 signifies very good performance of the model (Saleh et al. 2000; Bracmort et al. 2006; Moriasi et al. 2007). The value of NSE for streamflow (0.61) and sediment yield (0.56) indicates that the model adequately simulates the streamflow and sediment yield during calibration. The model underpredicted the streamflow peaks at few time steps, which might be due to the inconsistency of rainfall events or some localized event which was not captured by the

Shinde V.T. et al. Table 4 Sensitivity ranks and mean sensitivity index of streamflow and sediment parameters Streamflow

Sediment

Parameter

Rank Mean sensitivity Index

Parameter

Rank Mean sensitivity Index

GWQMN.gw (Threshold depth of water in shallow aquifer required for the occurrence of baseflow) CN2.mgt (Curve Number) ESCO.hru (Soil evaporation compensation factor) ALPHA_BF.gw (Base flow factor, 1/days) SOL_AWC.sol (Available water capacity) SOL_Z.sol (depth of soil profile)

1

1.23

CN2.mgt (Curve Number)

1

2.2200

2

0.2790

2

2.1600

3

0.2130

USLE_P.mgt (USLE practice support factor) SURLAG.bsn (Surface runoff lag coefficient)

3

2.1300

4

0.1350

4

1.6100

5

0.0895

5

0.8350

6

0.0609

6

0.3120

GW_REVAP.gw (Groundwater r evap coefficient)

7

0.0566

7

0.2450

REVAPMN.gw (Threshold depth of water in shallow aquifer for revap to occur) GW_DELAY.gw (Groundwater delay, days) CH_K2.rte (Channel hydraulic conductivity, mm/h)

8

0.0403

8

0.2410

9

0.0327

9

0.2330

10

0.0245

10

0.2240

CH_N2.rte (Manning’s ‘n’ for main channel) SLP.hru (Average slope steepness) SPEXP.bsn (exponent parameter for calculating sediment entrainment) SPCON.bsn (Linear parameter for calculating maximum sediment reentrained during routing) BLAI.crop.dat (Maximum potential leaf area index) ALPHA_BF.gw (Baseflow factor, 1/days) CANMX.hru (Maximum canopy storage, mm)

rainfall gauging station. However, the overall model behaviour during the calibration is in accordance with the rainfall events. The model underestimated sediment yield during high Table 5 Initial and optimal values with prior distribution of selected parameters used in calibration Parameters

Initial value in SWAT

Prior distribution of parameters

Optimal value

r__CN2.mgt v__ESCO.hru a__SOL_AWC.sol v__GW_REVAP.gw v__GWQMN.gw v__GW_DELAY.gw v__ALPHA_BF.gw v__SURLAG.bsn v__CH_N2.rte

61–92 0.65 0.1–0.17 0.02 25 31 0.048 4 0.014

U[−0.25,0.05] U[0, 1] U[0, 1] U[0.02–0.2] U[0, 500] U[0, 500] U[0, 1] U[0.05, 24] U[0, 0.3]

{1 + (−0.15)}a 0.37 AWC + 0.03 0.1 43 8 0.5 3 0.05

v - replace, a - absolute change, r - relative change a

The existing parameter values for different land uses and soil types are multiplied by (1+ a given value)

Impact of Abandoned Opencast Mines on Hydrological Processes of ...

Fig. 2 Precipitation, observed and simulated daily streamflow and sediment yield at outlet of the Olidih Watershed during calibration

peak flow but followed the trend of the observed sediment at other times with slight over- or underprediction. This may be a result of the fact that SWAT does not simulate extreme events efficiently and the model usually over- or underpredicts the largest flow events (Tolson and Shoemaker 2004). There were also periods, when the observed sediment yield was very low, during which the model overpredicted the sediment yields. This may be due to the fact that SWAT uses the simulated streamflow to estimate the sediment yield from the watershed, as reflected in Fig. 2. In addition, sediment yield peaks also corresponded to high rainfall events. Visual inspection of Fig. 3 indicates reasonable agreement between observed and simulated data with R2, NSE and RMSE of 0.83, 0.59 and 1.52, respectively, for streamflow, and 0.71, 0.53 and 5.52, respectively, for sediment during validation. The values of NSE for streamflow (0.59) and sediment yield (0.53) indicate the satisfactorily performance of the model during validation. The validation period received higher rainfall, and thus, higher observed runoff and sediment yield than the calibration period. During the validation period, the trend of simulated sediment follows the measured sediment yield during the monsoon months. Overall, the graphical comparison (Fig. 3) indicates that the model simulates the streamflow and sediment yield trend quite well with slight variations in simulating the peaks during the monsoon months which may be attributed to the rainfall characteristics and the approach of SWAT model for sediment estimation based on the total rainwater quantity rather than the intensity in the specific period (Kumar and Mishra 2015). Like in calibration, during the validation period the sediment yield corresponded to the simulated streamflow. Hence, the model underpredicted the sediment yield where the simulated runoff from the watershed was less than the observed runoff. It can be seen from the streamflow plot that the validated model also responds to the rainfall events occurring in non-monsoon months, where streamflow data was not measured. For example on 13th Feb, 2007 there was a corresponding rainfall event of 30 mm and due to lack

