Agricultural Engineering · Land & Water Resources Engineering · also Floods and Control Measures, Minor Irrigation and CAD, Management of Canal Irrigation Systems
Managing water at the scale of reservoirs, canals and river basins, including floods and farmer participation.
In the exams: ICAR AICE subject 40 Units 7–8; ICAR AIEEA-PG Unit VII; electives in the ICAR 6th Deans' Committee B.Tech
Water resources management is the engineering and governance of water at the scale where single fields and single watersheds stop being the unit of decision: a country's total supply, a river basin, a reservoir serving a hundred thousand hectares, a barrage that diverts a river into a canal network, a flood that threatens a district. The agricultural engineer meets this scale whenever a farm's water comes from a canal, a tank or a public tubewell, and whenever a flood plain is also a farm. Agriculture takes roughly 70% of the world's freshwater withdrawals by the FAO's AQUASTAT accounting, so the way reservoirs are sized, canals are run and floods are managed decides the water supply of most of the world's irrigated land.
This subject follows the curriculum's nine units. It starts with how a country assesses and plans its water, including basin management, inter-basin transfer and climate change. It then covers reservoirs (storage zones, mass-curve sizing, yield and sedimentation), diversion headworks and the theories used to design weir floors on permeable foundations, and how canal systems are operated and water is shared. Floods are treated as two units: estimating the design flood (empirical formulas, the rational method, the unit hydrograph and frequency analysis with Gumbel and log-Pearson Type III), and controlling floods with structural and non-structural measures, including Muskingum routing. The last three units are minor irrigation, water users' associations and participatory irrigation management, and water quality management.
The companion subjects Watershed Hydrology, Irrigation Engineering and Drainage Engineering teach the basics this subject builds on: rainfall analysis, infiltration, hydrograph derivation, crop water requirement, canal section design, drain spacing and the chemistry of saline and sodic soils. Here those tools are used for planning and operating systems, and the manual refers back to them rather than repeating them.
Worldwide, the subject is examined in India in ICAR AICE-JRF/SRF subject 40 (Soil and Water Conservation Engineering, units 7 and 8), in ICAR AIEEA-PG Unit VII and in state public service commission papers on dams, weirs and canals; flood frequency and routing also appear in the hydrology block of GATE AG. In the Philippines it sits in the Land and Water Resources part of the ABE licensure exam, and in the USA under dams, reservoirs and stormwater in the NCEES PE Agricultural and Biological exam. Practice differs between countries: Indian classification of projects, Philippine irrigators' associations, European and US water quality law are named where they apply.
A national water resources assessment answers four questions: how much water arrives, how much of it is renewable each year, how much can be used, and how much is already used. Precipitation is the gross input. Part returns to the atmosphere by evapotranspiration; the rest becomes surface runoff and groundwater recharge, which together make up the internal renewable water resource. Inflow from upstream countries is added to give the total renewable resource, with care not to count twice the groundwater that later discharges into rivers as baseflow.
Not all renewable water is usable. Utilisable water is the part that can be captured by storage, diversion or wells given the topography, the timing of flows and the engineering and economic limits of the day. In a monsoon climate most runoff arrives in three or four months, so the utilisable fraction depends heavily on storage. Assessments therefore separate the average annual resource from the utilisable resource, and the utilisable surface water from the replenishable groundwater.
India illustrates the bookkeeping. The Central Water Commission has long quoted average annual precipitation of about 4,000 billion cubic metres (BCM), an average annual water resource of 1,869 BCM, and utilisable resources of 1,122 BCM (690 BCM surface and 433 BCM groundwater); a 2019 reassessment using satellite inputs gave about 1,999 BCM for the average annual resource. Other countries publish comparable national balances, and FAO's AQUASTAT database compiles them on a common basis. Figures are revised as methods and records improve, so a planner always quotes the source and year.
Single numbers help compare countries. The Falkenmark indicator divides renewable water by population: above 1,700 m³ per person per year is taken as no stress, 1,000 to 1,700 as water stress, 500 to 1,000 as water scarcity and below 500 as absolute scarcity. It is simple and widely quoted but ignores where and when the water arrives and what people use it for. The withdrawal-to-availability ratio (criticality ratio) compares use with supply; values above about 0.4 are commonly read as high stress. The Sustainable Development Goal indicator 6.4.2 uses withdrawals as a share of renewable resources after environmental flow needs.
Per-capita availability falls as population grows even if rainfall is unchanged, which is why many monsoon countries have moved from comfortable to stressed on the Falkenmark scale within two generations. The indicator also hides spatial imbalance: a country may be stressed on average while one basin is in surplus and another is overdrawn.
Demand is grouped into irrigation, domestic and municipal supply, industry, energy (cooling and hydropower, which uses but does not consume most of its water), and environmental flows needed to keep rivers and wetlands functioning. Planners distinguish withdrawal, the water taken from a source, from consumptive use, the water evaporated, transpired or incorporated and not returned. Return flows from irrigation can be reused downstream, so improving field efficiency in one place may not free water for the basin as a whole.
Water resources planning sets objectives (food production, drinking water, energy, flood safety, ecosystems), projects demand over a planning horizon of 25 to 50 years, inventories supply, and screens alternatives: new storage, groundwater development, demand management, reuse, and transfers. Projects are compared by economic analysis. The capital cost is converted to an equivalent annual cost with the capital recovery factor, CRF = i(1+i)^n / [(1+i)^n − 1], for interest rate i and life n years; operation and maintenance is added, and the benefit-cost ratio is annual benefits divided by annual costs. A ratio above 1 is necessary but not sufficient; distribution of benefits, environmental and social impacts and resettlement are appraised as well.
