Advertisement

Modeling and assessing land-use and hydrological regimes to future land-use scenario for sustainable watershed management in a semi-arid region of southern India

  • Monika Saini
  • Venkatesh Dutta
  • Nagendra Prasad Singh
  • Omesh Bajpai
Original Article

Abstract

The present study investigates impact of land use land cover (LULC) change and water harvesting interventions in Kanva watershed, a rural catchment in Kaveri basin, located in semi arid region of southern India. Remote sensing data and Soil and Water assessment Tool hydrological model was used to assess changes in total water yield, groundwater recharge, percolation and evapotranspiration. Post classification change detection technique was used for LULC change analysis in 1992, 2001, 2008 and 2014. Modelling was also performed to assess potential impact of LULC using predicted land use scenario of 2022. A comparison between 1995–2003 (pre-watershed management) and 2004–2016 (post-watershed management) simulations indicated an increase of 1.01%, 4.65% and 1.74% in average runoff coefficient, groundwater recharge coefficient and percolation coefficient respectively, and decrease in evaporation coefficient by 0.51%. The results showed that LULC changes and water harvesting accounted for a difference in hydrological components between these two periods. Scenario analyses were performed for different adoption rates of rain water harvesting (RWH) interventions in sub-watersheds of Kanva watershed and its impact on hydrological parameters, especially on total water yield was identified. The analysis indicated that average runoff coefficient was 9.16% during pre-watershed period and 9.25% during post-watershed period. Further up-scaling of RWH would reduce the flow and runoff coefficient may decrease to 6.07% if RWH is adopted in whole of Kanva watershed. The results suggested that if extensive RWH is carried out in the Kanva watershed, it would result in unsustainable water management due to reduced flow downstream. The study found that the sustainability of watersheds in arid and semi-arid regions is very important given the uncertainties in future hydrologic regimes due to changes in land cover and changes in extreme rainfall patterns thus requiring suitable management interventions.

Keywords

Land-use planning Watershed hydrology Catchment responses Soil and Water assessment Tool (SWAT) Drought Sustainability Rain water harvesting 

Introduction

Semi-arid tropic (SAT) regions in the world are characterized by high rainfall variability, frequent occurrence of dry spells and severe droughts. Water scarcity is a significant problem in these regions and has direct implications on economic development, sustainable human livelihoods and environmental quality (Wang et al. 2007). High population pressure in the developing countries puts immense stress on the already scarce water resources (Falkenmark 1990). Excessive human induced abstraction of water uncompensated with poor rainfall leads to water crisis resulting in fall of water tables and drying up of rivers, wetlands and reservoirs (Ragab and Hamdy 2004; Troy et al. 2007). The SAT regions are also known for dynamic land use land cover (LULC). For example, between years 1700–1980, the total forests and grasslands area in Asia decreased by 313 m ha. It is the largest decrease in the two LULC classes in any regions of the world. Cropland area has increased with a maximum during the last 40 years (Chuluun and Ojima 2002). Mounting population pressure has led to agricultural extensification and intensification, crop transformation, agricultural encroachment into fragile ecosystems largely unsuitable for farming (e.g. steep slopes and marginal lands) and rapid urbanization (Bilsborrow 1992; Falkenmark 1997; De Koninck and Déry 1997; Lambin et al. 2001). In the recent times, human induced land use changes brought up by rain water harvesting (RWH), driven by the need to improve agricultural production and livelihoods, has gained a lot of popularity and acceptance in SAT regions like in India and Sub Saharan Africa (Li et al. 2000). Such LULC change adds complexity to hydrology in the SAT regions. The impact of LULC in assessing and understanding dynamicity of water resources is prerequisite for water resource management (Ajami et al. 2008; UN-Water 2008; Kumar et al. 2018; Sahoo et al. 2018). The need is critical in SAT regions as livelihood of majority of people is directly linked to management of land and water resources through rainfed agriculture (Rosegrant et al. 2002; Wani et al. 2009). The region is also characterized by low agricultural productivity and food insecurity (Fischer and Turner 1978; Hutchinson et al. 2010; Wani et al. 2011).

LULC changes have both short term and long term impacts on basin hydrology (Li et al. 2007). It alters water balance by changing the equilibrium between rainfall and evapotranspiration and the resultant runoff. In the short-term, destructive land use change disrupts the hydrological cycle either through increasing water yield (Li et al. 2007; Mugabe et al. 2011; Hyandye et al. 2018) or through diminishing or even eliminating the low flow in certain conditions (Bruijnzeel 1990; Pereira 1992; Croke et al. 2004). Long-term changes such as urbanization and deforestation, leads to reductions in evapotranspiration and water recycling which may initiate a feedback mechanism that results in reduced rainfall (Savenije 1995). Some of the studies to assess effect of LULC change on watershed hydrological are listed in Table 1. Such studies draw our attention towards the risk of worsening water availability and the impact on the hydrology (Bronstert et al. 2002; Ragab et al. 2012; Zhang et al. 2018) in the milieu of LULC change. However, studies on land use change, incorporating the dimension of RWH and their implication on catchment hydrology is rather limited. Research studies are often restricted to assessing land use change and land cover conversion and their further influence on water resources without considering the human induced abstraction of water through RWH. Incorporating the element of RWH in LULC change is important for hydrological assessment considering its great potential to impact hydrological processes of a watershed and its growing acceptance in SAT region. In such water-stresses region, apprehension has been raised about potential downstream implications from adoption of upstream water harvesting interventions (Ngigi et al. 2007; Garg et al. 2012). Integrating study on identifying the combined impacts of LULC change, together with RWH, on hydrological regime of a watershed would not only prove useful in framing appropriate land use policy but will also help in identifying limit of up-scaling water harvesting interventions in the watershed.
Table 1

Studies conducted globally to assess effect of LULC change on watershed hydrological along with the modelling framework used

S. no.

Study description

Region

Model used

Key reference

1

Impact of land-use changes on hydrological processes

Elbow River watershed, Canada

Land-use cellular automata (CA) model and hydrological model, MIKE-SHE

Wijesekara et al. (2012)

2

Impacts of land use changes on hydrological regime

Mohlapetsi River basin, South Africa

Hydrological model, GR4J

Troy et al. (2007)

3

Impact of land-cover changes on runoff in different LULC scenarios

River Nzoia catchment, Kenya

Hydrological model, SWAT

Githui et al. (2009)

4

Impacts of land cover changes on hydrological flow regimes

Upper Shire River, Malawi, South Africa

Hydrological model, SWAT

Palamuleni et al. (2011)

5

Impacts of soil and water management interventions on hydrological processes

Kothapally watershed, Southern India

Hydrological model, SWAT

Garg et al. (2012)

6

Quantification of contributions of changes for individual LULC classes on changes in hydrological components

Upper San Pedro watershed, Mexico

Hydrological model, SWAT

Nie et al. (2011)

7

Impact of land cover change on storm water

Sydney basin, Australia

Numerical model, RAMS

Gero et al. (2006)

8

Impact of potential land use change from different landuse change scenarios on storm-runoff generation

Lai Nullah Basin, Pakistan

Hydrological model, HEC-HMS

Ali et al. (2011)

9

Land use changes and hydrological impacts related to up-scaling of rainwater harvesting and management

Upper Ewaso Ng’iro River basin, Kenya

Conceptual model, HASR

Ngigi et al. (2007)

10

Impacts of catchment land-use change on surface–groundwater interactions

Tarcutta Creek catchment, Murraye Darling basin, Australia

Hydrological model, GWlag

Gilfedder et al. (2012)

11

Hydrological impact of water and soil conservation works

Merguellil catchment, central Tunisia

Hydrological model, GR4J

Lacombe et al. (2008)

12

Impacts of RWH on water resources

Modder river basin, South Africa

Hydrological model, SWAT

Welderufael et al. (2013)

13

Hydrological impact of land-use change

Lake Chad basin and Niger River basin, Africa

Terrestrial Ecosystem

Model, IBIS and Aquatic Transport Model, THMB

Li et al. (2007)

14

Effect of rainfall variability, land use change and increased reservoir abstraction on surface water resources

Mutangi and Romwe micro-catchments, Southern Zimbabwe

Hydrological model, ACRU

Mugabe et al. (2011)

15

Hydrological impacts of different land use

Caijiachuan watershed, Loess Plateau, China

Conceptual model based on Runoff coefficient and surface flow velocity

Wang et al. (2012)

