LAND USE LAND CHANGE IMPACT ON HYDROLOGY OF THE FOREST WATERSHED, INDIA

Land use change is the main factor influencing watershed hydrology and could serve for developing better watershed management practices. The behaviour of each process in hydrology is influenced by its attributes and other processes. Present study analyses the impact of land use change on watershed hydrology in Nallamala forest watershed,India.The Soil and Water Assessment Tool (SWAT) is used to simulate runoff using different land use land cover (LULC) maps of 2000 and 2010. In present study land use changes and hydrological responses were quantified to investigate the runoff responses on annual basis time scale using SWAT to study the impacts of land use land cover change. A calibrated SWAT model simulated annual runoff processes for a period of 10 years i.e. 2000 to 2010. Sensitivity analysis for input parameters is analysed using the SUFI-2 algorithm in SWAT_CUP (Calibration Uncertainty Programme). Four SWAT input parameters are more sensitive including CN2.mgt, Delay.gw, sol_awc.sol and sol_k.sol. Hydrological model could simulate runoff for each sub-basin using two land use scenarios 2000 and 2010,soil map and DEM.It is observed that model results using optimised parameters, hydrological processes are better predicted with statistical evaluation methods like Nash-Sutcliffe Efficiency(NSE) and Coefficient of determination(R). The results from present study help to quantify the potential impacts of land use land cover (LULC) change on total yield of water within watershed using SWAT model.

These models need data of input parameters relevant to land cover properties, soil properties, slope and meteorological parameters. Land cover change can cause change in vegetation which can alter flood frequency and extremes (Brath et al, 2006), mean annual discharge (Costa et al,2003) and change in base flow (Wang et al,2006). In watershed hydrology assessment, interactions and relationships between human activities and natural phenomena are important to understand the surface features with the help of change analysis which improves better resource management and decision making.
Change detection involves analysis of surface feature occurrence quantitatively to determine the changes associated with land cover land use with the help of multi-temporal satellite datasets.
In the past fifteen years, Nallamala forest watershed area has significantly experienced conversion of forest from agriculture land including deforestation. These land use changes have affected the hydrological cycle thus increasing rate of runoff and soil erosion. This hilly landscape associated with land cover change and rocky base layers has decreased rate of water infiltration thus raising major environmental concerns. Soil and Water Assessment Tool (SWAT) was developed by U.S. Department of Agriculture (USDA) which has proven to be a promising tool for assessment of hydrology (Arnold et al,1998).SWAT has a wide range of applications to deal with variety of watershed problems (Gassman et al,2007) like impacts of land use land cover changes (Breuer et al,2009).
Knowledge about annual and monthly mean runoff, stream discharge and groundwater is important for long-term watershed planning to protect water resources.The relationship between hydrological parameters and land use change are dynamic depending on topography, geology,soil type, climate and land use type (Hernandez et al,2000). Sensitivity analysis provides an opportunity to identify model sensitive parameters to improve model performance and reduce uncertainty. (Arnold et al,1998) used sensitivity analysis to identify sensitive hydrological parameters using SWAT model in Mississippi river basin. Several studies have been conducted using SWAT to evaluate the impact of land use change on runoff (Githui et al,2009;Li et al,2009;Wang et al,2006).The objective of our study is to investigate the hydrological responses to land use land cover change in semi arid forest watershed using Soil  to simulate long-term impacts of land use management practices (Neitsch et al,2005).
Components of SWAT simulation include weather, hydrology, soil , plant growth, nutrients, pesticides and land management practices (Gassman et al, 2007). This paper explains the long-term impacts of land use change and spatial variations in rainfall and temperature on the hydrology of the study area.

Data Preparation
SWAT requires data of topography, land use and land cover, soil variables, and meteorological data (Table 1) (Table 5).

Hydrological Response Units ( HRU) Generation
HRU is a homogeneous unit which describes different parts of the sub-basin in terms of soil, slope and land use land cover. Each sub-

Figure 2: Watershed Delineation
basin is then divided into hydrological response units (HRUs) with a total of 277 HRUs. Different soil types are described in Table 6 and area occupied within the watershed.   Figure 3. variables (e.g. rainfall), model conceptualisation, parameters and measured data (Abbaspour et al, 2004).All uncertainties associated are quantified by a measure as the P-factor, which is the percentage of measured data bracketed by the 95 per cent prediction uncertainty (95PPU).The 95PPU is estimated at 2.5 and 97.5 per cent levels of the cumulative distribution of an output variable obtained through Latin hypercube sampling, by omitting 5 per cent of the worst simulations.d-factor is the average thickness of the 95PPU band divided by the standard deviation of the measured data. Thus SUFI-2 seeks to bracket most of the measured data with the smallest possible uncertainty band (Abbaspour, 2007). The value of P-factor should tend towards one and have a d-factor close to zero. The optimisation process reflects the sensitivity of the 6 SWAT input variables. Sensitivity analysis is an integral part of model development and involves analytical examination of input parameters to aid in model validation (Yang et al, 2008) and provides guidance for future research (Jha,2011).Surface runoff and base flow variables were treated as the dependent variables and other model parameters were considered as independent variables.Initially, six SWAT parameters were selected to test surface runoff response sensitivity. These parameters are spatially distributed throughout the watershed and impart some indication of parameter ranges.
Curve number changes with changing land use land cover and soil type. Soil hydraulic conductivity varies when land cover changes (Shaw et al, 2014). Parameter sensitivity ranking is shown in Table 7.   (Moriasi et al,2007) ( Figure 5 and Table   9). The simulated and observed runoff for validation period yielded R2=0.82,NSE=0.63,p-factor=0.52 and d-factor=0.62.

Conclusion
The study investigated the impacts of land use change on runoff using hydrological modeling.Using SWAT model, hydrological parameters of the Nallamala watershed, India simulated using different land use types over a decade. Statistical evaluation between simulated and observed runoff values are found to be best fit.The sensitivity of four SWAT parameters is found to be most sensitive parameter for hydrological response (Arnold et al,1999;Spruill et al,2000). The results show that land use type is an important factor to determine the hydrological response in the study area. Major changes in the deciduous forest and scrub forest area were the major factors in the amount of runoff produced. Future conversion of forest to agriculture and barren land may lead to other ecological and hydrological problems like soil erosion and river sedimentation. Catchment area or watershed basins are often characterised with complex spatial and temporal nature with different land use types. To improve watershed management and decision making, new technological models such as SWAT are useful tools for investigation of hydrological responses. High degree of uncertainty associated with land use and hydrological models is often a major limitation. This uncertainty is due to lack of data and quality of data. A hydrologic model gives the needed quantitative information important to the decision-making process concerning land and water resources planning and management.