Integrated water resources management (IWRM) is the framework most countries now state in their policies. The 1992 Dublin principles hold that fresh water is finite, vulnerable and essential to life; that development and management should be participatory; that women play a central part in providing and safeguarding water; and that water has an economic value in all its competing uses.
Water moves within a basin regardless of district, state or national boundaries, so planning by basin lets upstream and downstream uses be weighed together. River basin organisations range from advisory committees to authorities that build and operate works. Well-known models include river basin agencies in France, the Murray–Darling Basin Authority in Australia, and river boards or tribunals in federal countries. Their tasks are a basin water balance, allocation among users and states, licensing of withdrawals, coordinated operation of reservoirs, flood management, and water quality planning.
Allocation rules differ. Prior appropriation gives senior rights first call in shortage (western USA); riparian doctrine ties rights to land beside the water (much of common-law practice); proportional sharing gives each user a fixed share of whatever is available; and in federal systems tribunals or compacts divide flows between states at a stated dependability. For international rivers the 1997 UN Watercourses Convention sets out equitable and reasonable utilisation and the obligation not to cause significant harm, though not every basin state is a party.
A basin plan should state the environmental flow: the quantity, timing and quality of flow needed to sustain the river ecosystem and the people who depend on it. Methods range from hydrological rules (a percentage of mean flow) to holistic methods built with ecologists and users.
An inter-basin transfer moves water from a basin judged to have surplus to one in deficit, by canal, tunnel or pipeline, often with pumping. Examples include the South–North Water Transfer Project in China and the Snowy Mountains Scheme in Australia; India's National Perspective Plan of 1980 proposed Himalayan and peninsular river-linking components.
The engineering questions are the transferable surplus at an agreed dependability, the conveyance route and its losses, pumping energy, and storage at both ends. The harder questions are whether the surplus is real after the donor basin's future needs and environmental flows are counted, the effects on sediment delivery and the delta downstream, displacement of people, mixing of aquatic species between basins, and cost per cubic metre delivered compared with demand management or local storage in the receiving basin. A planner presents transfers as one option among several, appraised on the same economic, environmental and social criteria.
A warmer atmosphere holds more water vapour, roughly 7% more per degree Celsius by the Clausius–Clapeyron relation, and the IPCC assessments report that this intensifies the water cycle: heavier short-duration rainfall in many regions, more variable monsoons, earlier snowmelt and shrinking glaciers that change the timing of river flow, and more intense droughts where evaporative demand rises faster than rainfall. Sea-level rise pushes salt water into coastal aquifers and deltas.
For engineering this means the past record may not be a reliable sample of the future, the assumption of stationarity behind classical frequency analysis. Responses include updating design rainfall and flood estimates as records lengthen, testing designs against a range of climate scenarios rather than a single forecast, adding safety margins where failure is costly, choosing measures that perform reasonably across many futures (storage that can be expanded, flexible operating rules, demand management, groundwater recharge), and monitoring so that plans can be revised. Agriculture adapts through crop and calendar changes, more efficient irrigation, and protecting soil moisture.
A reservoir stores water when it is plentiful and releases it when it is needed. A conservation reservoir serves irrigation, municipal supply, industry or hydropower; a flood-control reservoir keeps space empty to absorb flood peaks; a multipurpose reservoir does both, which means its purposes conflict, because conservation wants the pool high and flood control wants it low. Distribution reservoirs and service tanks in a supply system balance hourly demand and are a different scale of problem.
Site selection weighs a narrow gorge for a short dam with a wide valley upstream for storage, sound foundation and abutments, availability of construction materials, a site for the spillway, low sediment yield from the catchment, minimum submergence of land, villages and forests, and water tightness of the reservoir rim.
The volume of a reservoir is divided by levels. Dead storage lies below the minimum drawdown level (MDDL), set by the sill of the lowest outlet; it is not released in normal operation and is meant to receive sediment. Live or useful storage lies between the MDDL and the full reservoir level (FRL), also called normal pool level, and is the water actually used. Above the FRL, up to the maximum water level (MWL) reached during the design flood, lies surcharge storage, which is temporary and uncontrolled when the spillway is ungated. A flood-control reservoir also keeps a flood cushion, a controlled space below the spillway crest kept empty during the flood season. Above the MWL comes freeboard to the top of the dam, which allows for wind set-up and wave run-up.
Two further storages matter in operation. Bank storage is water that soaks into the permeable banks when the reservoir fills and drains back as it falls. Valley storage is the volume the natural river channel and flood plain already held before the dam; the net gain in storage from building the dam is the reservoir volume minus valley storage, which matters when assessing the flood benefit.
Every reservoir study starts with the area and capacity at each level. Areas are measured from contour maps or a digital elevation model at regular contour intervals. The volume between two contours with areas A1 and A2 a height h apart is found by the trapezoidal formula, ΔV = h(A1 + A2)/2, or more accurately for a basin-shaped valley by the cone (prismoidal) formula, ΔV = (h/3)(A1 + A2 + √(A1·A2)). Summing the increments gives the capacity curve. The trapezoidal formula always gives the slightly larger value. Prismoidal formulas using alternate contours are also used.