The current study focuses on Kanva watershed, a representative watershed of SAT region in the southern Karnataka, India. The watershed possesses a natural reservoir at its outlet, whose water is used to accomplish the drinking water and irrigation demand (to some extent) of the neighboring watershed communities. The land use pattern is constantly shifting under the influence of various factors such as population pressure, government policies and climate change. Rapidly increasing population is putting a lot of pressure on limited land available, to meet the demand of food, fuel and shelter. Shrub forest, scrubland, and wasteland are constantly being replaced by agricultural fields, plantations and built up land. This trend of land use change would certainly persist in the future too, as evidenced by present agricultural practices and population growth. Alteration in land use and accordingly to soil environment results in changes in hydrological regimes of the watershed. Key environmental change observed in the Kanva watershed due to modified LULC is high stream peak flows and increased sediment deposit in downstream reservoir. Government supported policy of promoting agro-forestry and up-scaling RWH are also likely to have significant impact on watershed hydrology. To ensure sustainable water resources management in the watershed, identifying the relation between LULC change and hydrological processes is prerequisite. In the present study it is accomplished with the following objectives: (1) to determine the impact of LULC change coupled with RWH on the hydrological behaviour of the catchment; (2) to study the catchment hydrologic responses to different LULC change scenarios including RWH.

Site description

The study site is the Kanva watershed, Ramanagram district, Karnataka state in southern India. It lies between 12°51′N and 77°12′E (Fig. 1). The watershed covers an area of about 352 km2 in the upstream part of Shimsha river basin. Shimsha is a tributary of Cauvery River, which is an inter-state river in southern India. The watershed is present in the maidan region of the South Deccan Plateau. The topography of the watershed is rolling to undulating surface with gentle slopes interspersed with many hills and hill ranges. The elevation ranges from 750 to 1000 m above sea level. The lithology is dominated by younger Closepet Granites which intrude the Archean aged Peninsular Gneiss (Radhakrishna and Vaidyanadhan 1994).
Fig. 1

Study area: location of the study area; stream network; location of precipitation stations and flow gauge station; sub-watershed (1–14)

The watershed is predominantly covered by red soil with soil texture varying from gravelly sandy to clayey and depth ranging from very shallow to deep. Very shallow red gravelly sandy soils are associated with hills and with rock outcrops. A considerable region in watershed has rock outcrops, habitation mask and dyke regions in it. Degraded shrub forest and open scrubland surrounding them mostly covers the hilly and upland regions of the watershed [source: Karnataka Remote Sensing Application Centre (KSRSAC)]. The region receives an annual rainfall of 847 mm and has bimodal type of rainfall pattern with two peaks in a year. May–June and September–October are the peak rainy months in a year. The important crops grown in Kanva watershed are finger millet, maize, horsegram, sorghum, mulberry and groundnut. Plantations in the region include mango, coconut, areca nut, eucalyptus etc. Agro-forestry and agro-horticulture is the dominant feature of LULC. Irrigation from dug/borewells has enabled cultivation of irrigated vegetables, mulberry, banana and rice.

Materials and methods

Material used

Cloud free remote sensing data of Landsat TM (12 January 1992), LANDSAT 7 ETM+ (24 November 2001), IRS-P6 LISS III (5 December 2008) and LANDSAT 8 OLI (20 November 2014) were used for LULC map generation. The data were acquired from US Geological Survey (USGS) Global Visualization Viewer server (GloVis) (USGS GloVis 2011 and 2016) and Indian Space Research Organization (ISRO), Bangalore. The digital elevation model (DEM) of 30 m resolution was obtained from Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) (NASA LP DAAC 2015). Soil map of 1:50,000 scale was obtained from Karnataka State Remote Sensing Application Centre (KSRSAC).

Daily rainfall data for the two gauge stations, located inside the study region, were acquired for the period 1992–2016 from Karnataka State Natural Disaster Management Centre (KSNDMC). Data of other weather parameters such as daily maximum and minimum temperature, wind speed, sunshine hours and relative humidity were obtained from nearest meteorological observation weather station located at Gandhi Krishi Vigyan Kendra (GKVK) campus, Bangalore. Observed water inflow data (from 1995 to 2016) of Kanva reservoir, present at the outlet of Kanva watershed, was obtained from Cauvery Neeravari Nigam Limited (CNNL), Bangalore. Data related to RWH structures (i.e. surface area and total storage capacity), built during watershed development programme (2003–2008) were obtained from District Watershed Development Department, Ramanagram. The physical specification details of the RWH structures are given in Table 2.
Table 2

Physical specifications of RWH structures in Kanva river basin

Structure type

Number

Storage capacity of RWH structures (Volume in TCM)

Storage surface area (ha)

Maximum storage depth (m)

Check dam

12

64.9

3.24

2

Nala Bund/percolation tank

10

106.3

3.54

3

Farm ponds

18

42.8

1.7

2.4

LULC mapping

Landsat series dataset were geometrically and radiometrically corrected. For geometric and radiometric correction of IRS-P6 LISS III data same procedure were used as suggested by global land cover facility while processing Landsat datasets (GLCN 2009). For image classification, reference data were acquired from two sources—primary and secondary. The primary source was 430 training points collected in September and November 2009. About 60 reference points per class (excluding forest and water) were identified to be used as training sets for classification. Training points selected from historical images available on Google Earth were used to classify 1992 and 2001 images. Based on the training sets collected, field knowledge, information gathered from Google Earth and visual interpretation, a supervised classification was performed using maximum likelihood classification algorithm on the images of years 1992, 2001, 2008 and 2014, using ERDAS Imagine 11.0. Maximum likelihood classification method has been one of the widely used in the field of remote sensing because of its robustness in image classification (Strahler 1980; Zheng et al. 2005).

LULC prediction for 2022

LULC scenario for year 2022 was predicted using Land Change Modeler (LCM) model of IDRISI TAIGA software (Eastman 2006). Two LULC datasets form the basis for the LULC change scenarios in LCM. The model was tested for predicting the LULC image of year 2014 for which an actual image was used. LULC map of 2001 (time-1) and 2008 (time-2) were used as inputs to change analysis tab of IDRISI that helped to analyze gains and losses and transition of areas among the LULC classes; and to quantify the changes that took place between 2001 and 2008. Based on the analysis of transitional matrix of LULC change, major driving forces for change were identified. Five significant driving factors: (1) distance from water body (reservoir); (2) distance from stream; (3) distance from built up; (4) elevation, and (5) slope derived from DEM were identified. These were used as probable variables, accountable for developments in LULC. The MLP neural network algorithm was used to create transition potential maps. Using transition potential maps, year 2014 image was predicted using Markov Chain modeler. The accuracy of the simulated map of 2014 was examined against the LULC map of 2014. Finally, a scenario for the year 2022 was formulated.

Scenarios for RWH

Scenarios for RWH system in sub-basins of Kanva watershed were developed considering the up-scaling of watershed development. The anticipated scenarios were based on the hypothesis that (1) adoption of RWH will increase progressively due to its promising impacts on soil and water conservation in agriculture and its potential benefits to farmers, and (2) increased retention of runoff in upstream of watersheds would reduce stream flow at outlet. The criteria for deciding storage capacity of RWH structures for scenarios were obtained from the ratio of storage volume of RWH structures to area of sub-watersheds in which they were already constructed during a watershed development programme. The achieved ratio was used as criteria for up-scaling RWH in different sub-basins in Kanva watershed. Seven scenarios of RWH system were generated for analysis. Figure 2 represents locations of sub-basins A, B and C in the Kanva watershed. The scenarios along with their specifications are depicted in Table 3. As an example to explain a scenario, the scenario of RWH in serial no. 1 can be read as: RWH is implemented in sub-basin A (area 104 km2). The water holding capacity of structures of RWH is 585 TCM and has water storage surface area of 23 ha. The maximum depth of storage is 2.5 m. To assess the impact of future LULC change on hydrological regimes, various LULC scenarios were generated and simulations were carried out for the period 2011–2026. Predicted LULC image of 2022 and weather datasets of years 2001–2016 were used as input in to the model. Simulated annual hydrological parameters using LULC of 2001, 2008 and 2014 and ‘without RWH’ condition was taken as baseline situation, with which simulated hydrological parameters of all scenarios were compared to assess the impact of potential LULC change on hydrological processes.
Fig. 2

