JABATAN PENGAIRAN DAN SALIRAN (JPS) KEMENTERIAN SUMBER ASLI DAN ALAM SEKITAR (NRE) MALAYSIA NOTA KURSUS TAHUN 2006 KM 7, JALAN AMPANG 68000 AMPANG, KUALA LlJMFW URBAN STORWATER MANUAL FORMALAYSIA (MANUAL SALIRANMESRA ALAM MALAYSIA) VOLUME4 DESIGNFUNDAMENTALS DEPARTMENTOF IRRIGATION ANDDRAINAGE MALAYSIA i3.2DESIGN RAINFALL INTENSITZES Rainfallis,obviously,thedrivingf ~r c e behindail stonwaterstudiesanddesigns.Anunderstandingof rainfali processes and the significance of the rainfall design dataisa necessary pre-requisiteforpreparingsatisfactory drainage and stormwater management projects. 13.1.1Rai nfal lPat t erns in Malaysia AnoverviewoftheclimateofMalaysia,withgeneral rainfall characteristics is given in Chapter 1. The frequencyandintensity ofrainfall in Malaysia is much higherthaninmostcountries,especiallythosewith temperate climates.Rainfall designmethods,whichhave beendevelopedinothercountries,maynotalwaysbe suitableforapplicationinMalaysia.Thedesign calculations forthesemethods havebeenadjustedinthis Manual to suit Malaysian conditions. 13.1.2Appl i cati on ThisChaptersupersedesHydrologicProcedureHP1-1982 forurban stormwaterdrainage only.The Chapter does not dealwithnon-urbansituations,suchasdamsorriver engineering,forwhichtheHP1 orotherspecialhydrologic procedures should continue to apply. Thematerial in thisChapter drawsupon thatinHP1-lW2, anditspresentationhasbeenrevised tobemoredirectly applicabletourbandrainageproblems.Noadditicnal analyses wereperformed.It isenvisagedthatbothHP1 andthisChapterwillberevisedinthefuture,using additional data that is becoming available. 13.1.3 Cl i mate Change Thcreispotentialforglobalclimatechanges tooccurdue tothe"GreenhouseEffect".Afurtherdiscussionon climate changeduetotheGreenhouseEffectisgivenin Chapter 46,Lowland, Tidaland Small Island Drainage. Someauthors havesuggested that climate changedueto theGreenhouseEffectwillcauseanincreaseinstorm rainfall intensity.Atthisstage theavailable evidence for anyeffectsonrainfallintensityisnotconclusive.This Manual does not recommend any increase in design rainfall intensities due to the Greenhouse Effect. Nevertheless,designersofmajorurbanstormwater drainage systems shouldconsider thepossibility of climate changeduetogreenhouseeffect.Sensitivity testingcan beperformedincriticalcases.Designsshouldbe sufficiently robustandincorporatesafetymargins to allow for this possibility. 13.2.1Defi ni ti ons Thespecification ofarainfdl eventasa"design storrn"is commonengineeringpractice.Therelatedconceptsof frequencyandaveragerecurrenceinterval(ARI)were discussed in Chapter 11. Althoughthedesignstormmustreflectrequiredlevels of protection,thelocalclimate,andcatchment conditions,it need not be scientificallyrigorous.It ismore important to definethestormandtherangeofapplicabilitysoasto ensure safe,economical and standardised design. Twotypesofdesignstormarerecognised:synthetic and actual(historic)storms.Synthesis and generalisation of a large number of actual storms is used to derive the former. The latter are events which have occurred in thepast,and whichmay have well documented impacts onthedrainage system.However,itistheusualpraaceinurban stormwaterdrainagetousesyntheticdesignstormsand most of this Chapter concentrates on these storms. Designsror,ndurationisanimportantparameterthat defines therainfall depth orintensity for a glvenfrequency, and thereforeaffects theresulting runoff peak and volume. Currentpracticeistoselectthedesignstorrndurationas equaltoorlongerthanthetimeofconcentrationforthe catchment(orsomemmimumvaluewhenthet!meof concentrationisshort).Intense rainfalls ofsnort durations usually occurwithinlonger-durat~onstormsrather thanas isolatedevents.Thetheoreticallycorrectmct i ceisto compute drschargeforseveral design storms ~41thdifferent durations,andthenbase thedesrgnonche"cmcal"storm wh~chproducesthemaximumdischarge.'.ioXvveverthe "critical"storm durationdetermined in this wajmaynotbe themostcriticalforstoragedesign.Recommended' practiceforcatchmentscontainingstorageistocompute thedesignfloodhydrographforseveralstormswith differentdurationsequaltoorlongerthanthetimeof concentrationforthe catchment,and touse tne one which producesthemostsevereeffectonthepondslzeand discharge fordesign.Thismethodisfurtherdiscussedin Chapter 14. 13.2.2Rai nfal lIntensi ty-Durati on-Frequency(IDF) Rel ati onshi ps The total stormrainfall depth at a point,for a given rainfall durationandARI,isafunctionofthelocalclimate. Rainfalldepthscanbefurtherprocessedandconverted intorainfallintensities(intensity=depthJduration), which arethenpresentedinIDFcurves.Suchcurvesare particularlyusefulinstormwaterdrainagedesignbecause many computationalprocedures require rainfall input in the form of average rainfall intensity. U&anStormwater Management Manual13- 1 Thevambles,frequeniy,intensityandduration,are aii r-elated to each other.Tinedataarenormallypresented ascurves dispiayinr; two of thevariables,such as intensity andduration,forarange offrequencies.Thesedataare thenusedastheinputinmoststormwaterdesign processes. 13.2.3 Areal Reduction Factor It isimportanttounderstandthatIDFcurvesgivethe rainfall intensity at a point Stormspatialcharacteristicsareimportantforlarger catchments.I n general,the larger the catchmentand the shorter therainfallduration,thelessuniformly therainfall isdistributedoverthecatchment.ForanyspecifiedARI andduration,theaveragerainfalldepthoveranareais less than the point rainfall depth. The,ratioofthearealaveragerainfallwithaspecified )durationandARI tothepointrainfallwiththesame duration and ARI is termed the areal reduction factor. Arealreductionfactors areappliedtodesignpointrainfall intensities only,toaccount forthefactthat it isnotlikely thatrainfall willoccur at the sameintensityovertheentire areaofastorm(theprincipleofdesignstormsassumes that the design storm is centred over the catchment).The arealreduction is expressed asa factorlessthan1.0.For largecatchments,thedes~gn rainfalliscaiculatedwtth Equation 13.1: whereF, isthearealreductionfactor,I,istheaverage rainfalloverthecatchment,andI,isthepointrainfall intensity. Suggested values of areal reductionfactorFAforPentnsular MalaysiaaregtveninHPNo.1-1982.Thesevaluesare reproduced in Table13.1 belowforcatchmentareasofup to2GGkm2.ThevaluesareplottedinFigure 13.1. Intermediate values canbe interpolated from thisFigure. Table13.1Values of Areal Reduction Factors (FA) --- -...... 