Water is an important natural resource and it is essential to sustain life over the Earth’ surface. Availability of abundant water on time is a prime concern for agriculture, energy and industrial sectors. The world has viewed an increase in demand of water due to rapid growth in population, urbanization and industrialization. However, usable water is reducing and becoming scarce day by day due to changed pattern of climate, reduced storage areas and increased anthropogenic disturbances leading to the changed land use/land cover patterns. These, in combined, have altered the hydrologic conditions and led to the reduced availability of water from various surface and sub-surface sources. The Intergovernmental Panel on Climate Change (IPCC) in its Fifth Assessment Report (AR5) have expressed concern for global warming and extreme climate events as these are anticipated to become more persistent in the future. In consonance with, changed in patterns of rainfall is also predicted, which may affect the spatial and temporal availability of regional water. This may, further, aggravate crisis in water sector. Therefore, supplying water for various sectors such as agricultural, industrial, energy and domestic is one of the great challenges for 21st century. This suggests for adopting some efficient approaches for analyzing the uses, depletion, and productivity of water to maintain continuous water supply. In this respect, geospatial techniques, combined with hydrological models, can be of great importance, and effectively be employed in hydrologic and water resources management studiesHydrological Models
Hydrological or watershed models, developed to simulate the catchment behavior, are complex mathematical models. These models use complex mathematical equations to explain how climatological forcings (precipitation and moisture) are converted into the watershed response in the form of runoff through processes such as interception, evaporation, transpiration and infiltration etc. These models are being used as state-of-the-art tools due to their economic feasibility in water resource management since many years. These assist in better understanding of hydrological phenomena functioning in the basin and also show how changes in the basin parameters affect these phenomena. Hydrological models, applied to predict/simulate hydrological behaviour of the basin under changing climate conditions and land use/land cover patterns, provide synthetic time series of hydrological data for facility design.
Different hydrological models have been developed for particular applications considering underlying difficulty of quantifying a basin-scale response to small-scale spatial complexity of physical processes. Based on levels of sophistication and complexity, these are categorized into a number of groups. Xu (2002) discussed the chronological development of various hydrological models as well as their applications in detail. These models can be grouped into stochastic, deterministic, empirical, physical, conceptual, lumped, distributed and semi-distributed models based on randomness, purpose and characteristics of model structure and spatial variations.
Stochastic models deal with large random variables to represent process uncertainty, and use probability distributions of hydro-climatic variables to simulate the conversion of basin precipitation into runoff. On the other hand, deterministic models are based on real physical processes that are involved in transforming precipitation into runoff. In case of empirical models, also known as black-box models, hydrological output is directly linked with climatic inputs through some empirical equations without defining the physical processes involved in modelling. These do not aid in physical understanding due to ignorance of internal structure and resultant response of the basin.
Conversely, physical models incorporate complex systems of equations based on physical laws and theoretical concepts that govern hydrological processes such as evapotranspiration, infiltration, overflow and saturated and unsaturated zone flow. These models have a logical structure similar to the real-world, and are data intensive and operate at fine temporal scales. Conceptual models are between empirical and physical models. They take into account physical laws but in very simplified form and characterise the catchment as an idealized representation of storages.
Lumped models consider the complete basin as a homogeneous whole that can be characterised by a single set of parameters. Generally, these are represented by a set of differential or empirical algebraic equations without considering spatial variability of processes, inputs, boundary conditions. Lumped models ignore spatial traits of the basin. They incorporate average values of the catchment characteristics i.e. vegetation, soils, geology or topography. In distributed models, the basin is divided into separate units or grids and flows are passed from one node to another as water drains through the basin. Their ability in representing spatial heterogeneity of the catchment and their realistic nature make them applicable over wide range of physical environment including gauged and ungauged watersheds.
Semi-distributed models use a combination of lumped and distributed model attributes by assigning separate units as homogeneous areas to simulate runoff. The basin is divided into number of small sub-basins where model parameters are allowed to vary partially in space. The major advantage of semi distributed models is that their structure is more physically based than the structure of lumped models and that these are computationally less demanding than fully distributed models. Soil and Water Assessment Tool (SWAT) is an example of this category.
Role of Geospatial Techniques in Hydrological Modelling
The ability of distributed and semi-distributed hydrological models in integrating with spatial distribution of various inputs and boundary conditions has raised the demand for spatial data. The spatial data applied in hydrological modelling can be divided into two classes, viz., topographical data and topological data. Elevation property of the terrain is defined by the topographical data whereas topological data represents spatial distribution of terrain attributes. This mainly involves catchment area, land use and land cover classes, flow length, surface roughness and soil types. These attributes help to describe the ability of a region to store and transmit water.
Recently, remote sensing techniques have emerged as an advanced tool in deriving spatial data required in hydrological modelling, and in some extreme cases it is the only source of getting data needed for hydrological models. Remotely sensed data, including topography, land use/land cover, precipitation, snow, climatic variables, ET, and vegetation characteristics, may be used directly as inputs in hydrological models. These data are available at different temporal and spatial resolutions. This makes remote sensing techniques as an indispensable and decisive tool for successful hydrological model analysis, prediction and validation.
The proper handling of large volumes of spatial and temporal datasets as well as their integration on a common spatial platform has always been a serious issue in hydrological modelling. This has also been acknowledged in past research studies. However, to the larger extent, this problem has been resolved with advent of Geographic Information System (GIS) and its integration with hydrological models. GIS is a system that facilitates the preparation and analysis of georeferenced data. The term GIS is spelled out in many ways. Burrough (1996) defined GIS as a tool that can be used in coding, storing and retrieving geographic information. Further, Morris (2006) defined GIS as a computer based system that can be used to store, manage, manipulate, analyses and retrieve large volumes of georeferenced spatial data and associated attributes collected from a variety of sources. The most obvious feature of GIS is it capability of performing spatial analysis. The ability of GIS in processing Digital Elevation Model (DEM) data has offered modelers with new platforms for data handling and visualization. Grid-based GIS is a very suitable tool for hydrologic modeling due to its capability in processing DEM.
The coupling of Geographic Information System with hydrological models can be established using four approaches. These are embedding GIS like capabilities in hydrological modelling package, embedding hydrological modelling in GIS package, loosely coupled integration and tightly coupled integration. In this way, distributed hydrological models, together with remote sensing and GIS, can be effectively applied in water resource management practices. The integration of SWAT model with GIS is an example of loose coupling approach where model input and output parameters are facilitated directly by GIS.
Water is a precious natural resource essential for existence of living organisms. Timely water availability is growing concern for agriculture, energy and industrial sectors. Increasing population in tandem with urbanization and industrialization increases the demand for water necessitating an in-depth understanding of hydrologic processes and spatio-temporal variability of water resources. In this respect, geospatial techniques, combined with hydrological models, have shown significant potential in hydrologic and water resource management studies. Hydrologic models, developed to simulate the catchment behaviour, are grouped categorically into stochastic, deterministic, empirical, physical, conceptual, lumped, distributed and semi-distributed models based on the level of sophistication and complexity. Further, to augment the performance of distributed and semi-distributed models, these models are integrated with geospatial techniques. The chapter endeavoured to dwell on the applicability of geospatial techniques in hydrological modeling.