Locations of sub-watersheds A, B and C in Kanva watershed

Table 3

Specifications of anticipated RWH structures for scenario analysis

S. no.

RWH system in sub-basin/s

Sub-basin/s area (km2)

Storage capacity of RWH structures (volume in TCM)

Storage surface area (ha)

Maximum storage depth (m)

1

RWH in sub-basin A

104

585

23

2.5

2

RWH in sub-basin B

171

963

38

2.5

3

RWH in sub-basin C

77

433

17

2.5

4

RWH in sub-basin A + B

275

1548

61

2.5

5

RWH in sub-basin B + C

248

1396

55

2.5

6

RWH in sub-basin A + C

181

1018

40

2.5

7

RWH in sub-basin A + B + C

352

1981

78

2.5

Hydrological modeling

Model description

The Soil and Water Assessment Tool (SWAT) version 2012 with ArcGIS 10.1 supported interface was used in the present study. The SWAT model is a continuous, long-term, physically based distributed model developed to assess impacts of climate and land management on hydrological components. It also assesses sediment loading, and pollution transport in watersheds (Arnold et al. 1998). The model uses water balance as a driver for simulating all processes in the watershed (Neitsch et al. 2005). For modeling, watershed is divided into number of sub-watersheds using DEM, stream network and the given outlet. Each sub-watershed is further divided into multiple hydrological response units (HRUs) that are comprised of unique LULC, soil, slope and management combinations. This facilitates the model to reflect differences in various hydrologic parameters (such as total water yield, groundwater recharge, percolation and evapotranspiration and others), sediment yield and nutrients for various LULC, soils and slope combinations. The detailed description of SWAT model is given by Neitsch et al. (2005).

Model inputs and modeling

The model inputs to SWAT are DEM, LULC map, soil data and weather data (including daily precipitation, maximum/minimum air temperature, solar radiation, wind speed and relative humidity at daily time steps). The reservoir’s location and specifications (Table 3) were added to the model. RWH structures were developed under watershed management programme (WMP) in sub-watersheds 6, 7 and 11 (Fig. 1). The total area of these sub-watersheds is 38 km2. The reservoir was added at the outlet of the treated sub-watershed to represent RWH storage. The LULC map (in grid format) for four time periods (1992, 2001, 2008 and 2014), soil map (in grid format with attributes such as soil depth, available water content, hydraulic conductivity, bulk density, sand, silt and clay ratio, carbon contents and soil albedo), slope map (in percentage) were used in the model. These were used to divide sub-watershed into 350 HRUs to account for diversity within each of the sub-basin. A threshold value of 5% for LULC, soil and slope were considered to ignore minor combinations. Other important model parameters which were selected for simulations were curve number (CN) method for generating surface runoff from precipitation; Penman–Monteith method for computing potential evapotranspiration (PET) and Muskingum method to simulate channel routing. Simulation was performed from 1992 to 2016, in which the first 3 years (1992–1994) were used as warm up period and not included in the analysis. Volume estimates of in situ RWH methods, adopted at field level, mainly aimed at increasing soil moisture and reducing runoff were given as input to model. While re-calibrating the model, alteration in model parameters associated with runoff, percolation, groundwater recharge and evapotranspiration (ET) were done for treated sub-watershed, as in situ interventions also have an impact on the total water balance in a watershed.

Model sensitivity analysis, calibration and validation

A sensitivity analysis was performed to identify important parameters for model calibration. It was followed by calibration and validation of the model. Simulations set up using year 1995 LULC map were used to calibrate streamflow from 1995 to 2000 at recording gauge of Kanva reservoir. After model calibration, simulations set up using 2001 LULC map were used to validate streamflow from 2001 to 2005 at the same gauging station. Re-calibration of model with water management interventions was performed for the period of 2006–2010 and re-validation was done for the period 2011–2016. Re-calibration is important as in situ and ex situ RWH interventions alter the watershed characteristics thus hydrological regime. The performance of model efficiency in simulating stream flow was evaluated using mean square error (R2), the Nash Sutcliffe Efficiency coefficient (NSE), RMSE-observations standard deviation ratio (RSR) and Percent Bias (PBIAS) (Moriasi et al. 2007). The details of the indices are given in Table 4.
Table 4

Details of the performance evaluation indices

Name

Equation

References

Coefficient of determination (R2)

\(= \left\{ {\frac{{\mathop \sum \nolimits_{i = 1}^{n} \left( {Q_{{{\text{obs}}i}} - Q_{\text{obsavg}} } \right)(Q_{{{\text{sim}}i}} - Q_{\text{simavg)}} }}{{\left[ {\mathop \sum \nolimits_{i = 1}^{n} \left( {Q_{{{\text{obs}}i}} - Q_{\text{obsavg}} } \right)^{2} \mathop \sum \nolimits_{i = 1}^{n} \left( {Q_{{{\text{sim}}i}} - Q_{\text{simavg}} } \right)e} \right]^{0.5} }}} \right\}^{2}\)

Santhi et al. (2001), Van Liew et al. (2003)

Nash Sutcliffe Efficiency coefficient (NSE)

\(= 1 - \left[ {\frac{{\mathop \sum \nolimits_{i}^{n} \left( {Q_{{{\text{obs}}i}} - Q_{{{\text{sim}}i}} } \right)^{2} }}{{\mathop \sum \nolimits_{i}^{n} \left( {Q_{{{\text{obs}}i}} - Q_{\text{obsavg}} } \right)^{2} }}} \right]\)

Nash and Sutcliffe (1970), Moriasi et al. (2007)

RMSE-observations standard deviation ratio (RSR)

\(= \frac{\text{RMSE}}{{{\text{STDEV}}_{\text{obs}} }} = \frac{{\left[ {\sqrt {\mathop \sum \nolimits_{i}^{n} \left( {Q_{{{\text{obs}}i}} - Q_{{{\text{sim}}i}} } \right)^{2} } } \right]}}{{\left[ {\sqrt {\mathop \sum \nolimits_{i}^{n} \left( {Q_{{{\text{obs}}i}} - Q_{\text{obsavg}} } \right)^{2} } } \right]}}\)

Chu and Shirmohammadi (2004), Singh et al. (2005), Vazquez-Amábile and Engel (2005)

Percent bias (PBIAS)

\(= \left[ {\frac{{\mathop \sum \nolimits_{i = 1}^{n} \left( {Q_{{{\text{obs}}i}} - Q_{{{\text{sim}}i}} } \right) \times 100}}{{\mathop \sum \nolimits_{i = 1}^{n} \left( {Q_{{{\text{obs}}i}} } \right)}}} \right]\)

Moriasi et al. (2007)

n is the number of events, Qobsi and Qsimi are the observed and simulated runoff at event i, Qobsavg and Qsimavg are the average observed and simulated runoff time period i. Model is judged to be performed satisfactory if R2 > 0.6; NSE > 0.5; RSR < 0.7 and PBIA < ± 25

Sensitivity analysis helps in concentrating on optimum values of limited numbers of parameters that govern the model output. The built-in SWAT sensitivity analysis tool that uses the Latin Hypercube One-factor-AT-a-Time (LH-OAT) (Van Griensven et al. 2006) was used in the study. This method combines the advantages of global and local sensitivity analysis methods and can efficiently give the rank orders of parameters. The sensitivity analysis was carried out for a period of 14 years (1992–2005), including initial 3 years of warm-up period. The details of parameters which hold importance for model calibration and their relative sensitivity rank, obtained after sensitivity analysis is presented in Table 5.
Table 5

Model parameters range and optimal values

S. no.

Parameter

Description

Initial parameter range/value

Final parameter range/value

Final parameter range/value in sub-watersheds with RWH

1

SOL_AWC_Sa–Cl

Soil available water storage capacity for the soil texture type “Sandy-clay” (mm H2O/mm soil)

0–0.26

0.09–0.35

0.12–0.38

2

SOL_AWC_Sa–Cl–Lo

Soil available water storage capacity for the soil texture type “Sandy-clay-loam” (mm H2O/mm soil)

0.08–0.10

0.09–0.19

0.12–0.22

3

SOL_AWC_Cl–Lo

Soil available water storage capacity for the soil texture type “Clay-loam” (mm H2O/mm soil)

0.08–0.12

0.09–0.20

0.12–0.23

4

CN_Kharif crop land

Curve number for kharif crop land (close grown crop-contoured)

