modern hydrology and sustainable water development (gupta/modern hydrology and sustainable water...

P1: SFK/UKS P2: SFK

c14 BLBK276-Gupta August 28, 2010 15:50 Trim: 246mm X 189mm Printer Name: Yet to Come

14 Epilogue

As the plan for writing this book was underway,it became obvious that hydrology forms part of al-most every aspect of life on our planet and it wouldbe unrealistic to address this diverse subject in a sin-gle book. It was, therefore, important to limit thescope of the book to the common issues, namely,understanding of the fundamental principles, meth-ods, and problems encountered in the field, withemphasis on the underlying science and its applica-tions to real life situations in the field of water. Thestudy of modern hydrology, therefore, involves, inaddition to conventional surface- and groundwaterhydrology, the application of new environmentaland artificial tracers, application of remote sens-ing, analytical and numerical models, and humandimensions.

The material of the book is drawn from basic con-cepts in geology, soil science, hydraulics, physics,chemistry, mathematics, engineering, and relateddisciplines as relevant to hydrology. The aim hasbeen to provide adequate understanding of the var-ious facets of modern hydrology in relation to ourenvironment, ecology, and life on the Earth fromlocal, regional as well as global perspectives, so asto present a book that can serve as a comprehen-sive textbook on modern hydrology at the grad-uate level and also serve as a reference book forresearchers and professionals in diverse disciplinesconcerned with the various facets of water on theEarth.

The aim is to enable students as well as profes-sionals to make informed analyses of any hydro-

logic dataset and plan additional investigations, ifneeded, to adequately address the hydrologic prob-lem at hand, while keeping in mind the issues ofwater ethics and the larger issue of global changeand the central role of water therein.

The first 12 chapters of the book are organizedunder three broad themes with the first 6 chaptersaddressing ‘water, its properties, and its movementand modelling’. The next 3 chapters deal with thebroad theme of ‘studying the distribution of wa-ter in space and time’. The next 3 chapters areintended to address the broad theme of ‘water re-source sustainability’. The last chapter (Chapter 13)presents a few case studies linking many of the as-pects discussed in the book with some real fieldsituations. This chapter is the wrap-up summaryto conclude in a comprehensive manner how hy-drologic investigations and analyses enable one tostudy the local, regional, and global water cycle andmanipulate and manage it, while keeping in mindconsiderations of sustainability and human welfare.

14.1 Water and its properties, qualityconsiderations, movement, andmodelling of surface- and groundwater

Many of the unique properties of water are largelydue to its V-shaped molecule with one atom of oxy-gen bound to two atoms of hydrogen. Perhaps themost important physical property of water is thatat the temperatures normally found on the Earth, itexists in all three states – liquid, solid (ice), and gas

Modern Hydrology and Sustainable Water Development S. K. Gupta

© 2011 S. K. Gupta. ISBN: 978-1-405-17124-3

P1: SFK/UKS P2: SFK


390 MODERN HYDROLOGY AND SUSTAINABLE WATER DEVELOPMENT

(vapour) – and moves around the globe continuallychanging its state. As it moves around in the ‘globalwater cycle’, transfer of energy is caused by thehigh values of its specific heat and the latent heatinvolved in its change of state.

Even though called the ‘water’ planet, the terres-trial component of the Earth contains only about 4%of all water on the Earth, with fresh water in lakesand rivers being about 0.007% and in groundwaterbeing only about 0.76% of the total water on theEarth. Yet rivers, lakes, and groundwater are thewater sources we are most familiar with and usein our daily lives. Due to a variety of geographic,geomorphic, and meteorological factors, the distri-bution of terrestrial water on the Earth is uneven.This gives every place on the Earth a unique char-acter in almost every aspect, including landscape,ecology, and environment and has governed theevolution of human societies and cultures through-out the world.

The concept of the hydrologic cycle is funda-mental to efficient management of water resources.When the flow of water is manipulated to meet hu-man needs, it is necessary to understand how itaffects the local and regional hydrologic cycle and,ultimately, the availability and quality of water todownstream users. To ensure availability of ade-quate water for human use, water managers needto be able to estimate the amounts of water thatenter, flow through, and leave a given watershed.

Of all the components of the hydrologic cycle,the most commonly measured element is precipita-tion, both as liquid rain and snow. Several methodsexist for measuring the magnitude and intensity ofprecipitation that include: (i) ground-based mea-surements at point sites; (ii) ground-based remotesensing measurements on a regional scale (weatherradar); and (iii) aircraft and satellite-based sensorsover still larger regions. Each measurement differsin respect of temporal and spatial scales and themeasurement technique employed.

Evapotranspiration (ET ) involves the vapourphase transfer of water from the land surface tothe atmosphere, through a combination of evapo-ration from open water surfaces (e.g. lakes, rivers,puddles) and transpiration by plants. Three majorfactors that limit ET are: (i) input of solar energy(particularly at high latitudes, and in winter sea-

son); (ii) water availability (in dry soils, and in theabsence of open water bodies); and (iii) turbulenttransport of vapour (under low winds, sheltered ar-eas). Perhaps the simplest way of measuring evapo-ration is with an evaporation pan. But hydrologistsand engineers are not really interested in what evap-orates from a pan; instead they wish to know theregional evaporation from a given land surface orfrom a nearby lake. Unfortunately, pan evaporationis often a poor indicator of these variables. ET canbe estimated indirectly by measuring percolationusing lysimeters and subtracting it from precipita-tion or by measuring soil water depletion during agiven time interval. Some other methods are purelytheoretical in nature. One such widely-used empir-ical method is the Thornthwaite method that pre-dicts the monthly potential ET (PE; in mm) basedon mean monthly air temperature (T ; in ◦C).