24hours; Catchment Area -6 hours 0.40--- -hour; *3 hours - - - -s-0.5hour --- 0.20 Storm Duration (hours) 101001000 Catchment Area(km2) Figure 13.1GraphicalArealReductionFactors Noareal reductionfactoris tobe used forcatchment arsas ofup to10 km2.Themajority of urban drainage areas wtll fall into this category. Arealreduction factors should not be applied to real ralnfall data,suchasrecorded dailyrainfalls.Instead anattempt shouldbemadetoobtainanduseallavailable datafrom other rain gauges in the catchment. Stcrmd~rectronandmovementcanhavemarkedeffects, pa~icularlyInareasw~t hpredommatmg weatherpatterns, andareparticularlyrelevanttothecaseofoperation ar;orconti01 ofalargesystemofstormwaterdraicage ner;$orks.However,forurban dra~nage~t IScustomary :oassume that deslgn storms are Gat~onary. 13.2.4 I DF Curves for Selected Cities and Towns Thepublication" Hydrolog~calData-RainfaNand Evd,ooratiOnRecordsforMalaysia"(199 1 ) and "HydrologicalProcedureNo.26"bytheDepartmentof IrrtgationandDrainage(DID),havemaximumramfall intensity-duration-frequencycurvesfor26and16urban areas in Peninsular Malaysia andEast Malaysia(Sabah and Sarawak),respectively.These curves willcover theneeds of the majority of users of thisManual. UsersneedtobeawareofthelimitationsoftheseIDF curves: Thecurveshavenotbeenrevisedsince1991.The patternsshould bereviewedusing theadditional data that is now available. The period of datafromwhichthe curves was derived wasveryshort,insomecasesonly7years.Fewof thestationshadmorethan20yearsofdata.This meansthatthereisalargepotentialerror~n extrapolating to long ARIsuch as100 years. ---- 13-2Urban Stormwater Management Manual -. ,.yeIc;;lier:;::;;:sfthedg;atjcnsanaj ypdwas15 mlnutes.DID shouldexpeditetheinstallationof a!rjltaipiuvlometerstocapturedatafromshortstorm C..-A- uul, u, down to5 minutes duration. ThelimitsofrainfallARIwerebetween2yearsand 100 years. Thecurveswerenotinaconvenientformforusein modern computermodels. Therewasno guidancegiven forurbanareas outside the 42 centres listed. I tisrecommended thatthecurves shouldbeupdatedby DIDtoincorporateadditionaldataandextendthe coverage as outlined above. 13.2.5IDF Curves for Other Urban Areas IDFcurvesarecalculatedfromlocalpluviometerdata. Recognising thattheprecipitationdatausedtoderivethe aboveweresubjecttosomeinterpolationandsmoothing, i tisdesirabletodevelopIDFcurvesdirectlyfromlocal rain-gaugerecordsiftheserecordsaresufficientlylong and reliable.The analyses involve the followingsteps: Data Series (identification) ii DataTests il Distrbution 1dentificaL;on C Estimation of D~stributionParameters Selection of Distrbutlon I! Quantile Estlmat~onat chosen ARI Therequired analyses arehlghly specialised andwouldke outside the scope of interest of most users of thisManual. LocalauthoritiesareadvisedtofindoutfromtheDIDcotheavailabilityofIDFcurvesorcoefficientsforthe~r respectiveareas,ortoobtainlocalpluviometerdatafor those wishing to conduct their own analysis. 13.2.6Polynomial Approximation of I DF Curves Polynomial expressionsintheformofEquation 13.2have beenfittedtothepublishedIDF curvesforthe35main citiesftowns in Malaysia. where, R ~ t =theaveragerainfallintensty(mmfhr)forARIand duration t R= average return interval (years) atodare fittlngconstants iependent on ARI. FourcoefficientsareconsideredinEquation 13.2tokeep the calculation simple fora yeasonabledegree of accuracy. Higherdegreeofpolynomialcanbeusedtogetmore accurate valuesof rainfall intensity.TheEquation canbe usedforderivingrainfallintensityvaluesforagiven duration and ARI,once the values of coefficients a to dare known.Theequationis in amoresuitable formformost spreadsheet of computer calculation procedures. Thecurvesin"HydrologicalData"(1991)arevalidfor durationsbetween 15 minutes and 72 hours.Extrapolation of the curve beyond these limitsintroduces possible errors, andisnotrecommended.Also,Equation 13.2shouldnot beusedoutsidetheselimits.Alternativeproceduresfor derivingIDFvaluesforshortdurationsaregivenin Section13.2.7. Thepossible uncertaintyrangeoftheI DF figuresderived in accordance withthisManual islikely to beup tort20/0. Amongthesourcesoferrornotedare:problemsof extrapolationtolongARIs,useoflocalratherthan generalisedanalysis,andproblemswiththeaccuracyof short-durationintensityrecords.Theerrorislikelytobe highestforthedurationsshorterthan30minutesand longerthan15hours,andforARIlongerthan50years. For particularlycritical applications it may be appropriate to conductsensitiv~ty testsfortheeffectsofdesignrainfall errors. Table132 glvesval ~es ofthefittedcoe%e.itsIn Equation13 2forKualai u ~ p u r , fcrrainfail ARI sbetween 2yearsand100yearsanddurat~ons~ l t h l n 30to1000 mlnutes(seeF~gure13.2forthegraphs)Appendrx13 A gwes derivedvaluesof the coeffic~entsInEquatlon 13.2 for the26and10urbarcentresinPeninsularandEast Malaysla,respectivelyDuetoirregularshapeofthe curves,coeffiaentsfor6otherurbancentresInEast Malaysia arenot sutable tobeused InEquatlon 13 2.IDF valuesforthese6stat~onsshouldbetakenfromtheir respectwe curves ava~lableIn HP-26 (1983) Table13.2Coefficients of theFitted IDF Equation for KualaLumpur (data period 1953 - 1983);\/a!!dity:3C c t 51000 minutes Urban Stmwater Management Manual 13-3 2Yr. . - - . - 1 yrARI10C Duration (minutes) Figure13.2IDF Cuwes forKuala Lurnwr 13.2.7 IDF Values for Short Duration Storms I t18 recommendedthatEquat~on13 2beusedt oderive designralnfallintensitiesfordurationsdownt oalower limit of 30minutes.Thisvaluecorrespondst o the original range of durations used in deriv~ngthe curves. Estimationofrainfailintensit~esfordurationsbetween5 and30minutesinvolves extrapolationbeyondthe range of the dataused In deriving the curve f i t t ~ngcoefficients.The eec ommended methodofextendingthedataisbasedon HPNo.1-1982,whichgivesarainfalldepth-duraticn plottinggraphf xdurationsbetween15minutesand3 - hours.Th~sgraphicalprocedurewasconvertedintoan equationandextendedasdescribedbelow.Anadditional _ adjustmentforstormintensitywasincludedbasedonthe methodusedin"PNGFloodEstimationManual"(SMEC, 1990),fortropicalclimatessimilartoMalaysia.This adjustmentuses the2 year,24-hourrainfall depth'p24h as a parameter. ThedesignrainfalldepthP,forashortdurationd (mmutes)is given by, wherePjOrP60 arethe30-minuteand60-minuteduration rainfalldepthsrespectively,obtainedfromthepublished designcurves.Foistheadjustmentfactorforstorm duration Equatlon13 3shouldbeusedfordurationslessthan30 minL:?sFordurat~onsbetween15and30mlnbtes,che