65–77

79–86

44–56

5

CN_Double/Triple crop land

Curve number for double/triple crop land (row crop-contoured and bunded)

70–84

70–86

49–63

6

CN_Plantations

Curve number for plantations/orchards (with understory)

40–71

53–55

19–50

7

CN_Scrub land

Curve number for scrubland

49–84

75–80

29–63

8

CN_Fallow land

Curve number for fallow land

71–88

80–87

50–66

9

ESCO

Soil evaporation compensation factor

0.95

0.20

0.10

10

Gwqmn

Threshold water depth in the shallow aquifer for return flow (mm)

0–1000

200

150

11

Alpha_Bf

The base flow alpha factor (days)

0.00–1.00

0.20

0.23

12

Revapmn

Threshold depth of water in shallow aquifer for “revap” (mm)

0–500

70

80

13

Gw_ Delay

Groundwater delay time (days)

0–500

20

20

14

SOL_K_Sa–Cl

Soil saturated hydraulic conductivity for the soil texture type “Sandy-clay” (mm H2O/mm soil)

2.3–12.6

4.3–14.6

5.3–15.6

15

SOL_K_Sa–Cl–Lo

Soil saturated hydraulic conductivity for the soil texture type “Sandy-clay-loam” (mm H2O/mm soil)

0–15

2–17

3–18

16

SOL_K_Cl–Lo

Soil saturated hydraulic conductivity for the soil texture type “Clay-loam” (mm H2O/mm soil)

1.8–12.2

2.8–14.2

3.8–15.2

17

Gw_Revap

Groundwater “Revap” coefficient

0.02–0.20

0.12

0.10

18

RCHRG_DP

Groundwater recharge to the deep aquifer

0.01

0.1

0.12

19

OV_N

Manning coefficient for channel

0.14

0.20

0.20

20

SURLAG

Surface runoff lag

4–10

3.5

3.5

Model calibration and validation

In the study area, all streams are ephemeral in nature and the water flow duration in these stream ranges from few hours to few days (1–2) after the rainfall events. From the observed data it was determined that the major contribution to the stream runoff is from overland and lateral flow and the role of base flow is almost negligible. So calibration was done keeping in consideration the contribution of overland flow and negligible role of base flow to total stream runoff. The SWAT user manual recommendations were used for model calibration (Neitsch et al. 2005). First the adjustments were made in the parameters to calibrate water balance. This was followed by calibration of temporal flow taking flow time lag and the hydrograph shape into consideration. The surface flow was calibrated by adjusting SOL_AWC, CN2 and SOL_K. After this, baseflow was calibrated by enhancing ET and revap, which involves removal of water from capillary fringe by evaporation or from shallow aquifer by deep rooted plants. Baseflow, revap and deep groundwater were calibrated with ALPHA_BF, GWQMN, GW_REVAP, REVAPMN and RCHARG_DP. For calibrating ET, ESCO and SOL_AWC were adjusted. SURLAG was also adjusted to match the shape of simulated hydrograph with observed data. The ranges between which the sensitive parameters were varied are given in Table 5. After each calibration, various model evaluation statistics were tested for satisfactory rating of the stream flow simulations. Fine tuning of the parameters were continued till maximum level of accuracy is achieved. Calibration of the model was followed by validation. Validation is comparison of simulated values with observed values without further adjustments of model parameters. The statistical criteria were checked for the validation output too.

Parameterization and calibration with RWH structures

The RWH structures impact the hydrology of a watershed and SWAT model has been well tested for evaluating such impacts (Arabi et al. 2007; Arabi et al. 2008; Ullrich and Volk, 2009; Sahrawat et al. 2010). To re-calibrate model taking RWH interventions into account, parameters affecting surface response (CN2, SOL_AWC and SOL K), subsurface response (GW_REVAP, REVAPMN, GWQMN and RCHRG_DP) and ET rate (ESCO and AWC) were altered in the treated sub-watersheds of Kanva watershed. Adjustment of the parameters in response to RWH practices enhanced the quality of model calibration. Final sets of model parameters for sub-watersheds with RWH are presented in Table 5.

Results and discussion

LULC mapping

Nine LULC classes are identified in the three classified images of 1992, 2001, 2008 and 2014. The classes include—kharif crop land, double/triple crop land (area sown more than once in a year), barren rock/stony waste/sheet rock, fallow land, plantation/orchards, water bodies, scrubland, shrub forest/degraded forest/forest scrub, and built up land (urban/rural) (Fig. 3). The overall classification accuracy achieved for classified images of year 1992, 2001, 2008 and 2014 are 82%, 84%, 81% and 85% respectively, and the overall kappa statistics are 0.79, 0.82, 0.76 and 0.83, respectively. The quantification of LULC changes are given in Table 6.
Fig. 3

Classified LULC map of Kanva watershed for the year 1992, 2001, 2008 and 2014

Table 6

LULC distribution in the Kanva watershed in 1992, 2001, 2008, 2014 and predicted image of 2022

LULC class

Area (km2)

Area (%)

Growth rate (%)

1992

2001

2008

2014

2022

1992

2001

2008

2014

2022

1992–2001

2001–2008

2008–2014

1992–2014

2014–2022

Kharif crop land

136.28

140.83

148.70

153.17

157.18

38.70

39.99

42.22

43.49

44.63

3.34

5.59

3.01

12.39

2.62

Double/triple crop land

25.20

28.31

31.37

33.15

35.71

7.16

8.04

8.91

9.41

10.14

12.34

10.81

5.67

31.55

7.73

Barren rock/stony waste/sheet rock

33.39

30.16

28.27

27.14

25.03

9.48

8.56

8.03

7.71

7.11

− 9.67

− 6.27

− 4.00

− 18.72

− 7.76

Fallow land

42.66

34.49

24.02

22.02

19.59

12.11

9.79

6.82

6.25

5.56

− 19.15

− 30.36

− 8.33

− 48.38

− 11.03

Plantations/orchards

23.52

38.33

45.38

52.27

58.80

6.68

10.88

12.89

14.84

16.69

62.97

18.39

15.18

122.39

12.48

Water bodies

8.30

6.43

4.92

3.19

3.12

2.36

1.83

1.40

0.91

0.89

− 22.53

− 23.48

− 35.16

− 61.57

− 2.12

Scrubland

58.91

50.67

45.97

39.26

31.21

16.73

14.39

13.05

11.15

8.86

− 13.99

− 9.28

− 14.60

− 33.36

− 20.51

Shrub forest/Degraded forest

22.28

19.10

19.03

17.06

16.29

6.33

5.42

5.40

4.84

4.63

− 14.27

− 0.37

− 10.35

− 23.43

− 4.49

Built up land (urban/rural)

1.64

3.86

4.52

4.92

5.24

0.47

1.10

1.28

1.40

1.49

135.37

17.10

8.85

200.00

6.52

Total area

352.18

Future LULC scenario analysis

Model validation was done by comparing the simulated map of 2014 with the ‘actual’ LULC map of 2014 and Kappa variations (Eastman 2006; Zadbagher et al. 2018). Obtained values had an acceptable level of accuracy, that is: Kno = 83%; Klocation = 84%; Kquantity = 84% and Kstandard = 80%. Finally, to model the pattern and tendency of changes in the future, a LULC for the year 2022 was predicted. Figure 4 shows the predicted LULC map of Kanva watershed for year 2022 and Table 6 presents anticipated area under different LULC classes. Analysis of the predicted image of 2022 and LULC image of 1992, 2001, 2008 and 2014 indicated that percentage area under different LULC classes in predicted image followed similar trends as that of previous LULC images. This evidently justifies the accuracy of predicted image. Comparison between the predicted image of 2022 and actual classified image of year 2014 indicated increase in area under kharif crop land, double/triple crop land, plantation/orchards and built up land. The rate of increase is about 2.62%, 7.73%, 12.48% and 6.52%, respectively. Fallow land, scrubland, shrub forest and water bodies showed a declining trend and the decrease in area are at the rate of 11.03%, 20.51%, 4.49% and 2.12%, respectively. Predicted image of year 2022 was used to simulate potential impacts of LULC changes on the hydrological processes.
Fig. 4