As precipitation reaches the ground surface itbegins to infiltrate into soil. Infiltration can be de-fined as the downward flux of water from the soilsurface into the soil profile via pore spaces. Theprocess becomes more complex, both with increas-ing non-uniformity of soil and rainfall rate. Infiltra-tion through a soil profile is governed by two ma-jor forces opposing each other, namely capillaryforce and gravity. It is affected by several otherfactors that include: (i) water input from rainfall,snowmelt, irrigation, and ponding; (ii) soil profileproperties, such as porosity, bulk density, conduc-tivity; (iii) antecedent soil moisture content and itsprofile; (iv) soil surface topography and roughness;and (v) soil freeze and thaw conditions. The infiltra-tion process is characterized by a general decreasein infiltration rate as a function of time and a pro-gressive downward movement and diffusion of aninfiltration front into the soil profile. The infiltra-tion rate is determined by measuring the time ittakes for a layer of water on a soil surface to dropby a fixed distance. In watershed studies for hydro-logic modelling, empirical and physical equationsenable the estimation of infiltration as a functionof time for a given set of soil properties. Empir-ical methods rely on regression equations basedon large datasets. Use of these equations enablesdetermination of infiltration with limited measure-ments. Some of the commonly used equations arethe Green-Ampt, Richards, and Horton equations.

P1: SFK/UKS P2: SFK


EPILOGUE 391

Runoff occurs when the portion of rainfall that isnot absorbed by the ground flows down-gradient.Part of the land from where runoff water collectsand joins a particular river or stream is known asthe watershed (or catchment) of the stream. Thestream hydrograph is a primary observable in ariver network draining a basin. Hydrologists uselong term measured stream discharge and attemptto relate it to rainfall characteristics in the basin. Atime-series of precipitation in a watershed is calleda rainfall hyetograph and a similar time-series ofstream discharge from the watershed is called astream flow hydrograph. There are a number ofmethods to measure the amount of water flowingin a stream or a canal. The objective is to estimatethe volume of water moving per unit time through agiven cross-section of a river channel. This is usuallycalculated using the ‘area-velocity method’. Flumesand weirs are constructed as ‘in stream’ struc-tures to make water flow through a well-definedcross-section so that flow velocity and hence dis-charge estimation can be made by measuringoverflow height above a standard fixed referencelevel.

Another commonly employed method to moni-tor surface water flow is to use a water depth-flowrate rating curve, which correlates the depth offlow to the flow rate based on measurements ofboth water depth and flow rate at the locationof interest over a range of flows. The other ap-proach, called an index velocity determination,uses a velocity meter, either magnetic or acoustic,to measure the index velocity of the flow at a gaug-ing station. This index velocity is used to calculatethe average velocity of the flow in a stream. A rat-ing curve, similar to that used for a stage-dischargerelationship, is constructed using discharge deter-mination to relate the indicated index velocity withthe stream discharge. As with other water cyclecomponents, empirical equations obtained by re-gression analysis using a large dataset are used fordischarge estimation, particularly in ungauged orinaccessible reaches of streams.

Rainfall-runoff analysis and modelling includesthe physical processes operating in a watershedleading to the transformation of rainfall into streamrunoff. The runoff hydrograph of a stream providesa reasonable indication of the processes that are op-

erative in a basin − perhaps more realistically thanany other measurement. There are a number ofways to estimate the amount of water that runs offa given surface. In addition to field-based methods,one can also use computer models and simulationsto estimate runoff volumes.

Transport of runoff through a channel systemalso gives rise to erosion and transport of sedi-ments (as suspended and bed loads) and nutrients(as dissolved load). In fact, transport of runoff wa-ter, together with dissolved, suspended, and bedload transported materials, is a major agent for land-scape evolution. Hill slopes, streams, and drainagebasins together form the fluvial network that trans-ports water, nutrients, and sediment through agiven landscape that itself changes continuouslywith time due to concomitant processes of erosionand deposition occurring in its different parts. Inits down-gradient journey, runoff produced at indi-vidual points is routed through a stream network.Several robust network laws relating to stream or-der and their number, length, and drainage area ofa basin have been identified through characteriza-tion of river channel morphology. The near univer-sality of these empirical stream networking ‘laws’seem to indicate that some basic properties of flu-vial geomorphology and the processes that shapethe landscape may be invariant over a wide rangeof spatial and temporal scales and that these empir-ical equations, in some way, exhibit the underlyingscaling laws.

In volume terms, groundwater is the most im-portant component of the active terrestrial hydro-logic cycle. But being subsurface, the main featuresof groundwater systematics are poorly known andthese can only be inferred indirectly. Below theground surface, water is contained: (i) in the topsoil as soil moisture; (ii) in the intermediate unsat-urated zone below the soil; and (iii) in the capillaryfringe as pellicular water and below it in the aquiferas groundwater. Some amount of water occurs un-der the Earth’s surface almost everywhere. Thisgives a unique advantage to groundwater in termsof its almost ubiquitous availability as compared tosurface water, which can only be harvested at cer-tain favourable sites and needs to be conveyed toother places by constructing appropriate engineer-ing structures.

P1: SFK/UKS P2: SFK



Groundwater and surface water are basicallyinterconnected. In fact, it is often difficult toseparate the two because they ‘feed’ each other.The aquifers are often partially replenished by seep-age from streams and lakes. In other locations, thesame aquifers may discharge through seeps andsprings to feed the streams, rivers, and lakes. Infact, for perennial inland rivers of non-glacial originin arid/semi-arid areas, the lean season base flowis contributed by groundwater discharge from theaquifers adjacent to the river banks.

Geological formations through which water canpass easily are said to be permeable and thosethat scarcely allow water to pass through or onlywith difficulty are described as semi-permeable orimpermeable, depending on the degree of per-meability. Saturated and permeable formations arecalled aquifers. An unconfined aquifer contains aphreatic surface (water table) as an upper bound-ary, being at the atmospheric pressure, that fluctu-ates in response to recharge and discharge of water.If, on the other hand, the effective thickness of anaquifer lies between two low permeability or im-permeable layers, it is called a confined aquifer. Animaginary surface joining the water level in bore-holes tapping a confined aquifer is called the po-tentiometric or piezometric surface.

External forces, which act on water in the sub-surface, include gravity, atmospheric pressure, andhydraulic pressure due to the overlying water col-umn, and molecular forces acting between aquifersolids and water. The driving force for groundwaterflow is known as hydraulic head – being the sumof gravitational and pressure potentials. Ground-water flows in the down-gradient direction – fromhigher to lower head. The flow of groundwaterthrough an aquifer is governed by Darcy’s Law,which states that the rate of flow is directly pro-portional to the hydraulic gradient. The coefficientof proportionality defines an intrinsic property ofporous media and is known as the hydraulic con-ductivity (K). This parameter is a measure of theease with which water flows through the variousmaterials that form aquifers and is a function ofproperties of the medium, collectively defining theintrinsic permeability (k), as well as the propertiesof the fluid. The overall permeability of a rock massor sediment depends on a combination of the size

of the pores and the degree to which the poresare interconnected. For well-sorted, granular ma-terials, hydraulic conductivity increases with grainsize.