resu:sshouldbecheckedagainstthepubhshedIDF c u n s Therelatiorship1s validforanvARI wthin:he ranceof 2to100 years The,:aiueofFD isobtainedfromTable13.3asafunction of 'i,,,, the2-yearAX124-hourrainfalldepthValuesof 'P.,.forPen~nsular MaiaysiaaregiveninFigwe13.3 I nt s- ~edi at evaluesshouldbe ~nterpoiated. NotethatEquatior!13.3!s!n termsofr3!n.f3!! depth,not intensit\/.I f intensityisrequired,suchasforroof dra~nage,thedepthP, (mm)ISconverted toanintensiP{i (rnm,'hr)by div~dingby the durationdi n hours: Table13.3Values of FDforEquation 13.3 II j Duration/2P2qh(mm) I I - 13-4Urban Stormwater Management Manual East Coast All1 1.39 1.03 0.741 0.48i 0.001 I West Coast :, 180 1.40 0.86 0.54 (minutes)11 100 512.08 120 1.85 1.13 0.72 10 15 150 1.62 0.99 0.62 1.28 0.80 - 20 30 0.42 O.O@ 0.47 0.00- 0.3610.32 0.00/0.00 13.2.8I DF Values for Frequent Storms Waterqualitystudies,inparticular,requiredataonIDF valuesforrelatively small,frequentstorms.These 9orms areofinterestbecause onanannualbasis,upto 90% of thetotalpollutantloadiscarriedinstormsofupto 3monthARI.Chapter 4recommendsthatthewater qualitydesign stormbethatwitha3monthARI.The typicalIDFcurvesgiveninAppendix13.Ahavealower limit of 2years ARI and thereforecannot be used directly. Thefollowingpreliminaryequationsarerecommendedf ~ rcalculatingthe1,3,6-monthand1 yearAR!rainfail intensities in the design storm,for all durations: . . ~. ,;Jhere," " r - ,'-'zI-,, "i,and' I , aretherequired1,3,6- nionthand!-yearARIrainfall~ntensitiesforanyduraticn D,and'1, :s th22-year ARIrainfallIntensityforthes a wduratior!D,obta~nedfromIDF curves. Users shouldbe aware of the limitations of these Equations 13.5a to13.5d.They werederivedbyfittinga distributlcn tothe1-hourduratjonrainfalls,andextrapolatingthe distributiontofrequentARIs.ThismethodISsubjectto considerableuncertainty.Thesepreliminaryequations werederivedusingIpoh rainfall data.Furtherresearch 18 required toconfirmtherelationships,particularlyinother parts of Malaysia where different climatic influences apply. 13.2.9IDF Values for Rare Storms Further research is required in order to allow design rainfall mformation tobegivenforstormswithARIgreaterthan 100 years. This Manual doesnot cover the design ofmajorstructures suchasdamsorbridges,forwhichaspecialhydrologic analysis is required. 13.3DESIGN RAINFALL TEMPORAL PATTERNS i 3. 3. 1Purpose Thetemporaldistribut~on ofrainfallwithinthedesign storm is an important factor that affeds the runoff volume, andthemagnitudeandtimingofthepeakdischarge. Design rainfall temporalpatterns are used torepresent the typical variationof rainfallintensities during a typical storm burst.Standardisationoftemporalpatternsallows standarddesignprocedurestobeadoptedinflow calculation. I tisimportanttoemphasise thatthesetemporalpatterns are intendedforuse indesign storms.Theyshould no:be confusedwiththerealrainfallvariabilityinhistorical storms. Realisticestimatesoftemporaldistributionsarebest obtainedbyanalysisof!ocalramfalldataFromrecarding gaugenetworks.Suchananalysismayhavetobedone forseveral widelyvaryingstormdurations tocover various typesofstormsandtoproducedistributionsforvarious designproblems.Differentdistributionsmayapplyto different climatic regions of the country. Temporalpatternsshouidbechosensothattheresulting rnnzffhydrqi aphsa x consistentwithobserved hydrographs.Thereforetheformofthetemporalpattern andthemethodofrunoffcomputationarecloselyinter- linked.The statistical basis of this approach 18 discussed in ".4usistr2/iar?.!?a~fA'.3,7C.?;.ncff' (P.R&R,1987). Arangeofmethodstodistributerainfallhavebeen suggested ~ntheliteraturs: 1Aderagetemporalpacerns developed fromlocal mi nt- rainfalldatameasuredInshorttrmeintervals(15 mlnutes crless) 2Si m~l e ~deallsed rairfalld~strlbutron fittedtolocal storm databy themethod of moments. 3Temporal patterns fromlocal IDF relat~onsh~ps. ThesecondmethodISnotrecommended,astheidealised patternsarenotrepresentativeofrealstormpatterns. Triangularpatterns,forexample,giveunrealistically high peak intensities. The thirdapproachfordistributingrainfallwithinadesign stormmakesuseofthelocalIDFrelationshipforthe designARI.Thisapproachisbasedontheassumption thatthemaximumrainfallforanydurationlessthanor equaltothetotalstormdurationshouldhavethesame ARI.Forexample,a10 year ARIthree-hourdesignstorm oft h~stypewouldcontainthe10 year ARIrainfalldepths foralldurationsfromtheshortest timeinterval considered (perhaps 5minutes)up to three hours.Theserainfalls are generally skewed. Urban Stormwater Management Manuai Deslgn Rainfall Figure 13.3Values of ' P ~ for use with Table 13.3 (source:HP 1, 1982) 13-6Urban Stomwater Management Manual 5 4Ithcijgh there are some theoretical sbjections to thethird approach,onthegroundsthatitcombinespeaksfrom diffsenthistoricalstorms,it 1s nevertheless conservative andveryconvenientfordesign.itistherefore recommended. This distribution canbe derivedfromthelocal IDF curves and the analysis of skewness of actual storms.The design temporalpatternspresentedinthisChapterhavebeen derived on this basis. 13.3.2Present Malaysian Practice. The1982updateofHydrologicalProcedureNo.1 gave recommendationson temporalpatterns tobeadopted for designstormsinPeninsularMalaysia.Patternswere prepared forsixstandard durations:0.5,3,6,12,24and 72 hours. Nine rainfall stations located in different parts of Peninsular Malaysia were used in this analysis.The data covered nine years from July1970 to June1979. Theprocedure usedwastoextract therainfallpatternof theannualmaximum stormbursts at eachstation.Thisis Method1 oftheprevious sub-section.Bycomparing the results,asetofrepresentative patterns werederived.I twasfoundthatdifferentpatternsappliedtotheWest CoastandEastCoast,exceptforthe0.5hourduration, andthereforedifferentregionalpatternswere recommended. Therecommended patterns werepresented in HPNo. 1 in graphical form(FiguresD.8to0.13ofHPNo.1).These graphsaredifficulttoreadandaresubjecttomis- interpretation.Therefore,the datahasbeen converted to tables in this Manual,asshown in Appendix13.8. 13.3.3Review of Standard Temporal Patterns ThisprocedureusedinHPNo.1-1982isless comprehensivethanthatusedin,forexample,AR&R (1987),because there wasa much smaller amount of data available. 