The same curves give the water spread area for evaporation estimates, the volume lost to sediment at each level after a resurvey, and the level-volume relation used in flood routing. Spillways pass flood water safely past the dam: the ogee (overflow) spillway on concrete dams, the chute and side-channel spillways common with earth dams, the shaft or morning-glory spillway at constricted sites, and the siphon spillway for small, closely regulated heads. Many dams also have a saddle or emergency spillway. The spillway design flood is chosen by the size and hazard of the dam: in India IS 11223 links large dams to the probable maximum flood, intermediate dams to the standard project flood and small dams to the 100-year flood; other countries classify by downstream hazard, so the rule must be checked locally.
The storage needed to sustain a given draft is found from the record of inflows. The mass curve (Rippl diagram, 1883) plots cumulative inflow against time. A uniform demand is a straight line with slope equal to the draft rate. Lines of this slope are drawn tangent to each peak of the mass curve; the largest vertical gap between a tangent and the following trough of the mass curve is the storage needed to meet the demand through the driest period of the record. If the tangent from a peak never meets the mass curve again, the demand exceeds the long-term mean inflow and cannot be met by any storage.
The method assumes the reservoir is full at the start of the critical period and that the record represents future inflows. It can include evaporation and other losses by adding them to the demand, and a varying demand by drawing a demand mass curve instead of a straight line.
The sequent peak algorithm does the same calculation in a table, which suits computers and long records. Let K be the cumulative deficit at the end of each period. Starting with K0 = 0, set K_t = max(0, K_(t−1) + D_t − Q_t), where D is the demand and Q the inflow in the period. The required storage is the largest K in the sequence. Running the record twice in succession handles a critical period that wraps around the end of the record.
Worked example: monthly inflows of 20, 35, 60, 80, 45, 25, 12, 8, 5, 4, 6 and 10 Mm³ (total 310 Mm³, mean 25.8 Mm³ per month) and a uniform demand of 20 Mm³ per month. Deficits build only in the six dry months with inflows of 12, 8, 5, 4, 6 and 10 Mm³: 8 + 12 + 15 + 16 + 14 + 10 = 75 Mm³. The live storage needed is therefore 75 Mm³. Common mistakes are to take the single worst month (16 Mm³) or the annual surplus (310 − 240 = 70 Mm³).
Yield is the amount a reservoir can supply in a stated period. Firm (safe) yield is the largest draft that can be met throughout the critical dry period of the record. Secondary yield is water available in excess of firm yield in wetter years, useful for purposes that can accept interruption. Design yield is set at a stated dependability: Indian practice commonly plans irrigation projects at 75% dependability, hydropower at 90% and drinking water at 100%, while other countries use reliability or risk criteria; always check the local norm.
Repeating the mass-curve or sequent-peak calculation for several drafts gives the storage-yield curve. Read the other way, it gives the yield from a given storage. With the inflows above and 60 Mm³ of live storage, the dry-period deficits (D − 12) + (D − 8) + (D − 5) + (D − 4) + (D − 6) + (D − 10) = 6D − 45 must equal 60, so D = 17.5 Mm³ per month. Reliability is often expressed as the fraction of time periods in which the full demand is met, alongside resilience (how quickly the system recovers from a failure) and vulnerability (how large a shortfall is). Simulation over a long record, or over synthetic sequences generated by stochastic models, gives these measures for a proposed size and operating rule.
Every reservoir on a sediment-carrying river loses capacity. As the river enters the pool its velocity falls, coarse material drops to form a delta at the head of the reservoir, and finer silt and clay settle further in or travel as a density current along the bed toward the dam. The fraction of incoming sediment that stays is the trap efficiency. Brune's curve relates trap efficiency to the capacity-inflow ratio (reservoir capacity divided by mean annual inflow); large storage reservoirs with a ratio near or above 1 trap 95% or more, while small reservoirs on large rivers pass much of their load. Churchill's curve, based on a sedimentation index, is used for settling basins and reservoirs that are often emptied.
The annual volume deposited is: sediment inflow (tonnes per year) × trap efficiency ÷ dry bulk density of the deposit (tonnes per cubic metre). Bulk density increases as deposits consolidate, so deposited volumes per year decrease with time for the same mass. Example: a catchment of 800 km² yielding 1,200 t/km²/yr with 95% trap efficiency and bulk density 1.2 t/m³ deposits 800 × 1,200 × 0.95 ÷ 1.2 = 760,000 m³ a year; a 12 Mm³ dead storage would hold about 16 years of such deposits, which shows why sediment does not stay neatly in the dead storage zone.
Reservoir life is taken as the time until sediment fills the dead storage, or until a stated fraction of the original capacity (often half) is lost, depending on the agency. Because trap efficiency falls as capacity shrinks, life is computed in steps: take a slice of capacity loss, find the mean trap efficiency for it, divide by the sediment inflow, and sum. Deposits do not stay in the dead storage; they spread through the live storage as well. Empirical methods such as the area-reduction method of Borland and Miller distribute the expected deposit over the elevation-capacity curve to estimate the new curve after a given number of years.
Sediment is managed at three points. In the catchment, soil conservation, check dams and afforestation reduce yield, though in large basins the effect on the reservoir is slow and partial. In the reservoir, sluicing passes sediment-laden flood water through low-level outlets before it settles, venting releases turbidity currents, flushing draws the reservoir down so the river scours deposits, and dredging removes deposits mechanically at high cost. Around the reservoir, bypass tunnels or off-channel storage keep sediment-laden floods out of the pool. Repeat hydrographic surveys measure actual losses.
