It is documented that the analysis of future climate impact involves large uncertainties. These uncertainties are due to several factors including different types of emission scenarios, hydrologic modeling setup, downscaling and bias correction methods. Therefore, to decrease these uncertainties of the model bias correction is required. Bias correction procedures employ a transformation algorithm for adjusting RCM output. The underlying idea is the identification of possible biases between observed and simulated climate variables, which is for correcting both control and scenario RCM runs. Bias correction methods are assumed to be stationary i.e. the correction algorithm and parameterizations for current climate conditions are also valid for future conditions. Teutschbein and Seibert listed a lot of bias correction methods, and provide a detailed discussion and state that a method that performs well for current conditions is likely to perform better for changed conditions than a method that already performs poorly for current conditions.
Hydrologic modeling has proved to be a powerful tool that can be applied to understand and explain the effects of land use and land cover change on hydrologic response of a catchment. Hydrologic models provide a framework to investigate the relationship between human activities, climate and water resources.
There are many different reasons why modeling of the rainfall-runoff processes of hydrology is required. The main reasons behind are a limited range of hydrological measurement techniques and a limited range of measurements in space and time. Therefore, it is necessary to develop a means of extrapolating from those available measurements in space and time to ungauged catchments and into the future to assess the likely impact of future hydrological change. Hydrological models are characterizations of the real-world system. The researchers use a wide range of hydrological models; however, the applications of those models are highly dependent on the purposes for which the modeling is made. stated that many rainfall-runoff models are carried out purely for research purposes as a means of enhancing knowledge about hydrological systems. He also Adds that other types of models are developed and employed as tools for simulation and prediction aiming ultimately to allow decision-makers to improve decision making about hydrological problems. Before developing the hydrological models, it is vital to understand how the catchment responds to rainfall under different conditions.
On the basis of the process description, the hydrological models can be classified into three main categories.
Lumped model; According to Moradkhani and Sorooshian in lumped models, the entire river basin is taken as a single unit where spatial variability is disregarded and hence the outputs are generated without considering the spatial processes. The parameters often do not represent physical features of hydrologic processes and usually involve certain degree of empiricism.
Distributed models; Parameters of distributed models are fully allowed to vary in space at resolution chosen by the user. Distributed modeling approach attempts to incorporate data concerning the spatial distribution of parameters together with computational algorithms to evaluate the influence of this distribution on simulated precipitation runoff behavior. Distributed models generally require large amount of data. Beven explains that th distributed model does have some problems with its non-linearity, scale, uniqueness and uncertainty.
Semi distributed models; to overcome the difficulties of fully distributed another semi distributed are compromise between lumped and fully distributed. According to Arnold the algorisms in semi distributed conceptual models are simple but physically based. Parameters of semi distributed models are partially allowed to vary in space by dividing the basin in to a number of smaller sub basins. The main advantage of these models is that their structure is more physically based than the structure of lumped models and need less input data than fully distributed models. SWAT, HEC-HMS and HBV are Considered as semi distributed models.
There are many criteria which can be used for choosing the right hydrologic model. These criteria always project dependent, since every project has its own specific requirements and needs. Further, some criteria are user dependent, such as the personal preference for graphical user interface, computer operating system, input out management system and structure. The four fundamental criteria that must be considered for model selections are:
- Predict the impact of land management practices on water, sediment, and agricultural yields in large complex watersheds with varying soils, land use and management conditions over long periods of time
- Requires specific information about weather, soil properties, topography, vegetation, and land management practices in watershed.
- Hydrological processes that need to be modeled to estimate the desired outputs adequately (Is the model capable of simulating single event or continuous processes?)
- Available of input data (Can all the inputs required by the model be provided within the time and cost constraints of the project?
In addition, the model must be readily and freely available within available documentation and should be applied over a range of catchment sizes from large to global.
For this study SWAT is selected because its structure is more physically based than the structure of the lumped model, freely available and meets the objective of this study in addition to the above criteria.
The Soil Water Assessment Tool, a semi-distributed, continuous-time, process based hydrology and water quality model, was developed by Dr. Jeff Arnold and his team at the Agricultural Research Service of the United States
Department of Agriculture to analyze the impacts of land use changes on discharge, erosion, sedimentation, and water quality in gauged and un gauged watersheds.
The special feature of SWAT is the use of HRUs where designated land use, soil type, and slope information can be grouped into files for each sub-basin. Outputs at the HRU level are aggregated at sub basin level, and eventually delivered from upstream to downstream sub-basin via channel routing. This approach is fairly useful back in the time when computational speed was still quite slow. On the other hand, users can assign one HRU per each sub-basin so that the SWAT project will be closer to a physically-based model with given modern computer technology.
SWAT model has been applied in agricultural watersheds and have been successfully calibrated and validated in many areas of the world. SWAT has been applied globally on various subjects including LULC change.
The studies indicated that; the SWAT model is capable of simulating hydrologic process from complex and data poor watershed with reasonable model performance statistical values, was applied SWAT model on Lake Tana Reservoir Water Balance and reported that, the overall model performance was satisfactory. Similarly, also applied SWAT model to evaluate surface runoff generation and soil erosion rates for a small watershed in the Awash River basin, Ethiopia, and recommended that, the SWAT model provides a useful tool for soil erosion assessment from watersheds and facilitates planning for a sustainable land management, was applied SWAT model for hydrological modeling of Katar watershed, Lake Ziway catchment and recommended the use of SWAT model for further future research. The above literature review indicated that the SWAT model is capable of simulating hydrological process with reasonable accuracy and can be applied to large and complex watersheds.