18yearshaveelapsedsinceHydrologicalProcedure No.1 wasupdated.Thepatterns shouldbe reviewed using the additional data that is now available. TherangeofStandardDurationsusedinthe Procedureisinsufficient forthefullrangeofdesign conditions.Design calculationsneed tobemade for periodsasshortas5 minutes,inthecaseofRoof Drainage(seeChapter 23).Thegapbetweeneach Standard Duration in HP No.1-1982is too wide. Temporalpatternsshouldbeinternallyconsistent (AR&R,1987).There are inconsistencies between the HPNo.1-1982 patterns. wDurations longer than 6hours are not coveredin this . Manualastheyareunlikely toberequired inurban stormwaterdrainagedesign.I f ararniaiiiuraticr! longerthan6hoursisrequired,theasershould consult HPNo.1-1982. Notemporal pattern datais available in HPNo.26for SabahandSarawak(1983).Forpreliminarystudies, thepatternsfortheEastCoastofMalaysia,in Appendixl3.B,couldbeadoptedforSabahand Sarawak.Becausetheclimaticconditionsaremore comparable to the East Coast than the WestCoast.A furtherstudytoderivetemporal patternssuitable for useinSabahandSarawakshouldbeundertakenby specialist hydrologists. There are too few datapoints found in each Standard Duration temporalpattern.Thiscausesasystematic biaswhichwillunder-estimatethepeakrainfallas shown in Figure 13.4.AR&R-1987estimates that this errorwillunderestimate the truepeakbyasmuch as 10%. Instantaneous Peak / Intensity Indicated Peak c in c s c I -- 2 c .- 2 Time Figure 13.4Example of the Under-estimation of a Hydrograph by Discretisation 13.3.4Temporal Patterns for Standard Durations Therecommended patterns inthisManualarebasedon thosefromAR&Rfordurationsofonehourorlessand from HPNo.1 (1982) for longer durations. Therecommended patterns followMethod 3 mentioned in Section 13.3.1.Checksweremadetoensurethatthe patterns complywiththerequirements inSection 13.3.3. Wherenecessary,theordinatesofthepatternswere adjustedtomeettheserequirements.Mostofthe adjustments made were relatively minor. Thestandarddurationsrecommended in thisManualfor urbanstormwaterstudiesarelistedinTable 13.4.The interim temporalpatterns tobeusedforthesestandard durations are given in Appendix13.8. Urban Stomwater Managem&Manual 13-7 Design Rainfall Table 13.4Standard Durations for Urban Stormwater Drainage Iote thatminutesareusedinthisTable,forconsistency lith the units in Equation 13.2. Standard Duration (minutes) 10 .3.3.5Temporal Patterns for Other Durations ;orotherdurations,thetemporalpatternforthenearest standard durationshouldbe adopted.It isNOTcorrectto average the temporal patterns for different durations. Number of Time Intervals 2 13.4RAINFALL TIME SERIES Time Interval i (minutes) 5 13.4.1Introduction Calculationsforstormwaterqualitymayinvolvethetime seriesofrunoff,whichInturnisrelatedtothevolumeof .airifall. jailyrainfallgaugesarewidespreadthraughoutMalaysia, Icomparisontothesmallernumberofpluviometers. ailyrainfallrecordsarealsooflongerduratior?than luviometer data.Both of these attributes make daily data aluableforstatisticalstudies.Dailyrainfalldatais ormallyreadilyavailableatorclosetoanylocationof rterest forurban stormwater studies. 13.4.2Sources of Rainfall Data I nMalaysia,rainfalldataiscollectedbyseveral departmentsandauthoritiesincludingtheMeteorological ServiceandDID.Otheragenciessuchaswaterand sewerageagencies,maycollectdatafortheirown purposes. The quality of the designs and analyses depends to a large degree,on the qualityof therainfall dataused.Therefore everyeffortshouldbe made t o searchforand obtaindata from the data collection agencies. - 23.4.3 Data Quality and Acceptance -L8l i e-dailyiainfa!lrecordshouldbeexaminedforquality, 3?i!inparticulartoidentifyanyinstancesofmissing records. Occasionalmissingdailyrecordsmaybeacceptable, depending onthepurpose forwhichthedataisused.I n general,forwaterqualitystudiestheamountofmissing record shouldnotexceed3%i.e.10daysperyear.For floodstudies,particular care is required because i t is often found that themissing record is the actualrecord ofmost interest,i.e.a large storm or flood event. SomeolderdailyrainfallrecordsarestillinEnglish (Imperial)units. 13.4.4Long-Duration Rainfalls Some of the design guidelines given in thisManual require calculation oflong-durationrainfalltotals,suchasthe75 percentile5-daytotal.Dailyrecordsshouldbeusedfor this purpose. Example 13.C.2inAppendix13.Cisaworkedexampleof the calculation of5-dayrainfall totalsforIpoh,Perak.The 75percentile 5-day rainfall is used in the designprocedure forsediment basins,in Chapter 39. 13.4.5Adjustmentof Daily Rainfalls Daily rainfall dataisread withfixedobservation times each day.Dailyrecordsdonotnecessarilyreflecttherainfall figuresfora24hourstormdurationbecausethestorm periodisunlikely tocommenceatexactlythetimeofthe daily reading. I fconversionisnecessaryfromdailyrainfalltotalsto24- haur totals,it should be done using Equat~on13.6: where,I . is the24-hour storm rainfall fora given ARI,IDis thedailyrainfall,andCi s aconversion factorwithvalues of1.16 (East Coast) and1.12 (West Coast). Thisequationisunchangedfromthe1982editionof HP No.1.TheconversionfactorCisobtainedfrom analyses of rainfall records.The values ofCrecommended aboveareconsistent withthose in SMEC(1990)andMiller et al (1973). 13.4.6Continuous Simulation Continuoussimulationmodelsareimportantforwater quality studies.Most of these models are designed foruse . with daily rainfall data. - 13-8Urban Stormwater Management Manual s~mulatrony n o d sf severalmonthstoach~evestabiiib.standards for water quality control works,asdixussec..Usually!t 6conuenienciornr!themodelsfcra12-mcnthChapter 4.The result is likely to be similar rnother part; simulatioii.Thisensuresthatseasonalvariations,ifany,of the West Coast of the Peninsula.Results in sttier a mare taken into account.of the countrymay be different due to different rainfall characteristics.Therefore,a 6milar calculation should be Thedatarequirementforsuchasimulationisafullyear,performed at other locations. ormore,ofcontinuousdailyrainfallrecords.Evendata withasmallnumberofmissingrecordscanbeaccepted (seeSection 13.4.3).Thedatashouldbepreparedin the formof a list orspreadsheet column of dailyrainfall totals, inrnillirnetres.If,possible,datashouldbeobtainedin digitalformtominimisedataentryand 'thepossibilityof transcription errors. 