Evaporation from the water surface is the main loss in hot climates. It is estimated from pan evaporation multiplied by a pan coefficient (commonly about 0.7 for the Class A pan, but the value varies with climate and pan type) and the water spread area. A 15 km² reservoir with 6 mm/day pan evaporation and a pan coefficient of 0.7 loses 15 × 10⁶ × 0.0042 × 30 = 1.89 Mm³ in a 30-day month. Seepage through the dam and foundation and into the rim adds to losses. Evaporation suppression with films and covers has been tried on small reservoirs.
Operation follows rule curves: target levels by date that the reservoir tries to follow, with zones that trigger restrictions when the level falls. The standard operating policy releases the full demand whenever water is available and spills when full; hedging rules deliberately cut releases early in a drought to avoid a deep shortage later. A multipurpose reservoir lowers its pool before the flood season to create a flood cushion and fills on the recession; the balance between missing the refill and keeping flood space is the central operating decision. Flood forecasts allow pre-releases that empty space ahead of a forecast flood.
A diversion headwork raises the river level enough to divert flow into a canal without large storage. A weir raises the water by a solid crest across the river, sometimes topped by small shutters; a barrage has a low crest and controls the pond level with gates across the full width, which gives better control of level and of silt passing downstream, and less afflux in floods, at higher cost. Both are built mostly on alluvial, permeable foundations, which is why their floors are designed for seepage as well as for the hydraulic jump.
The components of a typical headwork are the weir or barrage proper; undersluices (scouring sluices) in a deep pocket beside the canal head, which keep a clear approach channel and flush deposited silt; a divide wall separating the undersluice pocket from the rest of the weir; a fish ladder or fish pass so fish can move upstream; the canal head regulator, which controls entry to the canal; silt excluders in the pocket floor that draw off bed-load-rich bottom water before it reaches the regulator; silt ejectors (extractors) a short way down the canal; and river training works, such as guide banks that confine the river to a central waterway, marginal bunds and spurs (groynes).
Water ponded upstream seeps under the floor and emerges downstream. Two dangers follow: uplift pressure on the floor, and piping, where the emerging flow carries soil away and undermines the structure. Bligh's creep theory (1910) treats the seepage path, the creep length, as the total length of contact between water and the structure along the floor and both faces of every vertical cutoff. Head loss is assumed uniform along it. The safe creep length is L = C·H, where H is the head difference and C is Bligh's creep coefficient for the soil (larger for fine sand than for gravel); the safe hydraulic gradient is 1/C. A head of 4 m on a soil with C = 12 needs 48 m of creep length.
The floor must resist the residual uplift head h at a point. Without a safety factor, its thickness is t = h/(G − 1), where G is the specific gravity of the floor material; Bligh's own practice multiplies this by 4/3. Lane's weighted creep theory (1935), from a study of many structures, counts vertical contact fully but horizontal contact at only one-third, because vertical paths resist seepage better; the weighted creep ratio is (ΣL_v + ΣL_h/3)/H and must exceed a soil-specific value.
Khosla and his co-workers (1936) showed from observations of weirs in Punjab that seepage follows streamlines set by the geometry, not a straight creep path, and that the critical quantity is the exit gradient at the downstream end, where the upward seepage force can lift soil particles. For a floor of length b with a single downstream cutoff of depth d and head H, the exit gradient is G_E = (H/d) × 1/(π√λ), where λ = [1 + √(1 + α²)]/2 and α = b/d. Safe exit gradients quoted in textbooks are about 1/4 to 1/5 for shingle, 1/5 to 1/6 for coarse sand and 1/6 to 1/7 for fine sand. A deep downstream cutoff is the most effective way to lower the exit gradient.
Uplift pressures at the key points of a floor with cutoffs are found by the method of independent variables: the complex profile is broken into simple standard forms (a floor with one end pile, an intermediate pile, a depressed floor), the pressure percentage at each key point is read from Khosla's solutions, and corrections are added for floor thickness, the mutual interference of piles, and slope. Khosla's method is the basis of later design codes for weirs and barrages on permeable foundations.
A gravity canal system carries water from a headwork or reservoir through a main canal, branch canals, distributaries and minors to outlets, below which farmers' watercourses (field channels) take it to the fields. The area that can be irrigated by gravity from an outlet is its command; culturable command area (CCA) excludes uncultivable land within the gross command. Canal section design and lining are covered in Irrigation Engineering; this chapter deals with running the system.
Control structures make operation possible. A head regulator controls flow into an offtaking canal; a cross regulator across the parent canal raises or holds the water level so offtakes can draw their share, especially when the parent runs below full supply; an escape releases surplus water safely to a drain or river; canal falls drop the bed where the ground slope is steeper than the canal slope, and can carry a meter flume; and measuring structures record discharges at key points.
An outlet (turnout) passes water from a distributary to a watercourse. Outlets are classed as non-modular, where discharge depends on both upstream and downstream levels (a submerged pipe outlet); semi-modular or flexible, where discharge depends on the upstream level but not the downstream one as long as a minimum working head exists (pipe fall outlet, open flume, adjustable proportional module); and modular, where discharge is fixed within limits regardless of either level (Gibb's module).