Shinde V.T. et al.

Fig. 3 Precipitation, observed and simulated daily streamflow and sediment yield at the outlet of the Olidih Watershed during validation

of measured streamflow data, the graph of the observed value is rather smooth on this date. On contrary, the model responded well to this rainfall event and simulates streamflow. There are many such small rainfall events in non-monsoon months to which the model responded by predicting streamflow values. Similar trend as that of calibration was found for simulating sediment yield during validation.

4.2 Potential Hydrological Response of Opencast Mines The scenario of no-mines (unmined) was developed for the Olidih watershed by altering key factors for those sub-basins or HRUs where mines are present. The calibrated SWAT model was used to simulate the behaviour of different hydrological variables for the period 2005–2010, and their average annual variation for mined and unmined scenario is presented in Fig. 4. The considerable difference in the water balance parameters and sediment yield of both scenarios was observed. The annual average surface runoff of the unmined scenario (232 mm) was found 51.33% higher than that of the mined (153.3 mm) scenario. This may be due to absence of potholes in unmined scenarios in which water gets stored or trapped resulting in increased net surface runoff at the outlet. In addition, the annual average lateral flow was 2% higher than mined scenario which was negligible as compared to the surface runoff. Surface runoff gets accumulated in depressions of mines and alters the natural flow path of the water (López and Stoertz 2001). This results in more percolation of water in soil layers through fractures, causing less surface runoff and more lateral subsurface flow. The unmined scenario results in 13% and 4% lesser annual average groundwater flow and evapotranspiration, respectively, and 6% higher annual average water yield than mined. However, the annual average sediment yield of the unmined scenario (10.2 t/ha) was found less than that of the mined (11.9 t/ha). This may be due

Impact of Abandoned Opencast Mines on Hydrological Processes of ... 700

12.5 Unmined

600

12.0

500

11.5

400

11.0

300

10.5

200

10.0

100

9.5

Sediment yield (t/ha)

Depth (mm)

Mined

9.0

0 Surface runoff

Lateral flow

Groundwater

Water yield

ET

Sediment yield (t/ha)

Fig. 4 Mean annual hydrological response simulated over the period 2005–2010 for mined and unmined scenarios

to the surface soil disturbance in opencast mine areas which results in less surface runoff and slightly more groundwater flow due to more vertical infiltration through depression storage in these areas. The hydrogeology of mine areas behaves similarly to karst systems (Ford and Williams 1989), as depicted by the low surface runoff and high groundwater flow in mine areas though large depressions. It can also be assessed from Fig. 4 that a slight change in the topography and land use may result in a completely different behaviour of the water balance components of the hydrologic cycle of a particular basin. Furthermore, mean monthly hydrological response of these two scenarios were studied and presented in Fig. 5. Monthly hydrological impact of mines shows a similar trend to that of annual results. For the two scenarios, change in lateral flow was observed more dramatically than surface flow. Monthly lateral flow of the mined scenario was less in monsoon months which increases in later months of the year. But the absolute change of lateral flow between these two scenarios was negligible as compared to surface flow. The monthly variation in water yield follows a similar trend as that of lateral flow. In case of mined scenario, an increasing trend of groundwater flow was observed during the wet months. This may be due to more exchange of surface and groundwater in monsoon months which is controlled by the availability of soil water and the GWQMN. Opencast mines in the Olidih watershed have led to lesser annual and monthly water yields and higher sediment yields due to increase in groundwater flow and decrease in surface flow and lateral subsurface flow. In addition to it, the monthly comparison of hydrological variables between mined and unmined scenarios is more noticeable as compared to that of the annual comparison.