Simulated LULC map for the year 2022

Model performance and output values for stream flow

The general performance ratings and SWAT model performance results are shown in Tables 7 and 8, respectively. The comparison of simulated and observed hydrographs (Fig. 5) showed that the trends of flow changes are reasonably simulated for the calibration validation and re-calibration periods. Performance results of modelling exercise indicated that R2 and NSE values were above 0.75, RSR value less than 0.50 and PBIAS within the range of ± 20%. The ranges of these values suggested overall good model performance (Moriasi et al. 2007). From the performance rating it was concluded that SWAT model was efficient enough to simulate the runoff satisfactorily and can be used to predict the same for future scenarios.
Table 7

General performance ratings for recommended statistics for a monthly time step streamflow (Moriasi et al. 2007)

Performance rating

NSE

RSR

PBIAS

Very good

0.75 < NSE ≤ 1.00

0.00 ≤ RSR ≤ 0.50

PBIAS ≤ ± 10

Good

0.65 < NSE ≤ 0.75

0.50 < RSR ≤ 0.60

± 10 < PBIAS < ± 15

Satisfactory

0.50 < NSE ≤ 0.65

0.60 < RSR ≤ 0.70

± 15 < PBIAS < ± 25

Unsatisfactory

NSE ≤ 0.50

RSR > 0.70

PBIAS > ± 25

Table 8

SWAT model performance at calibration and validation processes

 

R 2

NSE

RSR

PBIAS

Calibration

0.85

0.83

0.39

− 13.98

Validation

0.82

0.81

0.43

− 17.81

Re-calibration

0.83

0.77

0.30

− 18.90

Re-validation

0.79

0.76

0.41

− 19.50

Fig. 5

Monthly average precipitation and simulated and observed streamflow in Kanva watershed for calibration period, 1995–2000; validation period, 2001–2005, re-calibration period, 2006–2010 and re-validation period 2011–2016

Catchment response to LULC change

A comparison of LULC maps for the years 1992, 2001, 2008 and 2014 indicated that the most significant changes occurred in four LULC classes: kharif cropland, plantations/orchards, scrubland and fallow land during 1992–2014. The proportional extent of kharif cropland increased from 38.7 to 43.49% while that of plantations increased from 6.68 to 14.84%. Conversely, during the same time period, the proportional extent of fallow land, scrubland and shrub forest decreased from 12.11 to 6.25%, 16.73 to 11.15% and 6.33 to 4.63%, respectively. The simulated average annual basinal values of total water yield, groundwater recharge, evapotranspiration and percolation simulated from each LULC map are shown in Table 9. The results obtained for the water balance components were characterized by spatial and temporal distribution of meteorological input variables and heterogeneities within the HRUs. Comparison of hydrological parameters of watershed pre-development period (1995–2003, as period 1) with watershed post-development period (2004–2016, as period 2) indicated an increase in simulated runoff coefficient [= (total water yield/rainfall) × 100] by 1.01% in period 2 from that of period 1. The average runoff coefficient for the period 1 and period 2 were 9.16% and 9.25%, respectively. For the same time periods, groundwater recharge coefficient increased by 4.65%, percolation coefficient increased by 1.74% and ET coefficient decreased by 0.51%. The average groundwater recharge coefficient values for period 1 and period 2 were 7.29% and 7.63%, respectively; average percolation coefficient were 9.97% and 10.14%, respectively; whereas average ET coefficient were 83.55% and 83.12%, correspondingly.
Table 9

Simulated average annual basinal values of hydrological components, and changes in hydrological components, for the Kanva watershed in the period from 1995 to 2016

Time period

Rainfall (PCP, mm)

Water yield (Q, mm)

Groundwater recharge (GWQ, mm)

Evapotranspiration (ET, mm)

Percolation (PERC, mm)

1995

688.61

44.20

40.91

603.50

61.97

1996

997.94

90.08

71.33

836.53

98.75

1997

938.58

81.57

79.36

777.65

93.63

1998

1088.57

129.61

95.18

863.78

124.63

1999

1069.43

138.83

102.64

827.96

119.51

2000

1309.78

190.31

128.76

990.71

133.37

2001

868.59

77.93

65.82

724.84

95.74

2002

566.64

18.53

18.67

529.44

44.55

2003

752.48

49.74

37.75

664.99

68.72

2004

925.38

82.64

70.48

772.26

110.79

2005

1161.88

138.38

112.12

911.38

133.62

2006

640.89

51.14

39.11

550.64

55.37

2007

1086.22

112.78

94.90

878.54

140.65

2008

976.58

93.62

84.43

798.53

117.42

2009

882.67

72.26

66.56

743.85

89.44

2010

817.93

65.34

60.61

691.98

75.43

2011

877.75

79.79

67.66

730.30

85.52

2012

540.10

28.86

22.63

488.61

37.48

2013

825.91

72.76

59.72

693.43

79.77

2014

757.47

66.88

46.77

643.82

71.34

2015

1108.21

142.38

114.64

851.19

127.85

2016

649.33

67.27

50.83

531.23

58.71

1995–2003

920.07

91.20

71.16

757.71

93.43

2004–2016

865.41

82.62

68.50

714.29

91.03

In further analysis, if LULC condition in 1992 was retained over the entire period of analysis (1995–2016), the runoff coefficient, groundwater recharge coefficient, percolation coefficient and ET coefficient for period 2, were found to be 8.89%, 7.42%, 9.49% and 83.69%, respectively. The results indicated decrease in runoff coefficient, groundwater recharge coefficient, percolation coefficient by 0.36 units, 0.21 units and 0.64 units, respectively and increase in ET by 0.57 units from the simulation runs for period 2, if temporally varied LULC were used. Similarly, if LULC condition in 2014 was maintained over the entire simulation period 1995–2016, then for the period 1, runoff coefficient, groundwater recharge coefficient, percolation coefficient by and ET coefficient was found to be 9.64%, 7.57%, 11.34% and 82.79%, respectively. The outcomes signify increase in runoff coefficient, groundwater recharge coefficient, percolation coefficient by 0.48 units, 0.28 units and 1.38 units, correspondingly and decrease in ET by 0.76 unit from the simulation runs for period 1, if LULC images of year 1992 and 2001 were used for simulations. The results show that LULC changes accounted for a difference in hydrological components between the two periods of analysis.

The pattern of LULC transformation indicated increase in cropped area at the cost of fallow land and scrubland and increase in plantations by replacing scrubland, shrub forest and crop land. As land cover under agricultural crop land increased, surface runoff also showed increasing trend. Crop land has the highest potential for runoff because the land is bare after crop harvesting and shortly after the sowing phase. During these phases of crop cultivation, plants do not cover the soil completely, consequently surface runoff increases (Rogger et al. 2017; Markert et al. 2018). In the current study, surface runoff was calculated with adjustments in the SCS curve numbers (CNs). Agricultural crop land and plantations were assigned CN ranging between 79–86 and 53–55 compared to CN between 80–87 and 75–80 for fallow land and scrubland, respectively. Since CN is a model parameter that indicates the runoff response of a drainage basin, so change in LULC results in altogether different runoff CN, which alters the rainfall runoff responses (Tasdighi et al. 2018). High CN indicates high surface runoff and vice versa. As there is increase in area under crop land at the cost of fallow land and scrubland, results in increase in surface runoff, as simulated by the model. Agro-horticulture, mostly comprising of plantations like mango and coconut trees grown between the crops, is rapidly increasing in the Kanva watershed. Plantations increased in area by replacing crop land, scrubland and shrub forest. Understory cover of the plantations, such as grasses along with mulching practices are adopted in the region. This resulted in opposite effect on runoff, that is, runoff decreased. Thus, overall there was not much change in the total water yield between the two assessment periods. Plantations with understory covers and mulch are also more favorable for infiltration of water into soil. With increase in area under plantations, more water infiltrated into deep soil and groundwater recharge/percolation increased (Bargués Tobella et al. 2014; Ilstedt et al. 2016; Chaturvedi et al. 2018). LULC in the concerned region has shown major shift in the direction of crop land and agroforestry. Community is extensively accepting this concept of farming due to economic gains associated with it, while government policies (through WMPs) are widely promoting it owing to its livelihood implications to community as well as ecosystem services it provides. Agroforestry is found to be one of the most desirable strategy for maintaining economic and ecological sustainability in the region.