For an outlet q = C·H^m and a distributary Q = K·D^n, where H is head on the outlet and D the depth in the distributary, flexibility F = (dq/q)/(dQ/Q) = (m/n)(D/H). A proportional outlet (F = 1) shares a change in canal flow fairly; a hyper-proportional outlet (F > 1) gains more than its share when the canal rises; a sub-proportional outlet (F < 1) gains less. For an open flume with m = 3/2 in a wide channel with n = 5/3, proportionality requires the crest to be set at H/D = m/n = 0.9. Setting outlets correctly is the main way designers protect tail-end farmers.
Water is delivered in one of three ways. Continuous flow keeps every channel running at a fixed share and suits large holdings and rice; it wastes water where demand is low. Rotational supply runs channels or outlets in turn at full capacity, which keeps velocities high enough to avoid silting and makes losses proportionally smaller. Demand-based delivery lets farmers order water, either on demand from a pipe network or by arranged demand with notice; it needs storage close to the users and good communication.
Warabandi, used in north-west India and Pakistan, is a rotational system in which each farmer on a watercourse receives the whole outlet flow for a fixed time, proportional to his landholding, on a fixed day and hour in a weekly cycle. A farmer with 6 ha in a 120 ha chak (outlet command) on a 7-day (168-hour) rotation gets 168 × 6/120 = 8.4 hours, before adjustments for the time water takes to reach his field. Rotation among distributaries is called osrabandi in that region. In Maharashtra, the shejpali system allots water against farmers' applications for crops and areas each season.
Conveyance efficiency is the fraction of water entering a canal that reaches the outlet or field. Losses in series multiply: if the main canal loses 15%, the distributaries 10% and the watercourses 20%, only 0.85 × 0.90 × 0.80 = 0.612 of the head discharge reaches the fields, so 10 m³/s at the head gives about 6.1 m³/s at the farm gate. Seepage is the largest loss in unlined channels in permeable soil; lining, compaction and maintenance reduce it, though water seeping from canals may recharge groundwater used by wells downstream.
Common operating problems are head-tail inequity (head reaches take more than their share and tails go short), unauthorised outlets and breaches, siltation and weed growth that reduce capacity, deferred maintenance, and mismatch between canal schedules and crop needs. Modernisation responds with measurement at every level, operating plans for each season, downstream control gates that respond automatically to demand, remote monitoring (SCADA), and transfer of lower levels to water users' associations. Performance is assessed with indicators such as relative water supply (water supplied including effective rainfall divided by crop demand), adequacy, equity among head and tail, and dependability over time.
The design flood is the flood a structure is designed to pass or withstand. Its magnitude depends on the hazard and the cost of failure: a farm culvert may be sized for a 10- or 25-year flood, a levee for 50 or 100 years, and a major dam spillway for the probable maximum flood (PMF), the largest flood reasonably possible from the most severe combination of meteorological and hydrological conditions, derived from the probable maximum precipitation (PMP) applied to the catchment through a unit hydrograph. The standard project flood (SPF) is derived from the most severe storm considered reasonably characteristic of the region rather than the extreme.
The choice of method follows the data. With no flow records, empirical formulas, envelope curves or the rational method are used for small catchments. With rainfall records and a unit hydrograph (observed or synthetic), a design storm can be converted to a design flood hydrograph. With a long record of annual peak flows, statistical frequency analysis is used directly, and regional analysis pools data from similar gauged catchments when the local record is short.
Empirical formulas relate peak flow to catchment area alone, with a coefficient fitted to a region. Dickens' formula, Q = C_D·A^(3/4), was developed for north and central India; Ryves' formula, Q = C_R·A^(2/3), for south India, with C_R of about 6.8 near the coast, 8.5 inland and up to 10.2 near hills; Inglis' formula for Maharashtra; and Fuller's and other formulas elsewhere. Q is in m³/s and A in km². Their coefficients carry all the hydrology, so they are only valid in the region they were derived for, and they give no return period. Envelope curves plot the largest observed floods against area for a region and give an upper bound.
The rational method, Q = C·i·A/3.6 (Q in m³/s, i in mm/h, A in km²), uses rainfall intensity for a duration equal to the time of concentration and of the design return period. It suits small catchments, commonly a few square kilometres to a few tens, where the assumptions of uniform rainfall in space and time hold. With C = 0.35, i = 60 mm/h and A = 2.5 km², Q = 14.6 m³/s; forgetting the 3.6 gives 52.5. Watershed Hydrology covers the derivation of C, time of concentration and IDF curves.
For catchments of a few hundred to a few thousand square kilometres, the unit hydrograph converts a design storm into a full hydrograph, which is needed whenever storage or routing matters. A design storm of chosen duration and return period is divided into blocks equal to the unit duration, losses are subtracted (φ-index or curve number), and each block of effective rainfall is multiplied by the UH ordinates, lagged and summed; baseflow is added. If a 6-h UH has a peak of 50 m³/s per cm, an effective rainfall of 3 cm in 6 h gives a direct runoff peak of 150 m³/s, and 170 m³/s with 20 m³/s baseflow. The area under a 1-cm UH equals 1 cm over the catchment, which checks the UH: ordinates summing to 300 m³/s at 6-h intervals imply a catchment of 300 × 6 × 3,600 ÷ 0.01 m = 648 km².
Synthetic unit hydrographs (Snyder, SCS dimensionless, or regional relations from catchment length, slope and shape) are used at ungauged sites.