5 Conclusions The current study provides valuable information on hydrological fluctuations due to opencast mines. Quantitative analysis of the water balance parameters infer that the change in the hydrological processes of the watershed occurs due to opencast mines. The satisfactory performance of the SWAT model, as evaluated by performance indicators such as R2, NSE and RMSE, indicates that the model can be effectively applied to assess the potential hydrological impact of opencast mines by defining large opencast mines as potholes during simulation. Abandoned opencast mines led to lesser annual and monthly water yields and higher sediment yield due to increase in groundwater flow and decrease in surface flow and lateral subsurface

Shinde V.T. et al.

Fig. 5 Mean monthly hydrological response simulated over period 2005–2010 for the mined and unmined scenarios on: a Sediment yield; b Groundwater; c Surface runoff; d Water yield; e Lateral flow; and (f) Evapotranspiration

flow. Evaluation of potential impact of these mines was possible only due to scenario analysis, which could neither be possible with field studies nor with landscape studies. Critical parameters are required to define opencast mines as potholes in SWAT model, which are possible to be determined manually for small watersheds with less number of potholes. But a tool or algorithm can be developed for depression-dominated areas which can be coupled with SWAT for determination of topographic details, including the characteristics, distribution, and hierarchical relationships of depressions, at the HRU level.

References Abbaspour KC (2007) User manual for SWAT-CUP, SWAT calibration and uncertainty analysis programs. Swiss Federal Institute of Aquatic Science and Technology, Eawag, Duebendorf, Switzerland Abbaspour KC (2008) SWAT-CUP2: SWAT calibration and uncertainty programs—a user manual. Duebendorf: Department of Systems Analysis, Integrated Assessment and Modelling (SIAM), Eawag. Swiss Federal Institute of Aquatic Science and Technology Abbaspour KC, Yang J, Maximov I, Siber R, Bogner K, Mieleitner J, Zobrist J, Srinivasan R (2007) Modelling hydrology and water quality in thepre-alpine/alpine Thur watershed using SWAT. J Hydrol 333(2–4):413–430