The specific discharge per unit area or Darcyflux (q = Q/A) gives the apparent volumetric flowvelocity through a given cross-section of an aquiferthat includes both solids and voids. The averagevelocity through the pores (the pore water veloc-ity) is estimated by dividing specific discharge bythe effective or dynamic porosity (ne). The effec-tive porosity may be significantly less than the totalporosity, n, of the media; the latter defined simplyas the ratio of void space (VV) to the total volume(VT) of the material. Transmissivity, T , is a mea-sure of the amount of water that can be transmittedhorizontally through a unit width by the fully satu-rated thickness of an aquifer under a unit hydraulicgradient. Transmissivity is equal to the hydraulicconductivity multiplied by the saturated thicknessof the aquifer through which the flow takes place.

The total load above an aquifer is supported bya combination of the solid skeleton of the aquifermatrix as well as by the hydraulic pressure exertedby the water in the aquifer. When an aquifer ispumped, water is released from the storage dueto release of hydrostatic pressure within the porespaces and compression of the solid skeleton bythe overburden resulting from loss of buoyancy.The storage coefficient or storativity S (dimension-less) is the volume of water that a permeable unitwill take into storage, or release from storage, perunit surface area per unit change in head. A relatedterm, specific storage Ss [L−1], is the amount of wa-ter per unit volume of a saturated formation thatis taken into or released from the aquifer storageowing to compression of the mineral skeleton andexpansion of the pore water per unit change in thehydraulic head.

The various flow parameters controlling the flowthrough the aquifer are defined for homogenousisotropic aquifers – an ideal situation rarely en-countered in nature. However, this problem is of-ten circumvented by considering mathematicallyequivalent properties for homogenous isotropicmedia. The two physical principles, namely Darcy’slaw and the mass conservation principle, definethe groundwater flow equations. The most general

P1: SFK/UKS P2: SFK


EPILOGUE 393

form of the saturated flow equation is (Eqn. 3.36),which describes flow in three dimensions, tran-sient flow (∂h/∂t �= 0), heterogeneous conductivi-ties (e.g. Kx being a function of x), and anisotropichydraulic conductivities (Kx �= Ky �= Kz). Thisequation simplifies to the familiar Laplace equa-tion (Eqn. 3.39) for steady flow with homogenous,isotropic K.

In saturated flow through porous media, veloc-ities vary widely across pore spaces of differentsizes, shapes, and orientations. As a consequence,in addition to molecular diffusion, an irreversiblehydrodynamic mixing process occurs during lam-inar flow of groundwater between its various ele-ments. This phenomenon is known as dispersion.

Many of the measurements related to ground-water and its flow characteristics are indirect be-cause most of the subsurface is inaccessible to di-rect observation. But groundwater levels, tempera-ture, and quality parameters can be determined insitu through wells to determine groundwater flowdirection and velocities for resource managementand pollution risk assessment.

Wells are used to extract groundwater and some-times also for recharging aquifers. Well hydraulicsdeals with the process of groundwater flow to wellstapping an aquifer system. Hydraulic properties ofaquifers that control groundwater flow are: (i) hy-draulic conductivity or permeability; (ii) transmis-sivity; (iii) storativity; and (iv) hydraulic gradient.Response of a well to any discharge of water fromor recharge into the well is a function of the aquiferproperties.

Under isotropic, homogeneous aquifer condi-tions, groundwater flows radially towards a pump-ing well from all directions. When water is pumped,the level of the water table in the vicinity of the welllowers in the shape of an inverted cone, knownas the cone of depression. The drawdown curvedefines the shape of the depressed potentiomet-ric surface/water table in three dimensions. Duringsteady state conditions, the head and cone of thedepression are in equilibrium between the pump-ing rate and aquifer properties, in contrast to un-steady flow when the head of the drawdown curvechanges continuously with time.

A number of equations have been derived todescribe the flow of water to wells, using calcu-

lus and application of Darcy’s Law to groundwaterflow from the surrounding aquifer to a pumpingwell. Implicit in these derivations are assumptionsthat the pumping well is 100% efficient in extract-ing water, it fully penetrates the aquifer, the watertable or potentiometric surface has zero slope ini-tially, and laminar flow conditions (characterizedby a low Reynolds number) prevail.

Pump tests are employed to determine perfor-mance characteristics of a well and to determinethe hydraulic parameters T , K, and S of an aquifer.Another use of the pump tests is to assess the per-formance of a well by monitoring the drawdownand yield to derive the specific capacity of the welldefined as a ratio of the yield to drawdown. Thisis then used to choose the appropriate size of thepump relative to the production capacity of thewell. Two types of commonly used aquifer pumptests are: (i) the constant-rate test involving pump-ing the well for a significant length of time at a uni-form discharge rate to allow the aquifer to cometo equilibrium with the well assembly; and (ii) thestep-drawdown test wherein the well is pumpedat successively higher discharge rates for relativelyshort duration intervals. When only a single wellis available for recording of pumping rates and thecorresponding drawdown, the pump test methodcan estimate only the transmissivity and specificcapacity of the well and not the storativity or thegeometry of the cone of depression.

Multiple well pump tests involving a pump-ing well and one or more observation wells arerequired to estimate storativity, and the three-dimensional geometry of cone of depression. Initialdata analyses involve making semi-log plots of draw-down versus time since the beginning of pumpingand matching the field curve with standard modelcurves.