13.4.7Role of Small,Frequent Storms Like most decisions ondesign standards, the selection of a suitabledesignstandardforwaterqualitycontrolworks involvesperformanceandeconomicconsiderations.I trequiresatrade-offbetweenthebenefitsofprovidinga higherlevelofprotection,performanceandthesizeand cost of works needed to provide that protection. Thelandareaandcostofatreatmentfacilitysuchasa pondisapproximatelyinproportionwithitsvolume. Increasingthedesignstandardfrom(say)3monthto 6 month ARIwillresultin significantsize/costincrease for onlyaminimalincreaseinpollutantremovalandthiscan beexpressedintermsofthelawofdiminishingreturns thatdefinesthepointatwhichincreasingthedesign standardisnolongercost-effective.It isgenerallyfound thatforwater qualitycontrol,mostbenefitisderivedfrom thetreatmentofsmallfrequentstormsandthati tisnot cost-effective to design facilities forlarge,rarer events. Theprelimmarycalculationspresentedlater(Chapter15) using rainfall dataforIpoh give a guidelinetodetermme a suitable cho~ceofdesign stormforsizingthewaterquai!?] treatmentmeasures.Theresultsofthecalculationsare plottedinFigure 13.5.FromFigure 13.5itcanbeseen that,for Ipoh (Perak): On a long-term average basis, 93% of total rainfall occurs in storms equal to or smai'ler than 3 month ARI Figure13.5AmountofTotalRainfallwhichOccursin Storms Less than a Given ARI (IpohData) looO/o -- 2 .E90% 2 - m 80% L 0 7o0/o 9 i jm - $60% u 50% 13.5HISTORICAL STORMS , I' 'I I +-- i 1 I . Adiscussion of rainfall data wouldnot be complete without mentionof historical storms.Historical stormdatais used in calibration ofmodels, aswell asforthe checking ofpast flood occurrences. 0.010.1I10 Event ARI (years) I n anurban drainage situation,i t isrelatively rare tohave goodhistoricalrainfalldataavailableclosetothestudy areaorcatchment.Nevertheless,everyeffortshouldbe made toobtainsuchdata.Datasourcesarediscussedin Section13.4.2. Goyenand O'Loughlin (1999)haveshowntheimportance of locating rainfall gauges closeto orpreferably within the studyareai faccuratecalibrationistobeachieved.I fdetailed studies arebeing undertaken and goodcalibration dataisrequired,thedensityofraingaugesshouldbeat least 1 per square kilometre. Urban Stomwater Management Manual 13-9 Design Rainfall APPENDIX13. A F I n ED COEFFICIENTS FOR IDFCURVES FOR 35 URBAN CENTRES Pulau PtnangPenang PerakIpoh PerakBagan Serai Pera kTeluk Intan Pera kKuala Kangsar 1205.785410.11751-0.1244:0.0044; ;506.5736 '-0.2903'-0.0482:0.00002' I -----. I j1 PerakISetiawan Selangor I '--. - - -i (Continued) U&n Stormwater Management Manual13-11 - IaGiei 3 . Ai Coefic~encsfor the i DF Equations for the Different Major Cities and Towns ~r ! Maiaysla (32 5r 51000 m~r.:! 1 25.25650.0719-0.13060.0065 5j 5.4663'0.0586-0.12690.0062 Neger~Semb~lanSeremban1970-1990106.12401-0.2191-0.08200.0039 206.3733,-0.2451-0.08880.0051 Coefficients of:he:CF?OI L~CCat E ~ ~ a t ~ a n sState~ocat i onDataPer~od-- 1 abt --- d -- 2/5.32550.1806,-0.13220.0047 I I55.10860.5037,-0.21550.0112 Federal TerrtoryKuala Lumpur1953-1983104.9696!0.6796;-0.2584;0.0147 i Negeri Sembilan 20 I SO ;100 '2 5 MaiaccaMalacca119;l-1990:10 i KualaPilah Johor 4.9781:0.7533-0.27960.0166 4.80470.9399-0.3218j0.0197 -- Johor i 5.0064 3.7091 4.3987 4.9930 -- Johor 20, 0.87091-0.3070,0.0186 1.1622/-0.32890.0176 0.7725/-0.2381j0.0112, 0.46611-0.1740/0.0069, Kluang Mersing 5.0856 Batu Pahat 0.5048'-0.1875,0.0082. 5014.8506 --10015.3796 Johor 0.7398,-0.23880.0117 0.4628-0.18260.0081 JohorBahru Johor I Segamat11970-1983 i I -.-- - ---- -- continued) -.- i 13-12Urban Stormwater Management Manudl TerengganuKuala Dungun1971-1983 A NCoeffk~eiitrof the i GF Polynom~alEquations Stareiocat~onData PemdI (year),a0, I C I d ! i II214.37160.3725/-0.127410.0026 I I I \54.54610.4017i-0.1348j 0.0036, 1Pahang, I RaubI1966-19831 10I5.4226i-0.1521/-0.0063!-0.00561 II2015.2525'0.0125I-0.0371-0.00351 TerengganuKuala Terengganu1951-1983 -- - + - - - KelantanKotaBharu1951-1990 I II I50 i /KelantanIGuaMusang I I II 4.86540.3420-0.10580.0012' 5.1818'0.2173-0.08340.0001 (Continued) 100 -2 5 4.9396 I IPahangCameron Highland11951-199010/4.3258!0.7684 Urban Stormwater Management Manual13-13 0.2645-0.16380.0082 I I 4.6471,0.4968 I / -0.25490.0134' I i2014.81780.5093i-0.2022j0,0100, 1I 505.32340.2213-0.1402i0.0059, I II1005.01660.46751-0.1887I0.0089 125.18990.2562 II 54.75660.6589 -0.1612j 0.0096 -0.2529!0.0167 PahangKuantan1951-1990i104.37540.9634I-0.30680.0198 III204.85170.7649-0.26970. 0176 ,505.03500.7267,-0.25890.0167 1005.2158,0.6752I-0.2450i0.0155 24.60230.4622-0.17290.0066 I 155.30440.0115-0.0590-0.0019 PahangTemerloh1970-1983110-14.58810.5465-0.1646'0.0049 204.43780.71 18-0.19600.0068 504.48230.8403-0.22880.0095 _- I -- 100'4.52610.7210-0.19880.0071 - ! able13.P.TCoefficients for the IDF Equations for the Different Major Cities and -OWESin Maiaysii!(30 i t 51000 min) - .- AP.! Crzefficients of :he:DFPolynom~alEquations StateLocat~onData Penod -j I (year)/abc, d13-14Urban Stormwater Management Manual 55.57220.0563-0.09190.0031 --- ------ 1953-1980 - SarawakB~ntulu106.1060-0.2520-0.02535.0012 --- - - -. 206.0081-2.Ll73-0.0574IO.OOX_- 6.2652- 0. 2584- -0.0244'-0.0008, 50---______- --- 23.22351.2714-0.3268,0.0164 - 5j4.54160.2745-0.0700-0.0032- SarawakKapt1964-1874104.51840.2886-0.0600,-0.0045 ------.----- 20,5.0785-0.0820,0.02961-0.0110- 2 -- 5.1719-,0.1558_-0.1093:0.0043 __ 54.88250.3871-0.145510.0068 SarawakKuch~ng1951-198010I5.16350.2268.-0.1039/0.0039 I20i5.2479-0.09680.0035 I ---- i5015.2780:0.2240 '2 I5 Sabah/Kota Kmabalu11957-1980l oI 1 II I20 5.19680.0414-0.07121-0.0002, 5.6093,-0.1034j -0.03591-0.0027 5.9468i-0.25951-0.0012-0.00501 '0.0038 -0.0021' -0.0048 -0.00444 I Ij2 I I1 550 2 5 10 20 50 I 4.93020.25641-0.1240 5.8216/-0.21521-0.0276 0.0026 0.0027 0.0191, 0.0095 0.0083I 0.0133 0.01231 0.0093 II 5.2150j 0.3033I-0.1164 Sabah I5016.3582 2i4.1091i0.6758i-0.2122 5.1922 3.7427 -0.3823/0.01701-0.0054 SarawakMI ~I I1953-1980106.1841'-0.385610.0114 I I SabahITawau1966-1978 ! I I 0.3652/-0.1224 1.2253/-0.3396 I 1 1 Sandakan 4.924610.51511-0.1886 5.272810.3693/-0.1624 4.9397!0.66751-0.2292 5.00220.6587'-0.2195 1957-1980 -0.3188j 0.0021i I II 513.10661.7041j -0.4717/0.0298- 2016.1591 10 20 b I ! I i24.1878'0.9320-0.3115/0.0183, i53.75221.3976-0.4086/0.0249 Sa ba hKuamutI1969-198010!4.15941.2539-0.3837j 0.0236 ;2013.8422,1.5659-0.4505/0.0282 ,505.62740.3053-0.1644i0.0079- /100I6.3202-0.0778-0.0849!0.0026 I j 2!4.3333 0.7773-0.2644/0.0144 5i4.98340.4624'-0.198510.0100 4.14191.1244I-0.3517/0.0220 4.46391.0439-0.342710.0220 SarawakS~manggang,1963-198010!5.6753:0.0623-0.1097j0.0038 --I2015.9006-0.0189'-0.0922i0.0027 1 23.08791.6430-0.4472;0.0262 15I3.45191.4161-0.3754,0.0200_ SarawakS~buI1962-1980103.64231.3388-0.3509'0.0177 - - - -20,3.3170,1.5906-0.3955,0.0202_- 25.27070.1314-0.09760.0025 -- - - ----- - -- -- Design