Impact of Abandoned Opencast Mines on Hydrological Processes of ... Almendinger JE, Murphy MS, Ulrich JS (2014) Use of the soil and water assessment tool to scale sediment delivery from field to watershed in an agricultural landscape with topographic depressions. J Environ Qual 43(1):9–17 Amatya DM, Jha M, Edwards AE, Williams TM, Hitchcock DR (2011) SWAT-based streamflow and embayment modeling of karst-affected chapel branch watershed, South Carolina. ASABE Trans 54(4): 1311–1323 Bhakdisongkhram T, Koottatep S, Towprayoon S (2007) A water model for water and environmental management at Mae Moh mine area in Thailand. Water Resour Manag 21(9):1535–1552 Bracmort KS, Arabi M, Frankenberger JR, Engel BA, Arnold JG (2006) Modeling long-term water quality impact of structural BMPS. ASABE Trans 49(2):367–374 Burbey TJ, Younos T, Anderson ET (2000) Hydrologic analysis of discharge sustainability from an abandoned underground coal mine. J Am Water Resour Assoc 36:1161–1172 Choubey VD, Rawat RK (1991) Hydrogeologic and environmental impact of Amjhore pyrite mines, India. Environ Geol Water Sci 17:51–60 Dvoracek J, Slivka V, Sterba J (2004) Problems connected with the underground mines closed. Proceedings of the 5th International Symposium on Mining Science and Technology, Xuzhou. Taylor & Francis, doi:10.1201/9780203022528.ch124 Evenson GR, Golden HE, Lane CR, D’Amico E (2015) Geographically isolated wetlands and watershed hydrology: a modified model analysis. J Hydrol 529(1):240–256 Ford D, Williams PW (1989) Karst geomorphology and hydrology. Unwin Hyman, London Galván L, Olías M, de Villarán RF, Santos JD, Nieto JM, Sarmiento AM, Cánovas CR (2009) Application of the SWAT model to an AMD-affected river (Meca River, SW Spain).Estimation of transported pollutant load. J Hydrol 377(3):445–454 Gong Y, Shen Z, Liu R, Wang X, Chen T (2010) Effect of watershed subdivision on SWAT modeling with consideration of parameter uncertainty. J Hydrol Eng ASCE 15(12):1070–1074 Hawkins JW (2004) Predictability of surface mine spoil hydrologic properties in the Appalachian plateau. Ground Water 42:119–125 Houghton JT, Ding Y, Griggs DJ, Noguer M, van der Linden PJ, Dai X, Maskell K, Johnson CA (2001) Climate change 2001: the scientific basis. Contribution of working group I to the third assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge Jhanwar MC (1996) Application of remote sensing for environmental monitoring in Bijolia mining area of Rajasthan. J Ind Soc Rem Sens 24:255–264 Kumar S, Mishra A (2015) Critical erosion area identification based on hydrological response unit level for effective sedimentation control in a river basin. Water Resour Manag 29:1749–1765. doi:10.1007/s11269014-0909-3 Li S, Chen YF, Li ZJ, Zhang K (2016) Applying a statistical method to streamflow reduction caused by underground mining for coal in the Kuye River basin. Sci China Technol Sci 59(12):1911–1920 López DL, Stoertz MW (2001) Chemical and physical controls on waters discharged from abandoned underground coal mines. Geochem-Explor Environ A 1:51–60 Menzel L, Koch J, Onigkeit J, Schaldach R (2009) Modelling the effects of land-use and land-cover change on water availability in the Jordan river region. Adv Geosci 21:73–80 Michalski SR (2004) The Jharia mine fire control technical assistance project: an analysis. Int J Coal Geol 59(1): 83–90 Moriasi DN, Arnold JG, Van Liew MW, Bingner RL, Harmel RD, Veith TL (2007) Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. ASABE Trans 50(3): 885–900 Narsimlu B, Gosain AK, Chahar BR (2015) SWAT model calibration and uncertainty analysis for streamflow prediction in the Kunwari River Basin, India, using sequential uncertainty fitting. Environ Process 2(1):79–95 Neitsch SL, Arnold JG, Kiniry JR, Williams JR (2005) Soil and water assessment tool: theoretical documentation, version 2005. Grassland, Soil & Water Research Laboratory, USDA-ARS, Temple, TX, USA Neitsch SL, Arnold JG, Kiniry JR, Williams JR (2011) Soil and water assessment tool theoretical documentation version 2009. Texas Water Resources Institute, TR No. 406, TX, USA Olias M, Nieto JM, Sarmiento AM (2011) Water quality in the future Alcolea reservoir (Odiel River, SW Spain): a clear example of the inappropriate management of water resources in Spain. Water Resour Manag 25(1): 201–215 Saini V, Gupta RP, Arora MK (2015) Environmental issues of coal mining - A case study of Jharia coal-field, India. International Conference on Engineering Geology in New Millennium, At IIT Delhi, New Delhi. doi: 10.13140/RG.2.1.4363.8805 Saleh A, Arnold JG, Gassman PW, Hauk LM, Rosenthal WD, Williams JR, MacFarland AMS (2000) Application of SWAT for the upper north Bosque River watershed. ASABE Trans 43(5):1077–1087

Shinde V.T. et al. Siddle RC, Kamil I, Sharma A, Yamashita S (2000) Stream response to subsidence from underground coal mining in central Utah. Environ Geol 39:279–291 Tolson BA, Shoemaker CA (2004) Watershed modeling of the Cannonsville basin using SWAT2000: model development, calibration and validation for the prediction of flow, sediment and phosphorus transport to the Cannonsville reservoir. Technical Report, School of Civil and Environmental Engineering, Cornell University, New York, USA Tripathi MP, Panda RK, Raghuwanshi NS (2003) Identification and prioritisation of critical sub-watersheds for soil conservation management using the SWAT model. Biosyst Eng 85(3):365–379 Van Roosmalen L, Sonnenborg TO, Jensen KH (2009) Impact of climate and land use change on the hydrology of a large-scale agricultural catchment. Water Resour Res 45:W00A15. doi:10.1029/2007WR006760 Wan R, Liu D, Munroe DK, Cai S (2013) Modelling potential hydrological impact of abandoned underground mines in the Monday Creek watershed, Ohio. Hydrol Process 27(25):3607–3616 Xiao W, Hu Z, Chugh YP, Zhao Y (2014) Dynamic subsidence simulation and topsoil removal strategy in high groundwater table and underground coal mining area: a case study in Shandong Province. Int J Min Reclam Environ 28(4):250–263 Yang J, Yanga J, Reicherta P, Abbaspour KC, Xiab J, Yanga H (2008) Comparing uncertainty analysis techniques for a SWAT application to Chaohe basin in China. J Hydrol 358(1–2):1–23 Yellishetty M, Mudd GM, Shukla R (2013) Prediction of soil erosion from waste dumps of opencast mines and evaluation of their impacts on the environment. Int J Min Reclam Environ 27(2):88–102