Hydrologic systems are complex, with pro-cesses occurring over different spatial scalescorresponding to the size and topography ofgeographical areas. In this situation, mathematicalmodels are valuable tools that enable one to makeassessments, investigate alternative scenarios,and assist in developing effective managementstrategies. Models are essentially aids to describeand evaluate the performance of relevant systemsunder various real or hypothetical constraints and

P1: SFK/UKS P2: SFK



field situations. A hydrologic model may be definedas a simplified conceptual representation of a partof the hydrologic cycle of a real-world system(here, a surface- or a groundwater, or a combinedsystem) that approximately simulates the relevantinput-output response of the system. Two majortypes of hydrologic models can be distinguished:(i) stochastic models, that use mathematical andstatistical concepts to link a certain input (e.g.rainfall) to the model output (e.g. runoff); and(ii) process-based models, that mathematically at-tempt to simulate the physical processes of surfacerunoff, subsurface flow, evapotranspiration, andchannel flow that can be quite complex. Thesemodels are known as deterministic models and canbe subdivided into single-event models, and contin-uous simulation models. Hydrological modellingcan be undertaken either as distributed or lumped-parameter simulation, differentiated by whetheror not spatial variation of hydrologic parameters isaccounted for.

Surface water hydrologic modelling is perceivedto meet two basic requirements: (i) to determinethe magnitude and frequency of flood flows; and(ii) to determine the long-term availability of waterfor consumption. The two requirements, however,involve different modelling approaches.

A groundwater model may be defined as asimplified, mathematical description of a realgroundwater system, coded in a programminglanguage. The accuracy of a model is dependentupon the level of understanding of the system andits realistic conceptualization consistent with mod-elling objectives and availability of data describingthe physical system that the model is meant to rep-resent. Any groundwater model essentially solvesthe governing equation of groundwater flow andstorage.

The three common methods of solution used ingroundwater modelling are: analytical, finite differ-ence, and finite element. Each method differs in itsapproach, assumptions, and applicability to real-world problems.

Analytical methods use classical mathematicalapproaches to solve differential equations and ob-tain exact solutions. These provide quick and reli-able results to simple problems but require assump-tions of homogeneity and are essentially limitedto one- and two-dimensional problems. However,

these can provide rough approximations for mostproblems with little effort.

Finite difference methods solve the partial-differential equations describing the system byusing algebraic equations to approximate the so-lution at discrete points in a rectangular grid. Thegrid can be one-, two-, or three-dimensional.The points in the grid, called nodes, representthe average of the surrounding rectangular block(cell). Although adjacent nodes have an effect onthe solution process, the value for a particular nodeis distinct from its neighbouring nodes. Grids usedin finite difference codes generally require far lesssetting-up time than those of finite element codes,but have less flexibility in individual node place-ment. Many common codes, such as MODFLOW,use the finite difference solution method.

Finite element methods differ from finite differ-ence methods in that the area (or volume) betweenadjacent nodes forms an element over which exactsolution values of the input parameters are definedeverywhere by means of basis functions. The es-sential difference is that finite element codes allowfor flexible placement of nodes, which can be im-portant in defining irregular boundaries. However,defining a unique location for each finite elementnode requires more effort in setting up the gridthan that of a finite difference code. FEMWATER isa common code using the finite element solutionmethod.

Some pre-processors allow superposition of thegrid and the site map, and then interactive assign-ment of boundary conditions, aquifer properties,etc. Post-processors allow the numerical output tobe presented as contour maps, raster plots, flowpath plots, or line graphs. Choosing a code thatdoes not have, or cannot easily be linked to, pre-and post-processors, should be avoided.

After selection of the modelling software, fea-tures of the conceptual model are transferred toan input file that defines the mathematical model.Features such as boundary conditions, grid dimen-sions and spacing, initial aquifer properties, andtime-steps are specified according to the require-ments of the selected code.

Groundwater models are useful in predicting theeffects arising from specific recharge and with-drawal stresses, usually employing injection andextraction wells that cause a relatively large volume

P1: SFK/UKS P2: SFK


EPILOGUE 395

of water exchange in a relatively small area. Theseanalyses can predict general aquifer response tosuch stresses.

Combining surface water and groundwatermodels to represent a real-life situation is a difficulttask. First, surface- and groundwater tend tooperate on very different temporal scales. Inaddition, in various numerical models, not only isthe space discretized (e.g. employing layers andgrid in MODFLOW the), the computations areperformed at discrete time intervals, called theoperating time-steps. The operating time-steps ina typical groundwater model and a typical surfacewater model are quite different. Groundwatermodels tend to run with time-steps of the orderof a few months to a year. For a surface watersystem, for example, flood or reservoir routing, thetime-step of a few minutes to a few hours wouldbe more appropriate. While it may take severaldays for the surface water flow to make its waythrough the section, it is required to be trackedat much shorter time-steps. Because of these andother difficulties in solving groundwater–surfacewater–unsaturated zone flow in an integrated man-ner, individual existing sectoral models are oftencoupled.

Model coupling conceptually requires a real-istic description of the hydrologic process andcombined calibration of, for example, inflows,outflows, recharges, discharges, and aquiferparameters. Coupling, therefore, provides ameans to realistically identify groundwater inflowinto and outflow from the catchment and theircoupling processes, for example, groundwaterrecharge/discharge. In practice, however, thesepotential benefits of coupled models are oftenoffset by high demand on computation time andrequired computer storage capacity. Often resultsobtained from the coupled models might be moreunrealistic than those obtained from the stand-alone sectoral models that require inputs fromhydrochemistry, natural chemical, and isotopictracers to achieve meaningful results.

One of the commonly-used codes is GSFLOW,which is a coupled groundwater–surface waterFLOW model based on integrating the US Geo-logical Survey Precipitation-Runoff Modelling Sys-tem (PRMS) and the US Geological Survey ModularGround-Water Flow Model (MODFLOW-2005).

Chemical composition of natural waters isderived from many different sources of solutes,including gases and atmospheric aerosols, weath-ering and erosion of rocks and soils, dissolutionor precipitation reactions occurring below theland surface, and more recently anthropogenicprocesses. Application of principles of chemicalthermodynamics can help in discerning broad in-terrelationships amongst these processes and theireffects on the surface water–groundwater system.Some of the processes, for example, dissolution orprecipitation of minerals, can be closely evaluatedby means of principles of chemical equilibrium,including the law of mass action. Other processesare irreversible and require knowledge of relevantreaction mechanisms and their rates.