Catchment response to future LULC change and RWH scenarios

A summary of the average annual hydrological parameters from the simulation of baseline situation and eight scenarios is provided in Table 10. Trend of change in the extent of LULC classes were similar to that of past, with agricultural land and plantations increasing in area and fallow land, scrubland and shrub forest loosing land to them. Comparison of hydrological parameters of baseline situation with scenario 1 helped in assessing the impact of different LULC conditions on hydrology keeping weather data sets unchanged. The simulation output indicated that there is increase in total water yield and ET by 4.67% and 0.89% respectively, whereas groundwater recharge and percolation decreased by 3.78% and 4.65%, respectively in scenario 1 as compared to the baseline. Scenario 2 to scenario 8 illustrated hydrological implications of prospective LULC change, including up-scaling of RWH at various levels. Comparison of simulation results of baseline situation with that of scenario 8 signifies the potential role of LULC change including RWH system in whole of watershed in altering water balance. In Scenario 8, there was reduction in total water yield by 31.81% and increase in groundwater recharge and percolation by 8.64% and 26.44% from baseline situation. ET increased by 3.57% in scenario 8 from the baseline. The independent effect of RWH system on water balance was also analysed by comparing scenario 1 with scenario 8. The results indicated 34.85% decrease in water yield, increase in groundwater recharge and percolation by 12.91% and 20.82%, respectively, and 2.66% increase in ET in scenario 8 from scenario 1.
Table 10

Simulated average annual basinal values (period 2011–2026) of hydrological components in Kanva watershed for baseline condition and eight scenarios

S. no.

Scenarios

Q (mm)

GWQ (mm)

ET (mm)

PERC (mm)

% Change from baseline

Q

GWQ

E T

PERC

 

Baseline (LULC image 2001, 2008 and 2014 and no SWC)

74.29

68.77

690.94

78.9

    

1

Scenario 1 (LULC Image 2022 and no SWC)

77.76

66.17

697.07

82.57

4.67

− 3.78

0.89

− 4.65

2

Scenario 2 (LULC Image 2022 and SWC in sub-basin A)

72.34

71.01

697.65

88.43

− 2.62

3.26

0.97

12.08

3

Scenario 3 (LULC Image 2022 and SWC in sub-basin B)

64.25

67.15

709.6

92.25

− 13.51

− 2.36

2.70

16.92

4

Scenario 4 (LULC Image 2022 and SWC in sub-basin C)

72.22

65.88

702.9

90.59

− 2.79

− 4.20

1.73

14.82

5

Scenario 5 (LULC Image 2022 and SWC in sub-basin A + B)

58.13

72.22

710.65

98.83

− 21.75

5.02

2.85

25.26

6

Scenario 6 (LULC Image 2022 and SWC in sub-basin A + C)

67.83

69.38

703.79

93.55

− 8.70

0.89

1.86

18.57

7

Scenario 7 (LULC Image 2022 and SWC in sub-basin B + C)

55.51

70.91

714.58

97.82

− 25.28

3.11

3.42

23.98

8

Scenario 8 (LULC Image 2022 and SWC in sub-basin A + B + C)

50.66

74.71

715.63

99.76

− 31.81

8.64

3.57

26.44

The assessment of the scenario based hydrological implications of potential LULC change including RWH suggests that the impact is significant on hydrological parameters, especially on total water yield. The total water yield showed continuous decreasing trends as level of RWH up-scaling increased in watershed. As RWH practices improve infiltration capacity and the water holding capacity of the soil, increase in groundwater recharge and percolation is also noticed. However, the impact of RWH on groundwater recharge and percolation is not as significant as that of water yield owing to the hard rock dominating the geology of the watershed which limits the infiltration of water into the aquifers. ET is also estimated to increase with up-scaling of RWH with more water loss by evaporation from water storage structures and loss of soil moisture.

The Kanva River feeding the Kanva watershed is basically an ephemeral river, with water flowing through it only during wet season, from July to October. Water stored in the Kanva reservoir is mainly used to accomplish the drinking water demand and irrigation demand, to some extent. For the promotion of rainfed farming, various WMPs are being promoted and implemented by government and non-government organizations. Most of the programmes are implemented at micro-scale, without considering overall impact on the downstream community. In the Kanva watershed too, few sub-basins were treated with watershed interventions. There may be up-scaling of watershed programme in near future in other sub-basins also. However, various RWH scenarios suggest that further watershed development can have serious implications on downstream community due to reduced outflow. These estimated river flows reduction present a serious challenge and hence the need to formulate sustainable solutions. Limit of upscaling RWH in a watershed should be evaluated so that it does not negatively impact downstream community (Dile et al. 2016; Vema et al. 2018). The current study is an effort towards identifying such challenges and proposes appropriate solutions. Analysis of results suggest that while the trend of LULC change, including RWH to some extent, has come out to be economically and ecologically favorable, but further upscaling of RWH can have serious implications on both—environment and community. The proposed modeling framework, that integrates both LULC and RWH interventions, emphasizes on LULC–RWH interface that could augment water balance and reduce conflicts at both micro and macro watershed scale. As a tool it provides preventive environmental security and timely integration of environmental considerations in decision making that ensures environmental sustainability.

Conclusion

A vital aspect of this study was to ascertain links between LULC change and hydrological responses in the Kanva watershed. The major LULC changes during the study period were the reduction in fallow land, scrubland and shrub forest and, increase in agriculture crop land and plantation. Hydrological assessments through simulations studies established that surface runoff and groundwater recharge patterns in the watershed have been considerably influenced by the LULC changes that have occurred between the time periods 1992–2014. Land use changes, especially development of rainfed agriculture are a practical attempt towards attaining food security. These changes ought to have positive impact on livelihood of the watershed community. Land use development that has taken place over the years in Kanva watershed has lead to a sustainable agriculture ecosystem. A combination of agricultural crop land, plantation, and RWH techniques has helped in creating an economically and ecologically viable ecosystem. Increase in runoff triggered by increased cropland is counterbalanced with increased plantation cover and RWH interventions, which reduce surface runoff, improve infiltration and augment groundwater recharges. The analysis suggested that with replacement of fallow land, scrubland and shrub forest by agricultural land (45.85–52.90% in 1992–2014); increase in plantation (6.68–14.84% in 1992–2014) at the cost of crop land, scrubland and shrub forest and adoption of RWH (of 214 TCM volume), contributed to increase in average runoff coefficient (by 1.01%) and hence total water yield from watershed pre-development period to post-development period. Adoption of agro-forestry enhanced infiltration of water and thus helped in increase in percolation (percolation coefficient increased by 1.74%) and groundwater recharge (groundwater recharge coefficient increased by 4.65%).

Adoption of RWH, a land use change, driven by the need to improve agricultural production and livelihoods by conserving water and soil resources is also justified in SAT regions. But at the same time it is imperative to scientifically determine the limit of upscaling RWH interventions. Adoption of RWH without considering the macro-scale impact on watershed hydrology augments the problem of decreasing flow downstream. Scenario analysis suggests that up-scaling of RWH in other sub-watersheds would reduce the flow to reservoir and runoff coefficient may decrease to 6.07% if RWH is adopted in whole of Kanva watershed. This can have huge impact on long-term sustainability of watershed especially the supply of water for irrigation and drinking in the command area of the reservoir in future. This assessment study presents quantitative information to stakeholders and decision makers to decide timely and better alternatives for sustainable land and water resource planning and management. This would enhance better understanding of various hydrological processes and related socio-economic aspects. It would prove useful in formulating sustainable regulations and policies that focus on actions and needs of watershed communities and along with ensuring environmental sustainability. Eventually, it would orient us toward sustaining our most vital natural resource, that is, water.