With n years of annual maximum peaks, frequency analysis fits a probability distribution and reads off the flood of return period T. The return period is the average interval between exceedances, so the annual exceedance probability is p = 1/T. Plotting positions, most commonly Weibull's T = (n + 1)/m for rank m, place the observed floods on probability paper.
Gumbel's extreme value Type I distribution gives x_T = x̄ + K_T·σ. For a very long record, K_T = −(√6/π)[0.5772 + ln ln(T/(T − 1))], giving K = 1.305 for 10 years, 2.044 for 25, 2.592 for 50 and 3.137 for 100 years. For a finite record of n years, K_T = (y_T − ȳ_n)/S_n, where y_T = −ln ln(T/(T − 1)) and ȳ_n and S_n come from tables for n; this gives a larger K for short records. The log-Pearson Type III method takes logarithms of the peaks, computes their mean, standard deviation and skew coefficient, and uses frequency factors K_T from tables for that skew: log Q_T = ȳ + K_T·σ_y. US federal guidance (Bulletin 17, now 17C) recommends it.
Risk, the probability that a T-year flood is equalled or exceeded at least once in n years, is R = 1 − (1 − 1/T)^n. A 100-year flood has a 39.5% chance of occurring at least once in 50 years. To keep risk at 10% over a 25-year life, T = 1/[1 − 0.9^(1/25)] ≈ 238 years. This is why the design return period is much longer than the design life.
Flood routing computes how a flood hydrograph changes as it passes through a reservoir or a river reach. Storage attenuates the peak and delays it. Routing is how a planner tests whether a flood-control reservoir, a detention basin or a set of embankments achieves the target reduction downstream. Reservoir routing with the level-pool (modified Puls) method uses the storage-outflow relation from the capacity curve and spillway rating; Watershed Hydrology covers it step by step.
The Muskingum method routes floods through river reaches, where storage depends on both inflow and outflow. Storage is S = K[XI + (1 − X)O], where K is the storage constant (close to the travel time through the reach) and X a weighting factor between 0 and 0.5, commonly 0.1 to 0.3 for natural rivers. The routing equation is O2 = C0·I2 + C1·I1 + C2·O1, with C0 = (−KX + 0.5Δt)/D, C1 = (KX + 0.5Δt)/D, C2 = (K − KX − 0.5Δt)/D and D = K − KX + 0.5Δt; the three coefficients sum to 1. To keep them non-negative, Δt should lie between 2KX and 2K(1 − X). For K = 12 h, X = 0.2 and Δt = 6 h, C0 = 0.048, C1 = 0.429 and C2 = 0.524.
Structural measures change the flood itself or keep it away from people. Storage reservoirs and detention basins hold part of the flood volume and release it later; the storage needed to cut a peak to a safe channel capacity is the volume of hydrograph above that capacity. For a triangular flood of 900 m³/s peak and 48 h base and a safe channel capacity of 500 m³/s, the part above 500 is a similar triangle 400 m³/s high and 48 × 400/900 = 21.3 h wide, holding 0.5 × 21.3 × 3,600 × 400 ≈ 15.4 Mm³.
Embankments (levees, dykes) and floodwalls confine the flow; they are cheap per kilometre and protect large areas, but they raise flood levels in the confined channel, can fail by overtopping, seepage, piping and slope failure, trap water behind them (drainage congestion), and on aggrading rivers allow the bed to rise. Channel improvement (widening, deepening, cutting off meanders, clearing) increases conveyance. Floodways and diversion channels carry part of the flood around a town or into a depression. Watershed treatment reduces the frequent small floods and sediment but has little effect on very large floods once the soil is saturated.
Non-structural measures reduce the damage a flood causes without changing the flood. Flood plain zoning restricts land use according to flood risk: no buildings in the zone flooded most often, restricted and raised construction in the next, and normal use with precautions beyond. Flood forecasting and warning gives people time to move themselves and their animals and to protect property; it combines rainfall and river gauging networks, hydrological and hydraulic models, and a warning chain that reaches the last village. Flood proofing raises plinths, uses water-resistant materials and builds raised platforms or mounds as refuge. Flood insurance spreads losses and can send price signals against building in danger zones. Emergency preparedness, evacuation plans and relief and rehabilitation complete the set.
Experience worldwide, set out in the WMO and Global Water Partnership programme on integrated flood management, is that complete flood control is neither possible nor always desirable: flood plains depend on floods for fertility and groundwater recharge, and structures create a false sense of security that draws investment into protected areas until a larger flood overtops them. Integrated flood management combines structural and non-structural measures within the basin plan, aims to manage risk rather than eliminate floods, and gives space back to rivers where possible, as the Netherlands did in its Room for the River programme.
Minor irrigation (MI) schemes are small, quick to build and often owned or run by farmers or local bodies. India classifies irrigation projects by culturable command area: major above 10,000 ha, medium 2,000 to 10,000 ha and minor below 2,000 ha. Other countries use their own thresholds or speak of small-scale or farmer-managed irrigation. MI is divided into surface water schemes (tanks, ponds, small diversion weirs and anicuts, check dams, surface lift schemes from rivers and canals) and groundwater schemes (dug wells, dug-cum-bore wells, shallow and deep tubewells). In much of South Asia, groundwater MI now irrigates more land than canals do; India runs a periodic minor irrigation census to track the number and status of schemes.