Basic data used in the determination of waterquality are obtained by chemical analysis of watersamples in the laboratory or onsite measurementsin the field. Most of the measured constituents arereported in gravimetric units, usually milligram perlitre (mg l−1) or milli-equivalent per litre (meq l−1).Chemical analyses may be grouped and statisticallyevaluated by determining the mean, median, andfrequency distribution or ion correlation that helpto consolidate and derive useful information fromlarge volumes of data. Graphical methods of anal-yses or groups of analyses aid in identifying andshowing chemical relationships amongst differentwaters, probable sources of solutes, regional wa-ter quality relationships, temporal and spatial vari-ations, and water resource evaluation. Graphicalmethods may enable identification of water typesbased on chemical composition, relationshipsamongst ions or groups of ions in individual waters,or waters from multiple sources considered simul-taneously. Relationship of water quality to hydroge-ologic characteristics, such as stream discharge rateor groundwater flow pattern, can be representedby mathematical equations, graphs, and maps.

Adverse human impacts on water quality arisefrom contamination and resulting pollution ofthe water sources from diverse anthropogenicactivities generating waste products. Transport andattenuation of point and non-point sources of pol-lutants and basic aspects of numerical modellingof solute transport also need to be considered.

Water quality standards for domestic, agricul-tural, and industrial uses have been published

P1: SFK/UKS P2: SFK



by various agencies. Irrigation and industrialrequirements for water quality are particularly com-plex. Basic knowledge of processes that governcomposition of natural waters is required for ju-dicious management of water quality.

14.2 Distribution of water in space and time

To obtain direct insight into the dynamics of sur-face and subsurface water flow, hydrologists usetracers − substances that tag water because of theirunique properties and follow its movement. Someof the commonly used tracers are dyes, solutes (e.g.chloride), radioactive and stable isotopes, dissolvedgases (e.g. helium, CFCs), and some physical pa-rameters (e.g. temperature).

When tracers are introduced artificially intoa system for carrying out a tracer based study,these are known as artificial or injected tracers.However, if a tracer substance is already presentin a system before the start of a tracer study, itis known as an environmental tracer. Dependingon the method of analysis, artificial tracers canbe classified into four broad groups – chemical,radioactive, activable, and particulate tracers. In re-cent years, gaseous tracers have also been used forvarious applications. These include dissolved inertgases used as geochemically conservative tracers ingroundwater systems. When using environmentaltracers, hydrologists exploit variations in the com-position of a large number of substances, elements,or their isotopes within and/or across hydrologicreservoirs. A special class of environmentalradioisotopes comprises tracers of cosmogenicorigin. These isotopes are produced in nature bycosmic radiation entering the Earth’s atmosphere.Cosmic rays produce nine radio-nuclides withhalf-lives ranging between 10 years and 1.5 Ma andfive with half-lives between 2 weeks and 1 year.These have been used as tracers for measuringgroundwater movement on timescales rangingfrom a few weeks to millions of years.

Groundwater age is generally considered as theaverage travel time for a water parcel from eitherthe ground surface or from the water table of theunconfined aquifer in the recharge zone to a givenpoint along the aquifer length. Various methods

for dating of young (<50 years) and old groundwa-ters are described. It is shown that, depending onthe conceptual mathematical model employed todescribe the aquifer system, the groundwater ageestimation can yield additional information such asvelocity, residence time, dispersion coefficient, andinflux of young shallow unconfined aquifer waterinto the underlying semi-confined aquifers.

Tracers, in particular Cl− and 3H, have also beenextensively used to estimate direct recharge ofgroundwater.

Amongst various isotopes used in hydrologyas tracers, stable isotopes of oxygen (18O) andhydrogen (2H or D) are the most commonlyused. Since these form an integral part of thewater molecule, they are ideally suited to tracethe movement of water in the hydrologic cycle.In hydrologic parlance, the two isotopes are alsocommonly referred to as water isotopes.

In addition, several chemical and gaseous trac-ers are available to investigate the hydrologic pro-cesses operating on different spatial and temporalscales.

Hydrologic data essentially represent randomphenomena – temperature, rainfall, wind, etc. –measured using various instruments or derivedfrom other measurements. All measurementsinherently contain errors, both random and thosearising from the measurement process itself. Statis-tics is the science of understanding and quantifyingthe uncertainty. The questions often relate to theoutcome of a random event. The expected value,equivalent to the mean or average, and other de-scriptive parameters, such as variance, skewness,and kurtosis, summarize and describe the observeddistributions quantitatively. Data is skewed whenthere is an imbalance in the number of high andlow values of observations. The fourth-momentabout the mean, or kurtosis, is used to describe thefrequency of low probability events – both withextremely high and extremely low magnitudes.The coefficient of variation of a dataset is the ratioof the standard deviation to the mean. A hydro-logic example of the coefficient of variation isstreamflow data – low flow variations are probablymuch smaller than under high flow conditions, buttheir coefficients of variation may be of a similarmagnitude.

P1: SFK/UKS P2: SFK


EPILOGUE 397

A frequency distribution represents the distri-bution of observed values, while a probabilitydistribution is a mathematical function that predictsthe expected likelihood of an unknown variable. Adiscrete distribution is used when only a count-able number of outcomes are possible, such as inthe tossing of a coin, or the number of studentsin a class. An example of discrete distribution isthe Binomial, which is used to predict the proba-bility of the number of heads when a coin is re-peatedly tossed ‘n’ times. Continuous distributionsare used for describing outcomes that can have anyfractional value, such as the monthly or annual rain-fall values. Examples of continuous distributions in-clude the Uniform, Normal, Exponential, Gamma,and Gumbel extreme value distributions. One cancreate additional distributions by taking the loga-rithm of the random variable, resulting in distribu-tions such as the log-normal, log-gamma, etc.

One reason for applying statistical methods is tobe able to make a definitive statement about occur-rence of a situation – whether a chance occurrenceis sufficiently unlikely that we can reasonably sayhow improbable it is. There are two types of errorswhen this method is used. Even an unbiased toss-ing of a coin has a small chance of yielding a rareoutcome. So by rejecting the coin itself, one may bemaking a mistake; the rejection of a fair coin maybe called a type-1 error. On the other hand, a cointhat would appear to give a biased outcome maystill give normal results and yet not be detected.The failure to reject such a coin is called a type-2error.