References

  1. Ajami NK, Hornberger GM, Sunding DL (2008) Sustainable water resource management under hydrological uncertainty. Water Resour Res 44:W11406CrossRefGoogle Scholar
  2. Ali M, Khan SJ, Aslam I, Khan Z (2011) Simulation of the impacts of land-use change on surface runoff of Lai Nullah Basin in Islamabad, Pakistan. Landsc Urban Plan 102:271–279CrossRefGoogle Scholar
  3. Arabi M, Govindaraju RS, Hantush MM (2007) A probabilistic approach for analysis of uncertainty in the evaluation of watershed management practices. J Hydrol 333:459–471CrossRefGoogle Scholar
  4. Arabi M, Frankenberger JR, Engel BA, Arnold JG (2008) Representation of agricultural conservation practices with SWAT. Hydrol Process 22:3042–3055CrossRefGoogle Scholar
  5. Arnold JG, Srinivasan R, Muttiah RS, Williams JR (1998) Large area hydrologic modeling and assessment part I: model development. JAWRA J Am Water Resour As 34:73–89CrossRefGoogle Scholar
  6. Bargués Tobella A, Reese H, Almaw A, Bayala J, Malmer A, Laudon H, Ilstedt U (2014) The effect of trees on preferential flow and soil infiltrability in an agroforestry parkland in semiarid Burkina Faso. Water Resour Res 50:3342–3354CrossRefGoogle Scholar
  7. Bilsborrow RE (1992) Population growth, internal migration, and environmental degradation in rural areas of developing countries Croissance de la population, migration interne et dégradation de l’environnement dans des aires rurales de pays en développement. Eur J Popul 8:125–148CrossRefGoogle Scholar
  8. Bronstert A, Niehoff D, Bürger G (2002) Effects of climate and land-use change on storm runoff generation: present knowledge and modelling capabilities. Hydrol Process 16:509–529CrossRefGoogle Scholar
  9. Bruijnzeel LA (1990) Hydrology of moist tropical forests and effects of conversion: a state of knowledge review. Hydrology of moist tropical forests and effects of conversion: a state of knowledge review. UNESCO-IHP, Humid Tropics Programme, Paris, p 224Google Scholar
  10. Chaturvedi OP, Dagar JC, Handa AK, Kaushal R, Pandey VC (2018) Agroforestry potential for higher productivity from degraded ravine watersheds. In: Dagar J, Singh A (eds) Ravine lands: greening for livelihood and environmental security. Springer, Singapore, pp 335–360CrossRefGoogle Scholar
  11. Chu TW, Shirmohammadi A (2004) Evaluation of the SWAT model’s hydrology component in the piedmont physiographic region of Maryland. T ASAE 47:1057CrossRefGoogle Scholar
  12. Chuluun T, Ojima D (2002) Land use change and carbon cycle in arid and semi-arid lands of East and Central Asia. Sci China Ser C 45:48–54Google Scholar
  13. Croke BF, Merritt WS, Jakeman AJ (2004) A dynamic model for predicting hydrologic response to land cover changes in gauged and ungauged catchments. J Hydrol 291:115–131CrossRefGoogle Scholar
  14. De Koninck R, Déry S (1997) Agricultural expansion as a tool of population redistribution in Southeast Asia. J Southe Asian Stud 28:1–26CrossRefGoogle Scholar
  15. Dile YT, Karlberg L, Daggupati P, Srinivasan R, Wiberg D, Rockström J (2016) Assessing the implications of water harvesting intensification on upstream–downstream ecosystem services: a case study in the Lake Tana basin. Sci Total Environ 542:22–35CrossRefGoogle Scholar
  16. Eastman R (2006) Idrisi taiga tutorial. Clark Labs, Clark University, WorcesterGoogle Scholar
  17. Falkenmark M (1990) Rapid population growth and water scarcity: the predicament of tomorrow’s Africa. Popul Dev Rev 16:81–94CrossRefGoogle Scholar
  18. Falkenmark M (1997) Meeting water requirements of an expanding world population. Philos Trans R Soc B 352:929–936CrossRefGoogle Scholar
  19. Fischer RA, Turner NC (1978) Plant productivity in the arid and semiarid zones. Annu Rev Plant Physio 29:277–317CrossRefGoogle Scholar
  20. Garg KK, Karlberg L, Barron J, Wani SP, Rockstrom J (2012) Assessing impacts of agricultural water interventions in the Kothapally watershed, Southern India. Hydrol Process 26:387–404CrossRefGoogle Scholar
  21. Gero AF, Pitman AJ, Narisma GT, Jacobson C, Pielke RA (2006) The impact of land cover change on storms in the Sydney Basin, Australia. Glob Planet Change 54:57–78CrossRefGoogle Scholar
  22. Gilfedder M, Rassam DW, Stenson MP, Jolly ID, Walker GR, Littleboy M (2012) Incorporating land-use changes and surface–groundwater interactions in a simple catchment water yield model. Environ Model Softw 38:62–73CrossRefGoogle Scholar
  23. Githui F, Mutua F, Bauwens W (2009) Estimating the impacts of land-cover change on runoff using the soil and water assessment tool (SWAT): case study of Nzoia catchment, Kenya/Estimation des impacts du changement d’occupation du sol sur l’écoulement à l’aide de SWAT: étude du cas du bassin de Nzoia, Kenya. Hydrol Sci J 54:899–908CrossRefGoogle Scholar
  24. GLCN (2009) Global land cover network. Global land cover 2000. http://www.glcn.org. Accessed 23 March 2017
  25. Hutchinson CF, Varady RG, Drake S (2010) Old and new: changing paradigms in arid lands water management. Water and sustainability in arid regions. Springer, Dordrecht, pp 311–332CrossRefGoogle Scholar
  26. Hyandye CB, Worqul A, Martz LW, Muzuka AN (2018) The impact of future climate and land use/cover change on water resources in the Ndembera watershed and their mitigation and adaptation strategies. Environ Syst Res 7(1):7CrossRefGoogle Scholar
  27. Ilstedt U, Tobella AB, Bazié HR, Bayala J, Verbeeten E, Nyberg G, Sanou J, Benegas L, Murdiyarso D, Laudon H, Sheil D (2016) Intermediate tree cover can maximize groundwater recharge in the seasonally dry tropics. Sci Rep 6:21930CrossRefGoogle Scholar
  28. Kumar N, Singh SK, Singh VG, Dzwairo B (2018) Investigation of impacts of land use/land cover change on water availability of Tons River Basin, Madhya Pradesh, India. Model Earth Syst Environ 4(1):295–310CrossRefGoogle Scholar
  29. Lacombe G, Cappelaere B, Leduc C (2008) Hydrological impact of water and soil conservation works in the Merguellil catchment of central Tunisia. J Hydrol 359:210–224CrossRefGoogle Scholar
  30. Lambin EF, Turner BL, Geist HJ, Agbola SB, Angelsen A, Bruce JW, Coomes OT, Dirzo R, Fischer G, Folke C, George P (2001) The causes of land-use and land-cover change: moving beyond the myths. Glob Environ Chang 11:261–269CrossRefGoogle Scholar
  31. Li F, Cook S, Geballe GT, Burch WR Jr (2000) Rainwater harvesting agriculture: an integrated system for water management on rainfed land in China’s semiarid areas. Ambio 29:477–483CrossRefGoogle Scholar
  32. Li KY, Coe MT, Ramankutty N, De Jong R (2007) Modeling the hydrological impact of land-use change in West Africa. J Hydrol 337:258–268CrossRefGoogle Scholar
  33. Markert KN, Griffin RE, Limaye AS, McNider RT (2018) spatial modeling of land cover/land use change and its effects on hydrology within the Lower Mekong Basin. In: Vadrevu K, Ohara T, Justice C (eds) Land-atmospheric research applications in South and Southeast Asia. Springer Remote Sensing/Photogrammetry. Springer, Cham, pp 667–698CrossRefGoogle Scholar
  34. 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. T ASABE 50:885–900CrossRefGoogle Scholar
  35. Mugabe FT, Chitata T, Kashaigili J, Chagonda I (2011) Modelling the effect of rainfall variability, land use change and increased reservoir abstraction on surface water resources in semi-arid southern Zimbabwe. Phys Chem Earth A/B/C 36:1025–1032CrossRefGoogle Scholar
  36. NASA LP DAAC (2015) ASTER Orthorectified Digital Elevation Model (DEM) Version 3. NASA EOSDIS Land Processes DAAC, USGS Earth Resources Observation and Science (EROS) Center, Sioux Falls. https://lpdaac.usgs.gov. Accessed 1 Oct 2015
  37. Nash JE, Sutcliffe JV (1970) River flow forecasting through conceptual models part I-A discussion of principles. J Hydrol 10:282–290CrossRefGoogle Scholar