Tanks are small reservoirs formed by an earthen bund across a shallow valley. The parts are the bund, the surplus weir (waste weir) that passes floods, one or more sluices that release water to the command, the supply channel that brings water from the catchment or from an upstream tank, and the foreshore that is submerged when the tank is full. In South India and Sri Lanka tanks are often linked in cascades, where the surplus of one feeds the next down the valley, so the cascade must be planned and rehabilitated as a whole.
Tank yield is estimated from catchment area, rainfall and a runoff coefficient (or a regional runoff table): a 15 km² catchment with 900 mm rain and a runoff coefficient of 0.25 yields 3.375 Mm³. If 30% is lost to evaporation and seepage and the crop needs a delta of 0.5 m, the irrigable area is 3.375 × 0.7 ÷ 0.5 = 4.73 Mm², about 473 ha. Siltation, encroachment of the foreshore and supply channels, and decline of community maintenance are the main causes of tank decline; rehabilitation programmes desilt, repair sluices and weirs and restore user organisations.
Lift irrigation pumps water from a river, canal or tank to a delivery chamber, from which it flows by gravity. The water power is ρgQH, and the input power is ρgQH/η: lifting 0.05 m³/s through 25 m with an overall efficiency of 0.7 needs 9.81 × 0.05 × 25 ÷ 0.7 ≈ 17.5 kW. Energy cost and reliable power supply decide whether a lift scheme survives; solar pumping is increasingly used for small lifts.
A well's command area follows from its discharge, pumping hours and the peak crop demand. A tubewell giving 30 L/s for 10 h a day delivers 1,080 m³ a day; with 75% application efficiency and a peak demand of 6 mm/day it serves 1,080 × 0.75 ÷ 0.006 = 135,000 m², or 13.5 ha. Well design, aquifer tests and pump selection belong to Groundwater, Wells and Pumps; in this subject the question is planning: spacing wells to avoid interference, conjunctive use of canal water and groundwater so that waterlogging in one place and depletion in another are both avoided, and artificial recharge.
Large canal projects in many countries were found to irrigate far less than designed because the works stopped at the outlet. India's Command Area Development programme, begun in 1974, funded on-farm development below the outlet: field channels and drains, land levelling and shaping, warabandi, consolidation of holdings, and extension. The same logic applies to minor schemes: the last few hundred metres to the field decide how much of the water stored or pumped becomes crop. On-farm works are also the natural responsibility of water users' associations, the subject of the next chapter.
Participatory irrigation management (PIM) means involving water users in the planning, operation, maintenance and financing of irrigation systems. Irrigation management transfer (IMT) goes further: the agency hands responsibility, and sometimes ownership, of part or all of a system to a users' organisation. The case for both grew from experience that agency-run canals often had poor maintenance, low fee recovery, head-tail inequity and little accountability to farmers, while farmer-managed systems in Nepal, the Philippines, Bali (subak) and elsewhere had run for generations.
Countries that undertook large transfers include Mexico (from 1989, transferring irrigation districts to user associations), Turkey (transfers to irrigation associations in the 1990s), the Philippines (irrigators' associations under the National Irrigation Administration), and India, where Andhra Pradesh passed the Farmers' Management of Irrigation Systems Act in 1997 and other states followed with their own laws.
A water users' association (WUA) is usually formed for a hydraulic unit, such as the command of a minor or distributary, not an administrative village, so that everyone sharing a channel is a member. A typical structure is a general body of all water users, an elected managing committee with a president, and sometimes higher-level federations at the distributary and project levels. Its functions are to prepare the seasonal operation plan, distribute water within its area, maintain channels, collect water charges and keep part of them for its own work, settle disputes, and keep records.
Water charges may be by area and crop, which is easy to administer but gives no incentive to save water, or volumetric, where the agency supplies a measured bulk volume to the WUA and the WUA charges its members; volumetric supply needs a measuring structure at the head of the WUA's area and is the stronger basis for accountability in both directions. Legal recognition, clear water rights or entitlements, a share of fees retained locally, and training in accounts and operation are the conditions most often found where WUAs work.
Elinor Ostrom's study of long-lived common-pool resource institutions, including irrigation systems, set out design principles that recur in successful WUAs: clearly defined boundaries of the system and its members; rules suited to local conditions, with benefits proportional to contributions; collective-choice arrangements in which those affected help make the rules; monitoring by people accountable to the users; graduated sanctions; cheap and quick conflict resolution; recognition by government of the right to organise; and, for larger systems, nested layers of organisation.
Reviews of transfer programmes by the World Bank, FAO and the International Water Management Institute found mixed results: fee collection and maintenance often improved and agency costs fell, but gains in water productivity and equity were less consistent, and many WUAs stayed weak where they had no real control over water deliveries or where they were formed on paper only. Performance is judged with indicators such as relative water supply, delivery performance ratio (actual over planned delivery), equity between head and tail, fee collection ratio, and the area actually irrigated against the target.
Water quality is judged against the intended use. Physical parameters are temperature, turbidity and suspended solids; chemical parameters include pH, electrical conductivity (EC) or total dissolved solids, major ions, nutrients (nitrate, ammonia, phosphate), dissolved oxygen (DO), and toxic substances such as arsenic, fluoride, heavy metals and pesticides; biological parameters are faecal indicator bacteria such as Escherichia coli and, for treated wastewater, helminth eggs. Organic pollution is measured by biochemical oxygen demand (BOD), the oxygen used by microbes to oxidise organic matter, usually over 5 days at 20 °C, and chemical oxygen demand (COD).