In statistical tests, one generally assumes that ob-servations follow a Normal distribution – giving thefamiliar bell-shaped curve. If a particular observa-tion is too far from the mean, then one might thinkthat it is fundamentally different from the otherobservations. To check this, one should first findthe standard normal variable (Eqn. 8.47) and thenuse this variable to make a decision. Using the nor-mal distribution, one can calculate the likelihoodof this occurrence (variable). If the probability isquite small, one might infer that observations donot follow the Normal distribution.

One generally uses the least squares method tofit a distribution to a series of data points (xi,yi ± σ i) under the assumption of negligible un-

certainty in the variable x. In this method, val-ues of arbitrary constants corresponding to chosencurve/distribution are obtained by minimizing theweighted sum of squares (S) of deviations betweenobserved values of yi and their computed values foreach xi, using equations similar to (Eqn. 8.119 andEqn. 8.120). The least squares method is quite gen-eral and is applicable to any type of distribution,for instance, quadratic or second-degree polyno-mial, harmonic, exponential, etc. The Chi-Square(χ2) test is used to estimate the probability thatSmin is exceeded for the given degrees of freedomby less than 5% or 1% and the hypothesis that thechosen distribution can be accepted at 95% or 99%confidence levels, respectively.

Time series (TS) data are used to describe manyaspects of the hydrologic cycle. TS data containseveral pieces of information that can be utilizedby a user for various analytical purposes. The datais usually collected at regular intervals, referred toas the time-step. Hydrologic models can also gen-erate time series data. Methods for TS analyses canbe divided into two broad classes: (i) time-domainmethods, and (ii) frequency-domain methods. Atime domain analysis aims to describe the patternof the series over time. A frequency domain anal-ysis, on the other hand, aims to determine thestrength/power of periodicity(ies) inherent in theseries within each given frequency band over arange of frequencies.

Forecasting with classical time domain, TS meth-ods may be viewed as an attempt to decompose theseries into component parts and then predict thefuture pattern of each part. The component partsare the trend, cycle, seasonal, and irregular compo-nents. These procedures require some knowledgeof the mathematical model of the hydrologic pro-cess. However, in real-life situations, patterns ofthe data are not clear, as individual observations in-volve considerable errors. The Auto-Regressive In-tegrated Moving Average (ARIMA) methodology de-veloped by Box and Jenkins enables one to uncoverthe hidden patterns in the data and generate fore-casts. Lags of the differenced series appearing inthe forecasting equation are called ‘auto-regressive’terms, lags of the forecast errors are called ‘mov-ing average’ terms, and a time series which needs

P1: SFK/UKS P2: SFK



to be differenced to make it stationary is said to bean ‘integrated’ version of a stationary series.

Spectrum analysis is a frequency-domain methodof TS analysis and is concerned with the explo-ration of periodicities inherent in the data. Thepurpose of the analysis is to decompose a com-plex time series with periodic components into theinherent sinusoidal (sine and cosine) functions ofparticular wavelengths. Employing spectrum analy-sis, one might uncover recurring cycles of differentperiodicities in the time series of interest, whichat first would appear more or less like randomnoise.

It is sometimes of interest to investigate the jointstructure of two series, that is, the dependenceor degree of coherence between the two series.This is achieved by examining coherency and phaserelationships between the two series.

The hydrologic cycle integrates atmospheric, hy-drospheric, cryospheric, and biospheric processesover a wide range of spatial and temporal scales andthus lies at the heart of the Earth’s climate system.Studies of the integrated, global nature of the hydro-logic cycle are crucial for a proper understandingof natural climate variability and prediction of cli-matic response to anthropogenic forcing. Only inrecent years, particularly with the advent of satel-lite remote sensing, the required global data seemsto be within reach.

Although most hydrologists believe that re-motely sensed data is valuable for global hydro-logic studies and even for regional hydrologic mod-elling and field operations, these data are rarelyused in practice, possibly due to: (i) lack of neces-sary technical expertise in processing/interpretingthe data; and (ii) the form of emitted and reflectedradiances not being the type of data traditionallyused to run and calibrate models. Remote sensingdata also represent averages over finite areas, orpixels, and thus mask much of the detail at indi-vidual points to which most hydrologists are accus-tomed. In addition, current remote sensing obser-vations are not optimized to provide the temporalresolution needed to measure certain changes inhydrologic processes. Furthermore, algorithms forconverting these reflectances into physical quanti-ties are often empirical in nature and are subject tonoise present in the calibration data.

Nevertheless, remote sensing is beginning toprove its usefulness in providing hydrologic in-formation. Hydrologic remote sensing can revealcomplex spatial variations that cannot be readilyobtained through traditional in-situ approaches.Development of such datasets and models in whichthese data can be used requires field experimentsthat combine appropriate remote sensing measure-ments with traditional in-situ measurements inregions that are hydrologically well understood.Once hydrologic models are developed for use withremote sensing data in well-monitored basins, theycan possibly be extended to regions where little orno in-situ measurements exist.

Different remote sensing satellites carry sen-sors of varied characteristics. Often data are com-plementary in nature, for example panchromaticdata have high spatial resolution and multispec-tral data have low spatial resolution. Fine spatialresolution is necessary for an accurate descriptionof shapes, features, and structures, whereas finespectral resolution enables better discriminationbetween attributes (e.g. for classification of landcover). Hence, merging of these two types of datato form multi-spectral images with high spatial res-olution is useful for various applications such asvegetation mapping, land cover classification, pre-cision farming, and urban management.

Geographical Information System (GIS) is acomputer-assisted system for capturing, storage,retrieval, analysis, and display of spatial data anddata with non-spatial attributes. Some of these in-volve reclassification, aggregation, overlays, suit-ability analysis, network, and route analysis, opti-mization, allocation, etc. The data can be derivedfrom alternative sources such as survey, geograph-ical/topographical/aerial maps, or archived data.Data can also be in the form of location data (i.e.latitude/longitude) or tabular (attribute) data. Ap-plications of GIS range from simple database querysystems to complex analysis and decision supportsystems. Areas of application range from naturalresources management to near real-time applica-tion such as flood forecasting. GIS techniques areplaying an increasing role in facilitating integrationof multi-layer spatial information with statistical at-tribute data to arrive at alternative developmentalscenarios.