Ra~nfall - APPENDIXi 3. BDESIGN TEMPORAL PATTERNS Tabie 13.B1Temporal Patterns - West Coast of Peninsular Malaysia I 12 Time Period Duration (min) 10 15 30 60 120 180 360 15 min Duration 123 Time Period No. of Time Periods 2 3 6 12 8 6 6 60 minute Duration Fraction of Rainfall in Each Time Period 0.5700.430- 0.3200.5000.180- 0.1600.2500.3300.0900.1100.060- 0.0390.070q.1680.1200.2320.1010.0890.0570.0480.0310.0280.017 0.0300.1190.3100.2080.0900.1190.0940.030- 0.0600.2200.3400.2200.1200.040- 0.3200.4100.1100.0800.0500.030- 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2Time Period 180 minute Duration 123456 Time Period I30 minute Duration I 1 2 3 4 5 6Time Period 120 minute Duration 10.5- I 1 2 3 4 5 6 7 3Time Period I 360 minute Duration 123456 Time Period U&n Stormwater Management Manual13-15 Design Rainfall 1 -(min)Periods Table13.82TemporalPatterns - E a nCoast nf seninsularMaiaysia" Fraction of RamfalljnEach Time Period 10 min DurationI Time Period 3 15 min DurationI 1 2 3 Time Period I 1 23 45 67 89 1 0 1 1 1 2Time Period I I 180 mlnute Duration1 12345 Time Period I 30 minute DurationI 1 2 3 4 5 6Time Period I 120 minute Duratm I 1 2 3 4 5 6 7 8Time Period 360 mlnute Duratlon 0 5 - -. ---. 12 3 45 Time Period ( #these patterns can also be used in Sabah and Sarawak,until local studies are carried out) - - - - - - - 13-16Urban Stormwater Management i%nua/ Design Rainfall APPENDIX13.CWORKEDEXAMPLE 13.C.1Calculation of 5 minute Duration Rainfalls Table:3.C:Exampie of Caiculation ofDaily and 5-day Rainfall Totals Rainfall Station 45:111, Politeknik Ungku Ornar at Ipoh,Perak Problem:Calculatethe5minuteduration,20yearARI rainfall for use in a roof design in Kuala Lumpur. L D a t e 1994*1996 DailyISDay ~ o t a l lDaily(5Day Total Solution:Fiveminutesdurationisshorterthanthe periodofvalidityofEquation 13.2.Therefore,referto the procedure for other durations i n Section 13.2.7. From Equation 13.2,for 20 year AMand't = 30 minutes weobtain20~30=142.4mm/hrandthecorresponding rainfall depth is 20~'0 = 71.2mm. Similarly,Z0~60=91.0mml hrandthecorresponding rainfall depth is= 91.3mm. From Figure 5 in HPl-1982,P' forKuala Lumpur is read asapproximately100 mm.ThecorrespondingFD factor from Table 13.3 is 2.08. Substituting these values inEquation 13.3, j (Intermedizte Data from 1511 to 19/f 2 are not Shown)I Canvertthisdepthtoarainfallintensityusing Equation 13.4: 13.C.2Use of Daily Rainfall Data Problem:Use daiiyrainfall records to calculate the5-day rainfalltotalsforIpoh,Perak,Computethe25 percentile,5G percentile and75percentiie5-dayrainfali totals.The75percentile5-daytotalisrequiredfor designofawetsedimentbasin,inaccordancewith Chapter 39. Solution:TheresultsarepresentedinTable13.C1. Someoftheintermediatelinesof datahavebeen omitted.Calculationsarecarriedoutusingstatistical functions in a spreadsheet such as EXCEL. ?-.iqdicates Missing Dara i20/12I0.0I0.00.012.0 j21/12I0.5j0.5-9.5 22/120.50.0 21.5 21.5 30.5 I0.5I0.0 20.0 25/12/7.518.010.0I20.0- i26/12j7.0j 14.5j0.511.0 i27/120.0014.5i2.5 1- 13.5 '3/ 12 ' 0.0114.5!0.0!3.0! ,rp- 29/12~ - 7 0 . 0 i3.0! !301120.0 --7- 7.0143.5/46.5 I31/120.011.5!47.5 L-__i-._.- 1Total1783.0 I!2351.01 _X____-_- IMissing days!151 91 j /25percentile;- perce~ti l ei HYDROLOGICAL PROCEDURE NO. 4 MAGNITUDE AND FREQUENCY OF FLOODS IN PENINSULAR MALAYSIA (REVISED ANDUPDATED) JABATAN PENGAIRAN DAN SALIRAN KEMENTERIAN PERTANIAN MALAYSIA HydrologicalProcedureNo. 4 MAGNITUDE OF AND FREQUENCY FLOODS IN PENINSULAR MALAYSIA (REVISED AND UPDATED) 1987 BAHAGlAN PARIT DAN TALIAIR KEMENTERIAN PERTANIAN, MALAYSIA HydrologicalProcedureNo. 4 MGGIMTUDE AND FREQUENCY OF FLOODS IN PENINSULAR MALAYSIA (REVISED AND UPDATED) 1987 Contributor: OngCheeYan Price:$lo/= DrainageandIrrigationDepartment Ministry ofAgriculture Malaysia. First published in1 974 Revised and updated in1987 SYNOPSIS ThisprocedureisarevisedandupdatedversionoftheDrainageandIrrigation DepartmentHydrological ProcedureNo.4(1974) - "MagnitudeandFrequencyof Floods in Peninsular Malaysia".The Hydrological ProcedureNo.4 whlchwas firstpublishedin1974 wasdevelopedbasedonhydrologicaldataup t o year 1970 and regionalanalysis was usedt o estimate designfloods for PeninsularMalaysia. Thisrevisedandupdatedversionalsoestimatesfloodsusingthetechniqueofregionalana'lysis.However, anadditional10yearsormorehydrologicaldata( up toyear1982) wasusedt o establishthenewflood frequency regionsfor PeninsularMalaysia. Theregionalanalysiscarriedoutinthisproceduregenerallyconsistsofthe t wo major parts - (i)Develop- mentofasetofregionaldimensionlessfloodfrequencycurves and(ii)Developmentofasetofregional regression equations relatingmeanannualflood t o the catchmentcharacteristics(catchmentareaandmean annual catchmentrainfall). Hencetwomapsareincludedinthis~r ocedur e- Map1identifiesthe variousfloodfrequencyregions ( FFregions)inPeninsularMalaysiaandMap1 identifiedthevariousmeanannualflood(MAFregions) inPeninsularMalaysia.Byknowingthefloodfrequencyregionandmeanannualfloodregion a riverbasin ofinterestbelongsto, thedesignfloods of thebasincanbeestimatedusingtheregionalfloodfrequency curveandtheregionalMAF equation. Thisprocedurewillberevisedandupdatedagainwhenanadditionaltenyearsofhydrologicaldatais available. iii CONTENTS Page INTRODUCTION 1.1Regional FloodFrequency Analysis 1.2Frequency Distribution Used- Gwnbel Type I DEVELOPMENT OF PROCEDURE 2.1Selection of Catchments 2.2ExtractionofAnnual Flood Data 2.3Frequency Analysis of Individual Station FloodData 2.4Derivation of Regional Dimensionless FloodFrequencyCurves 2.5Derivation of Regional MAF (Mean Annual Flood) Equations APPLICATIONOF PROCEDURE 3.1Method of Application 3.2Worked Examples ACCURACY OF PROCEDURE 4.1Comparison withObserved Data 4.2Comparison withW 4 (1974) 4.2.1ComparisonofResults usingHPq 1988) andHP4(1974) RELIABILITY OF PROCEDURE REFERENCES APPENDIXI- List of Catchments and Catchment Characteristics used in Procedure APPENDIXII - Results of Individual Station Frequency Analysis MAP1- Flood FrequencyRegions of Peninsular Malaysia MAP 2.