Drinking water standards follow WHO guidelines, adapted nationally; WHO guideline values include 10 µg/L for arsenic, 1.5 mg/L for fluoride and 50 mg/L for nitrate as NO3. Irrigation water is classed by salinity (EC), sodicity (SAR and residual sodium carbonate) and specific ion toxicity (boron, chloride, sodium) following FAO Irrigation and Drainage Paper 29; the calculation of SAR and leaching needs is taught in Drainage Engineering.
Point sources, such as sewage and industrial outfalls, discharge at a known place and can be treated and permitted. Non-point (diffuse) sources, of which agriculture is the largest, arrive across the landscape: nitrate leaching to groundwater, phosphorus attached to eroded soil, pesticides, salts in irrigation return flow, and animal waste. Where a waste stream enters a river, the concentration downstream after full mixing is the flow-weighted mean: 20 m³/s at 2 mg/L of nitrate mixed with a 5 m³/s drain at 22 mg/L gives (40 + 110) ÷ 25 = 6 mg/L.
BOD exerted by time t follows BOD_t = L0(1 − e^(−kt)), where L0 is the ultimate BOD and k the rate constant (base e). With L0 = 200 mg/L and k = 0.23 per day, BOD5 = 200(1 − e^(−1.15)) ≈ 137 mg/L. Some texts write the same law with base 10, BOD_t = L0(1 − 10^(−Kt)), where K = k/2.303; mixing the two forms is a common error. Downstream of an organic load, dissolved oxygen falls to a minimum and then recovers as reaeration catches up, the oxygen sag described by the Streeter–Phelps model.
Reservoirs add their own problems. In warm seasons they stratify into a warm upper epilimnion and a cold lower hypolimnion; the hypolimnion can lose its oxygen, releasing iron, manganese and phosphorus from sediments, and water released from low-level outlets may be poor. Nutrient enrichment causes eutrophication, with algal blooms, some toxic, and oxygen loss.
Management combines regulation, treatment and land management. Regulatory frameworks include the US Clean Water Act with permits for point sources and total maximum daily loads (TMDLs) for impaired waters, the EU Water Framework Directive (2000), which requires member states to reach good status in water bodies through river basin management plans, and India's Water (Prevention and Control of Pollution) Act, 1974, enforced by pollution control boards. Each sets limits, monitoring and enforcement in its own way.
On farms, best management practices reduce diffuse loads: nutrient management matching fertiliser to crop need, conservation tillage and cover crops that cut erosion, vegetated buffer strips along streams, constructed wetlands to treat drainage water, controlled drainage, and careful handling of manure and pesticides. Irrigation managers protect quality by keeping salinity in balance through leaching and drainage, and by planning the reuse of drainage and treated wastewater. The WHO 2006 guidelines for the safe use of wastewater in agriculture use a multi-barrier approach, combining treatment, crop restriction, application methods that avoid contact with the edible part, withholding periods and protective equipment, rather than relying on treatment alone. Monitoring networks with regular sampling and published results make all of this enforceable.
Learn to move between the storage zones, the mass curve and the sequent-peak table until reservoir sizing is routine, and practise the flood chain from design return period to risk to Gumbel or log-Pearson estimate to routed outflow. Draw the layout of a diversion headwork from memory and be able to say what each part does. For canal operation, be able to compute a warabandi turn, an outlet's flexibility and a conveyance efficiency in series. These are the calculations that appear as numerical questions in PG entrance and PhD papers, and the vocabulary (MDDL, FRL, exit gradient, osrabandi, IMT) appears as recall questions.
A five-year engineer in an irrigation department, a consultancy or a development project prepares seasonal canal operation plans, checks outlet settings during a tail-end complaint, computes tank yields for rehabilitation proposals, sizes lift schemes, and supports water users' associations with operation plans and fee records. In flood-prone districts the same engineer estimates design floods for embankment repairs and culverts, reads the dam's rule curve before pre-release decisions, and contributes to flood plain zoning maps. The decisions are practical: which outlet to reset, which tank to desilt first, what return period a repair should withstand.
A senior engineer with 25 years of experience leads basin plans and project appraisals: confirms dependable yield and sediment life before a reservoir is approved, sets the spillway design flood with the dam safety panel, negotiates allocations between states or users, and decides between new storage, demand management, reuse and transfer. In floods, the senior engineer owns the reservoir operation call and the trade-off between refill and flood cushion. In management, they design the legal and financial basis for irrigation management transfer and judge whether WUAs are real or only on paper.
A researcher with 25 years of experience works on the open problems: flood frequency under non-stationary climate, reservoir operation with forecasts and optimisation, sediment management for reservoirs that are losing capacity, conjunctive use models, environmental flow methods, and the institutional economics of irrigation management transfer. They design the long-term experiments and monitoring networks that test whether interventions worked, review national water assessments, and advise on policy, where the main contribution is to state clearly what the data can and cannot support.
This manual gives the planning and operating logic and the standard textbook calculations. It does not reproduce Khosla's pressure-percentage charts, the Brune and Churchill curves, Gumbel finite-sample tables or log-Pearson frequency-factor tables, so questions supply those values. It does not cover optimisation models for reservoir operation (linear and dynamic programming), hydraulic (unsteady-flow) flood routing, detailed spillway and energy dissipator design, dam safety engineering, or the water law of individual countries beyond the examples named. National figures (India's water resources, project classification, dependability norms, spillway criteria) are quoted as published by Indian agencies and differ elsewhere; check the current national document before using them in design.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.