P1: SFK/UKS P2: SFK


EPILOGUE 399

Combining the ability of RS to measure spatial,spectral, and temporal information on the state ofthe Earth’s surface and that of GIS to handle andmanipulate geospatial data for a multitude of ap-plications in resource management, quantification,and process understanding, have become possible.These include understanding and simulating land-scape changes and hydrology, urban flood mod-elling, urban environment and impact assessment,generic ecosystem patterns, regional climate mod-els, and a host of other fields.

14.3 Water resource sustainability

The historic shift from population settlement basedon resource exploitation to economy-driven, trans-portation, and amenity-based settlement patternsposes great challenges for achieving sustainable wa-ter supplies and water management. The amenity-based settlement patterns put increasingly largerpopulation and land-use pressure on areas that pre-viously served as ‘water banks’ for meeting the re-quirement of urban population inhabiting the area.

Concentrated human settlements with theirpropensity to create hard, impermeable surfacesfor building houses and roads and the need forwater intake and outflow in a variety of forms,are not in harmony with the natural hydrologiccycle. The adverse effects of creating impervioussurface cover in urbanized watersheds, reducingthe groundwater recharge and the consequent re-duction in the base flow of the stream/river flow-ing through the area, are well documented. Sewer-age and water supply systems serving dense settle-ments can further interfere with groundwater andsurface water hydrology. An urban settlement alsocreates a ‘heat island’ effect, reduces evapotranspi-ration due to reduction in the vegitative cover, andmodifies the local microclimate.

The major driving force for change is essentiallypopulation growth coupled with a rising livingstandard globally, a combination that has resultedin over-exploitation of resources, including water.The current world population is about 6 billion,which is expected to grow to about 9 billion by theyear 2050. When the population was much smaller(e.g. <2 billion) and the per capita use of resources

was also much smaller, the traditional pattern of re-source consumption, namely, ‘take, make, waste’was sustainable. However, what is needed is to re-cycle and reuse all resources (including water) andalso increase the use of renewable resources. Wa-ter stress currently affects only a modest fractionof the human population, but it is expected to af-fect 45% of the population by the year 2025. Thissituation may be further exacerbated by global cli-mate change, which may alter water supply andstorage patterns in ways that could render existingwater-management infrastructure ineffective.

Recycling technologies can significantly reducethe net water abstraction from the environment,but many of these technologies require an increasein the use of other resources, especially energy.In our resource-constrained world, increasing theconsumption of any resource, such as water, mustbe carefully considered.

Another aspect of water stress caused by urbanwater-management systems is increased load of nu-trients, especially phosphorus, entering the aquaticenvironment. Mined as phosphate rock, phospho-rus is used for manufacturing fertilizers that arewidely used to increase the yield of crops for hu-man consumption. Phosphorus and other nutrientsthen pass through the human body metabolism andend up in the wastewater discharge. When theseeffluents are discharged into the aquatic environ-ment, the excess nutrients can cause eutrophica-tion of surface water bodies. At the current rate ofconsumption, the supply of phosphate, an essentialnutrient for crops with no known replacement, isexpected to be exhausted in about 100 years. Thus,there is an urgent need to recover phosphate fromwastewater.

Two other factors must be taken into consid-eration. First, although water supply is uniformlyprovided in the developed world, approximately 1billion people in the developing world do not haveaccess to safe drinking water, and more than 2.5billion do not have access to adequate sanitation.Clearly, to meet global needs, more efficient urbanwater and waste management systems are needed.

Some of the aspects of urbanization that exertthe most obvious influence on hydrologic pro-cesses are the increase in population density andthe proportion of built-up areas within urbanized

P1: SFK/UKS P2: SFK



areas. With an increase in population, water de-mand begins to rise. The increase in water demandis accelerated with rising living standards, whichfurther compounds the problem of developingadequate water resources – the first of the majorurban hydrologic problems.

Due to increased urbanization and with the in-stallation of sewerage systems for both domesticand storm water, the amount of water-borne wasteload also increases in proportion to populationgrowth. The resultant water quality changes areintimately linked to the increase in population den-sity. As the latter rises, the extent of imperviousbuilt-up area also increases, the natural drainagesystem gets modified, and the local microclimatechanges. Owing to the larger proportion of areabecoming impervious, a higher fraction of the inci-dent rainfall appears as runoff compared to the situ-ation when the catchment was in its pristine state.Furthermore, laying of storm sewers, realignmentof natural stream channels, and construction of cul-verts result in more rapid transmission of runoff tothe drainage network. The increase in inflow veloc-ities directly affects the nature of the runoff hydro-graph. Since a large volume of runoff is dischargedwithin a short time interval, peak rates of flow in-evitably increase, giving rise to flash floods, the sec-ond of the major urban hydrologic problems, suchas water-logging during heavy rain spells.

Municipal wastewater is a combination of liquid-or water-transported wastes originating in the sani-tary systems of dwellings, commercial or industrialunits, and institutions, in addition to any ground-water, surface water, and storm water that maybe present. Untreated wastewater generally con-tains high levels of organic matter, nutrients, andtoxic compounds as well as numerous pathogenicmicro-organisms. It thus entails environmental andhealth hazards and, consequently, must immedi-ately be conveyed away from its source locationsand treated appropriately before its final disposal.The ultimate goal of wastewater management isprotection of the environment in a manner com-mensurate with public health and socio-economicconcerns of the area.

Physical, chemical, and biological methods areused to remove contaminants from waste water.In order to achieve different levels of contaminant

removal, individual wastewater treatment proce-dures are combined into a variety of systems, classi-fied as primary, secondary, and tertiary treatments.Physical unit operations, in which physical meth-ods are applied to remove contaminants, includescreening, comminution, flow equalization, sedi-mentation, flotation, and granular medium filtra-tion. Chemical processes used in wastewater treat-ment are designed to bring about some form ofchange in the redox conditions by means of chemi-cal additives and/or reactions. These are invariablyused in conjunction with physical unit operationsand biological processes and include chemical pre-cipitation, adsorption on activated carbon, disinfec-tion, dechlorination, etc. Biological unit processesare used to convert the finely divided and dissolvedorganic matter in waste water into flocculent set-tleable organic and inorganic solids. In these pro-cesses, micro-organisms, particularly bacteria, con-vert the colloidal and dissolved carbonaceous or-ganic matter into various gases and build their celltissues that are subsequently removed in sedimen-tation tanks. Biological processes are generally usedin conjunction with physical and chemical pro-cesses, with the main objective of reducing theorganic and nutrient loads of waste water. Biolog-ical processes used for wastewater treatment maybe classified under five major categories, namely:(i) aerobic processes; (ii) anoxic processes; (iii)anaerobic processes; (iv) combined processes; and(v) pond processes. The commonly used biologicalprocesses include trickling filters, activated sludgeprocess, aerated lagoons, rotating biological con-tactors, and stabilization ponds.