- Mean Annual FloodRegions of Peninsular Malaysia Intheplanninganddesignofwaterresourcesprojects,engineers andplannersareofteninterestedto determinethemagnitudeandfrequencyoffloods thatwilloccurattheprojectareas. Anestimateofthe magnitudeofaaoodofacertainrecurrenceinterval(commonlyknownasthe"designflood")thatis likelytooccurisfundamentdtoensurethateconomicengineering designwithadequatestandardsof safety can beachieved. InMalaysia, streamflowrecords from gauged rivers offer a fairly accurate means of estimatig design floods throughtheapplicationofvarious statisticalmethods.However, noteveryriverinMalaysiais gauged and moreover,gaugedriversareonlygaugedatcertainstrategic pointsoftherivers.Ifaprojectislocatedin anungaugedcatchmentwithnostreamflowrecords,thenthe designfloodsforthecatchmentwillhave to be estimatedby other flood estimationtechniques. Hydrologists havedevelopednumeroustechniquesforestimatingdesignfloods, themainones beingthe rationalmethod,unithydrographmethod,conceptualandstatisticalrainfall-runoffmodels andregional frequencyanalysis.Thisproceduredescribestheuseofregionalfloodfrequencyanalysestoestimate designfloods for Peninsular Malaysia. 1 . 1 Regional Flood Frequency Analysis Theregionalapproachtofloodfrequencyanalysis hasbeenwidelyusedinmanycountries,inUnited Kingdom (NERC,1975), inUnitedStates (Riggs,1973) and in NewZealand (Beable etai,1982). Regiona- lizationorregional analysis is concerned with the extension ofrecords from gauged sites ofclose proximity to coverthatofaregion:itprovidesa meansofapplyinginformationfromgaugedsites inoneregionto ungaugedsitesinthesameregion.Inregionalfrequencyanalysis, individual frequency curves from gauged sitesareaveragedt oformaregionalcurvewhichispostulatedto applytoall catchments intheregion. Inthisstudy, theregionalfloodfrequencyanalysis methodusedbytheNaturalEnvironmentalResearch Council(NERC,1975)isadopted.Basically,themethodinvolves thedevelopmentoftwocomponents: (i)Asetofdimensionless regionalfrequencycurvesrelating QT/MAFtoTwhereQTisthepeakdis- chargeofT-years recurrenceinterval, MAFisthemeanannualfloodorpeakdischargeandT is the recurrence interval inyears. (ii)Asetofregionalregressionequationsrelatingthemeanannualpeakdischargetothecatchment characteristics of catchmentarea and mean annual catchmentrainfall. 1.2Frequency Distribution Used- Gumbel Type 1 Aprobabilitydistribution commonlyusedinfloodfrequency analysisistheGumbelType1 distribution. This distributionhasbeenadoptedfor regional floodfrequency analysis inUntedKingdom (NERC,1975). NewZealand(Beableetal,1982) andothercountriesthroughouttheworld.TheGumbelTypeIdis- tribution was adoptedfor flood frequency analysis inthis procedure. Thereturnperiodorrecurrenceintervalofanyrankedfloodina Gumble Type1 distributionisskewed towardsthemodeofthetheoreticaldistribution.Thetheoreticalfitisthendeterminedbythe methodof moments andis inthe form ofthe following equation (Haan, 1977): whereXTis the magnitude ofthe event having a returnperiod of T years. - Xis the arithmetic mean value of the magnitudes of the events. v- is the standard deviation from the mean. Kis the Chow Frequency Factorfor the Extreme Value Type I distribution. Thedata werealso fittedusingLog-Pearson 111probabilitydistribution.This distributionis recommended bytheUnitedStatesWaterResource Council (USWRC, 1967) forfloodfrequencystudies intheU.S.A. However, the Log-Pearson I11 distribution was notadoptedfor this procedure because: (i)thecurvatureofthisdistributionvaries greatly fromone dataset -to anothermaking itdifficultto obtain a regional curve. (ii)there is a high degree ofskewness inthe sample distributions. (iii)20% of the sample data failed the Smirnov-Kolmogorovgoodness-of-fit test. 2.DEVELOPMENT OF THE PROCEDURE The methodusedin developing this procedure is summarized below: (a)Selectionof catchments. (b) Extraction of annual floodflow (annualpeakdischarge) data. (c) Frequency analysis of individual station flow data. (d) Derivation of regional dimensionlessflood frequency curves. (e) Developmentof regional meanannual peakdischarge (MAF) regressionequations. 2 1. Selection of Catchments Streamflowrecordsfromal3riverstationsoperatedbytheDrainageandIrrigationDepartment were inves- tigatedforthestudy.Atotalof61stations withlengthsof recordsvaryfrom8 t o 36 years in Peninsular Malaysia were selected for the analysis.The stations selectedare listedinAppendix I.The flow records from each station wereassessedandselected basedon the following criteria: (i)There must be at least eightyears of goodquality data. (ii)Thecatchmenthasnotchangedsignificantly'overtheperiodofrecordsbecauseofurbanisation, agricultural developmentor industrialdevelopment. (iii)Thereisno substantialregulationof flowupstresmof thestationdue t o r e s e ~ o r storage or diversion offlowfrom theriver. (iv)The area of the catchmentis greaterthan20 square kilometres. (v)There is no tidal influence atthestation. (vi)The catchmentispredominantlyrural. 2.2Extractionof AnnualFlood Data Themaxin~umpeakdischargeineachyearofrecordisextractedfromD.1.DStreamflowRecordPubl~ca- tions(Pre-1960 to1980). Each datasample is thoroughiyinvestigatedforincompleteddata (missing record duringanyyear).Ifrecords weremissingforaparticularyearthentheyareexamined to determine ifthe missingrecordsoccur duringthedryseason(whentheannualpeakdischargeisunlikelytooccur).I fitis suspectedthatthemissingrecordsincludestheannualpeakdischargethenthatparticularyear'sdatais treatedas missing. 2. 3Frequency Analysisof Individual Station FloodData Theannualfloodpeaksthatwerecollectedforeachstationwerereducedtothedimensionlessformof QJMAF. The MAF is def i ed as thearithmeticmeanof the annual floodseries: MAF= -Qi............................................ (2) where Qi=annual floodpeakforthe i th year n=number of records in years (number ofannual flood peaks inthe series) Theplottingpositionsof eachsampleisdeterminedbasedonthefollowing Gurnbel criteria (Haan,1977): (a) The plottingposition mustbesuch that all observations canbeplotted. (b) Theplottingpositionshouldliebetweentheobservedfrequenciesof(m- l)/nandmln wherem istherankoftheobservationbeginningwithm=lfor thelargestvalueand n is the number of years of records or the number ofobservations. (c) The returnperiodofavalueequaltoorlargerthanthe largest observation and the return periodof a value equal to or smallest observation should converge toward n. (d) The observations should be equally spaced on the frequency scale. (e) Theplottingpositionshouldhaveanintuitive meaning, beanalytically simple, andbeeasyto use. The Weibull plottingpositionformula meets all the criteria statedabove. Therefore, the pjottingpositions (recurrence interval in years) of each sample are calculated using the Weibullformula: where T =plottingposition of the peak discharges in years n=length of recordinyears m=rank of the peak discharge in the series Thedimensionless frequencycurveforeachsample is obtainedbyplottingratios ofQ,/MAF againstthe recurrenceintervalofQT usingtheGumbeldistribution.TheSrnirnov-Kolmogorov Goodness-of-Fit Test wasusedto testeachdistribution'sfitto the data. Only those distribution which fit the data at 95% confi- dence limits were accepted. AcomputerprogramwasdevelopedtofittheCumbeldistributiontothedimensionlessannualflood seriesbymethod-of-moments.Theprogramalsocalculatesthe95%confidenceh i t sandtestoutthe goodness-of-fit. Theprogramoutputs theQT/MAF values whereQTis QiforreturnperiodT. The results of' frequency analysis cdrried out on individual station'sdata are presentedin Appendix11. 