  38. Neitsch SL, Arnold JG, Kiniry JR, Williams JR, King KW (2005) Soil and water assessment tool theoretical documentation. Grassland. Soil and Water Research Laboratory, TempleGoogle Scholar
  39. Ngigi SN, Savenije HH, Gichuki FN (2007) Land use changes and hydrological impacts related to up-scaling of rainwater harvesting and management in upper Ewaso Ng’iro river basin, Kenya. Land Use Policy 24:129–140CrossRefGoogle Scholar
  40. Nie W, Yuan Y, Kepner W, Nash MS, Jackson M, Erickson C (2011) Assessing impacts of Landuse and Landcover changes on hydrology for the upper San Pedro watershed. J Hydrol 407:105–114CrossRefGoogle Scholar
  41. Palamuleni LG, Ndomba PM, Annegarn HJ (2011) Evaluating land cover change and its impact on hydrological regime in Upper Shire river catchment, Malawi. Reg Environ Change 11:845–855CrossRefGoogle Scholar
  42. Pereira H (1992). Keynote paper. In: Parde J, Blanchard D (eds) Forests, a heritage for the future. Proceedings of the 10th World Forestry Congress, Paris, September 1991. Revue Forestiere Francaise, Paris, pp 139–150Google Scholar
  43. Radhakrishna B, Vaidyanadhan R (1994) Geology of Karnataka. Geological Society of India Publication, IndiaGoogle Scholar
  44. Ragab R, Hamdy A (2004) Water management strategies to combat drought in the semiarid regions. In: 2nd Regional conference on Arab water 2004: water management for drought mitigation in the Mediterranean, Cairo, pp 37–101Google Scholar
  45. Ragab R, Bromley J, D’Agostino DR, Lamaddalena N, Luizzi GT, Dörflinger G, Katsikides S, Montenegro S, Montenegro A (2012) Water resources management under possible future climate and land use changes: the application of the Integrated Hydrological Modelling System, IHMS. Integrated water resources management in the Mediterranean Region. Springer, Dordrecht, pp 69–90CrossRefGoogle Scholar
  46. Rogger M, Agnoletti M, Alaoui A, Bathurst JC, Bodner G, Borga M, Chaplot V, Gallart F, Glatzel G, Hall J, Holden J (2017) Land use change impacts on floods at the catchment scale: challenges and opportunities for future research. Water Resour Res 53:5209–5219CrossRefGoogle Scholar
  47. Rosegrant M, Cai X, Cline S, Nakagawa N (2002) The role of rainfed agriculture in the future of global food production. Environment and Production Technology Division Discussion Paper. International Food Policy Research Institute (IFPRI), Washington DCGoogle Scholar
  48. Santhi C, Arnold JG, Williams JR, Dugas WA, Srinivasan R, Hauck LM (2001) Validation of the swat model on a large RWER basin with point and nonpoint sources. J Am Water Resour Assoc 37:1169–1188CrossRefGoogle Scholar
  49. Sahoo S, Dhar A, Debsarkar A, Kar A (2018) Impact of water demand on hydrological regime under climate and LULC change scenarios. Environ Earth Sci 77:341CrossRefGoogle Scholar
  50. Sahrawat KL, Wani SP, Pathak P, Rego TJ (2010) Managing natural resources of watersheds in the semi-arid tropics for improved soil and water quality: a review. Agric Water Manag 97:375–381CrossRefGoogle Scholar
  51. Savenije H (1995) Does moisture feedback affect rainfall significantly? Phys Chem Earth 20:507–513CrossRefGoogle Scholar
  52. Singh J, Knapp HV, Arnold JG, Demissie M (2005) Hydrological modeling of the Iroquois river watershed using HSPF and SWAT. J Am Water Resour Assoc 41:343–360CrossRefGoogle Scholar
  53. Strahler AH (1980) The use of prior probabilities in maximum likelihood classification of remotely sensed data. Remote Sens Environ 10:135–163CrossRefGoogle Scholar
  54. Tasdighi A, Arabi M, Harmel D (2018) A probabilistic appraisal of rainfall-runoff modeling approaches within SWAT in mixed land use watersheds. J Hydrol 564:476–489CrossRefGoogle Scholar
  55. Troy B, Sarron C, Fritsch JM, Rollin D (2007) Assessment of the impacts of land use changes on the hydrological regime of a small rural catchment in South Africa. Phys Chem Earth A/B/C 32:984–994CrossRefGoogle Scholar
  56. Ullrich A, Volk M (2009) Application of the Soil and Water Assessment Tool (SWAT) to predict the impact of alternative management practices on water quality and quantity. Agric Water Manag 96:1207–1217CrossRefGoogle Scholar
  57. UN-Water (2008) Status Report on IWRM and Water Efficiency Plans for CSD16. http://www.unwater.org/app/uploads/2017/05/UNW_Status_Report_IWRM.pdf. Accessed 13 April 2016
  58. U.S. Geological Survey (USGS) Global Visualization Viewer server (GloVis) (2011) Landsat 4–5 TM and Landsat 7 ETM+ image data. https://glovis.usgs.gov. Accessed 13 Dec 2011
  59. U.S. Geological Survey (USGS) Global Visualization Viewer server (GloVis) (2016) Landsat 8 OLI/TIRS image data. https://glovis.usgs.gov. Accessed 18 Nov 2017
  60. Van Griensven A, Meixner T, Grunwald S, Bishop T, Diluzio M, Srinivasan R (2006) A global sensitivity analysis tool for the parameters of multi-variable catchment models. J Hydrol 324:10–23CrossRefGoogle Scholar
  61. Van Liew MW, Arnold JG, Garbrecht JD (2003) Hydrologic simulation on agricultural watersheds: choosing between two models. T ASAE 46:1539–1551CrossRefGoogle Scholar
  62. Vazquez-Amábile GG, Engel BA (2005) Use of SWAT to compute groundwater table depth and streamflow in the Muscatatuck River watershed. T ASAE 48:991–1003CrossRefGoogle Scholar
  63. Vema V, Sudheer KP, Chaubey I (2018) Hydrologic design of water harvesting structures through simulation-optimization framework. J Hydrol 563:460–469CrossRefGoogle Scholar
  64. Wang G, Liu J, Kubota J, Chen L (2007) Effects of land-use changes on hydrological processes in the middle basin of the Heihe River, northwest China. Hydrol Process 21:1370–1382CrossRefGoogle Scholar
  65. Wang S, Zhang Z, McVicar TR, Zhang J, Zhu J, Guo J (2012) An event-based approach to understanding the hydrological impacts of different land uses in semi-arid catchments. J Hydrol 416:50–59CrossRefGoogle Scholar
  66. Wani SP, Rockström J, Oweis TY (eds) (2009) Rainfed agriculture: unlocking the potential, vol 7. CABI, WallingfordGoogle Scholar
  67. Wani SP, Rockstrom J, Sahrawat KL (2011) Integrated watershed management in rainfed agriculture. CRC Press, FloridaGoogle Scholar
  68. Welderufael WA, Woyessa YE, Edossa DC (2013) Impact of rainwater harvesting on water resources of the Modder river basin, central region of South Africa. Agric Water Manag 116:218–227CrossRefGoogle Scholar
  69. Wijesekara GN, Gupta A, Valeo C, Hasbani JG, Qiao Y, Delaney P, Marceau D (2012) Assessing the impact of future land-use changes on hydrological processes in the Elbow River watershed in southern Alberta, Canada. J Hydrol 412:220–232CrossRefGoogle Scholar
  70. Zadbagher E, Becek K, Berberoglu S (2018) Modeling land use/land cover change using remote sensing and geographic information systems: case study of the Seyhan Basin, Turkey. Environ Monit Assess 190(8):494CrossRefGoogle Scholar
  71. Zhang L, Nan Z, Yu W, Zhao Y, Xu Y (2018) Comparison of baseline period choices for separating climate and land use/land cover change impacts on watershed hydrology using distributed hydrological models. Sci Total Environ 622:1016–1028CrossRefGoogle Scholar
  72. Zheng MG, Cai QG, Wang ZJ (2005) Effect of prior probabilities on maximum likelihood classifier. In: Proceedings of the 25th IEEE international geoscience and remote sensing symposium, 25–29 July 2005, Seoul, pp 3753–3756Google Scholar

Copyright information

© Society for Environmental Sustainability 2018

Authors and Affiliations

  1. 1.DST-Centre for Policy ResearchBabasaheb Bhimrao Ambedkar UniversityLucknowIndia
  2. 2.Department of Civil EngineeringMNITBhopalIndia
  3. 3.Department of Environmental ScienceBabasaheb Bhimrao Ambedkar UniversityLucknowIndia

Personalised recommendations