Natural systems for wastewater treatment aredesigned to take advantage of physical, chemical,and biological processes that occur in the naturalenvironment where water, soil, plants, micro-organisms, and the atmosphere are in constantinteraction with each other. Natural treatmentsystems include land treatment, floating aquaticplants, and artificially created wetlands. All naturaltreatment systems are preceded by some formof mechanical pre-treatment for removal of grosssolids. Where a sufficient land area suitable forthis purpose is available, these systems can oftenbe the most cost-effective option in terms oftheir construction as well as operation. They are

P1: SFK/UKS P2: SFK


EPILOGUE 401

generally well suited to small communities andrural areas.

Even though modern water supply and sanitationis considered the most significant contribution topublic health in the past 150 years, rapid urbaniza-tion is challenging the sustainability of the devel-opment process and new approaches to water andsanitation management are urgently needed. Threeaspects of urban water management are emerg-ing as increasingly significant and will continue tobe important in the foreseeable future. These as-pects are decentralized wastewater management(DWM), wastewater reclamation and reuse, and in-creased attention to wet-weather flow (WWF) man-agement.

Harvesting of rainwater, either directly fromhouse rooftops, the runoff from private/publicland, or natural water collection areas, is an optionthat holds significant promise in any sustainablewater resource management strategy. In addition,groundwater, another natural resource represent-ing natural subsurface accumulation of rainwaterover timescales ranging from a few minutes tocenturies to millennia, provides an importantreserve to be exploited during periods of failure/deficit of rain. Due to its ubiquitous nature andrather simple technology required for its exploita-tion, the groundwater resource has been widelyover-exploited in the last few decades, while atthe same time rainwater harvesting at householdand community levels has been neglected. Severaltechnological tools and practice of rainwaterharvesting for three important applications,namely potable use at household and communitylevels, agriculture including horticulture at a farmscale, and for artificial groundwater recharge,are available. The scope of RWH is wide and thetechnologies discussed can be applied, with orwithout minor modification and innovation, in anyregion facing water scarcity (for whatever reason),but still having unutilized potential of rainwater.

Fortunately, most water resources are renew-able (except some groundwaters), albeit with hugedifferences in availability in different parts of theworld and wide variations in seasonal and annualprecipitation in many places. Human influence onuseable water is now a global phenomenon andplays a significant role in the hydrologic cycle.

Per capita use of water is increasing (with betterlifestyles) and the world population is also con-stantly growing. Human activity, in turn, impactsthe availability of clean water with its concomitanthealth implications. Access to water has also di-rect implications on poverty alleviation, economicgrowth, and development in general. Competitionfor water is intensifying day by day with the pro-gressive collapse of traditional water-based ecolog-ical systems, diminishing river flows, and ground-water depletion arising from over-exploitation.Thus water as a symbol of life, purity, and re-generation is under threat in large parts of theworld.

The human dimensions in water resource de-velopment and management encompass a diverserange of issues and policy domains. Ensuring sus-tainable and equitable resource development re-quires consideration of ethical principles involvingquestions of right or wrong from socio-economicas well as moral perspectives. Some of the impor-tant water-related challenges, with significant ethi-cal considerations, have been identified as:

� meeting basic needs – for safe and adequate wa-ter supply and sanitation;

� securing the food supply – especially for thepoor and vulnerable section of the populationthrough effective use of water;

� protecting ecosystems – ensuring their integritythrough sustainable water resource manage-ment;

� sharing water resources – promoting coop-eration between different users of water andbetween concerned states/countries, throughapproaches such as sustainable river basinmanagement;

� managing risks – to provide security from arange of water-related hazards, such as droughts,floods, pollution, etc.;

� valuing water – to manage water in the lightof its different values (economic, social, environ-mental, and cultural) and to move towards pric-ing of water to recover the costs of providing theservices, taking into account the equity and theneeds of the poor and vulnerable sections ofthe population;

P1: SFK/UKS P2: SFK



� managing water prudently – involving the in-terests of the public as well as various stakeholders;

� water and industry – promoting cleaner in-dustrial environment, particularly with regardto water quality and the needs of varioususers;

� water and energy – assessing the key role ofwater in energy production to meet rising energydemands;

� ensuring the accessibility of knowledge base –so that water knowledge becomes more easilyavailable globally to all concerned;

� water and cities – recognizing the distinctivechallenges posed by increase in the number ofurbanized regions throughout the world.

In addition to ‘water ethics’, the amount ofwater embedded in the production of goods orservices, referred to as ‘virtual water’, is anotherrecent concept that may significantly influence theregional and global commodity trade and waterallocation for the various competing demands. It

is likely to lead to more productive uses of water,even though water is not the only componentof any decision-making process. As with ‘carbonfootprints’, adding up all virtual water in theproducts that are used in daily life and the watercoming out of a tap leads to the idea of one’s ‘waterfootprints’ – a concept that helps to understandthe impact of an activity, individuals, communities,and nations on limited freshwater resources of theEarth.

Lastly, four case studies from three continentscovering regions with high water stress have beendescribed in Chapter 13, to highlight the variousissues related to understanding of the hydrologicsystem, technology, and societal concerns and howthese are being addressed to ensure sustainabledevelopment of water resource of each of theseregions. Evolving region-specific adaptation mea-sures to mitigate the situation of water stress arealso described. Sound knowledge of fundamentalprinciples and advancements in various disciplinesprovide important clues to ensure the survival ofhumankind on the Earth.

modern hydrology and sustainable water development (gupta/modern hydrology and sustainable water...

Documents