2.4Derivationof Regional Dimensionless Flood Frequency Curve Stationswhichexhibitedsimilardimensionless frequencydistributionwereindentified andgroupedinto variousgroups.Thegroupingofstationsintoregions isdonebysuperimposing thedimensionless curves togetherandexaminingthesimilarityofthecurves.Ifacurvefromonestationliescloselyto a curve fromanotherstationthenthecatchmentsofthesetwostationsaregroupedtogetherunderoneregion. Considerationisalsogiventocatchments whicharelocatedclosely to oneanother.Catchmentsinclose proximity with one another are more lkely to beclassified under thesame region. Factorslrkeclimate, topographyandhydrologicalcharacteristics whichinfluence the flood flows ina river basinarealsotakenintoconsiderationbeforefinalkingtheregionalfloodfrequencyboundaries.The delineationofthefloodfrequencyregionsisguidedbytheMeanAnnualRainfailMaps(D.I.D,1973,HydrologicalRegionMaps(Goh,1974),AverageAnnualWaterResourcesMapsforPeninsularMalaysia (Teh,1982) andtopographicalmapsofPeninsularMalaysiapublishedbySurveyDepartment,Malaysia. Afterthefloodfrequencyregionshasbeenestablished,theregionalcurvefor eachregionis derivedby averagingthedimensionlesscurvesofstationsbelongingtothatregion.Theregionalcurveisthenthe representative flood frequency curve for all rivers in thatregion. Atotalof6floodfrequencyregionswereestablishedforPeninsularMalaysia.These regions areshown inMap1. and theregional curves belonging to each region are shown in Figure1. 2.5Derivation of Regional MAF (Mean Annual Flood) Equations ThemagnitudeoftheMAFofariverbasinisaffected byboththephysiographical, meteorological and catchmentcharacteristicsofthebasin.Catchmentarea,meanannualcatchmentrainfall,meanchannel slope,meanchannellengthanddrainage patternaresomeoftheeasily defined characteristics thatcould affect the catchment'sMAF. Ina studyonthedegree of correlation of catchment characteristics with the MAF (Nash and Shaw, 1965), itwasfoundthatacombinationofthecatchment areaandthemeanannual catchmentrainfall exhibits thehlghestcorrelationwiththe MAF.Table1showsthecorrelation coefficient for different combination of catchment characteristics with the MAF derived from the Nash and Shaw Study. Edd b.dd Sdd S'O S'E 1)ARS ii)AR iii)AS iv)RS v)A vi >R vii)S Table1:CorrelationCoefficientsofCatchmentCharacteristics with the MAF derived byBash andShaw A-catchment area. R-mean annual catchmentrainfall S-meanchannel slope Catchment Characteristics Inthis procedure,the catchment areaand mean annual catchment rainfall are the catchment characteristics chosenforthederivationoftheMAFequations.These two parameters are easily available. The catchment areasofstreamflowstationsusedintheprocedureareobtainedfromtheD.1.DHydrologicalStations Inventory(D.I.D,1987). ThemeanannualcatchmentrainfallforcatchmentsinPeninsular Malaysiawere abstractedfromthePeninsular MalaysiaMeanAnnualRainfalllsohyetalMap(1950-1975)(D.I.D, 1976) byplanirnetering of isohyets withineach catchment. Coefficient of Correlation Squared Therelationshipexisting betweencatchmentcharacteristicsandits MAFis assumedto bein theform of: MAF=cA' R'............................................ (4) WhereMAFis the mean annual flood Ais the catchment area Ris the mean annual catchment rainfall c,a,bare catchment characteristics constants to be estimated TheMAFequationisreducedtoa multipleh e a rregression bytransformingEquation(4) intoitsloga- rithmicform: Log MAF =log c + a log A + b logR............................................ ( 5 )The residual r, from the equation (5)is definedas: r=log (MAF obs.)- log (MAF est.)............................................ ( 6)whereris the log residual of the MAF equation MAFobs.is the observed MAF MAFest..is the estimatedMAF TheMAF regionsare established usingthesame methodadoptedinD.1.D Hydrological Procedure No.12 - "MagnitudeandFrequencyofLowFlowsinPeninsular Malaysia".(Toong,1985).Groupingofcatch- mentsintoregions isdone byseparating the residuals into positive and negative residuals. Catchments with positiveresidualsformedoneregionandcatchments withnegativeresiduals formedanotherregion.Re- fmementiscarriedoutbyrepeatingthegroupingineachregionuntilanideal numberofMAFregions isattained.SixfinalMAF regions areestablishedandtheseregions areshown in Map2.TheMAFequa- tions withthecatchmentcharacteristicconstantsandthecorrelationcoefficients squared derived for each region are JistedinTable 2. Table 2: Regional MeanAnnualFlood(MAF) Equations CatchmentCharacteristics Constants Region MAF1 MAF2 MAF3 MAF4 MAF 5 MAF 6 Multiple Coefficient of CorrelationSquaredMAF Equation 0. 79010. 1980 MAF= 0. 6582 AR 0. 65410. 8093 MAF= 0. 9630 AR 0. 61 753. 0571 MAF = 0. 1 192 AR 0. 71773. 0224 MAF = 0. 1048 AR 0.79Srl5. 0354 MAF0. 0140 AR 0. 90660. 9463 MAF= 0. 4783 AR A-CatchmentAreainkm. R-MeanAnnual CatchmentRainfallinmetres. NOTE: Ris measuredinmetres. 3.APPLICATIONOF PROCEDURE inthedevelopmentofthisprocedure,manyconstraints weresetbythenatureofthehydrologicaldata usedinderivingtheregionalfloodfrequencycurvesandtheMAFequations. Thereforein the application of this procedure, the catchmentofinterestshouldalso satisfythe following criteria: (i)Thecatchmentmustnotbesignificantlyregulated( byreservoir, diversion, etc). (ii)Thecatchmentmustnotbeinfluencedbytidaleffects. (iii)The catchmentareamustbegreaterthen20squarekilometres. (iv)The catchment is predominantlyrural. 3.1Methodof Application Step1 Step 2 Step 3 Step 4 Step 5 Step-6 Step.7 Themethodofapplicationofthisproceduret o estimatedesignfloods for an ungaugedcatchment involves the following steps: Determine the catchmentareaA in squarekilometres. Estimatethe meanannual catchment rai nfdRin metres (1000 xrnillimetres). i) Thecatchmentmeanannualrainfallcanbe estimatedfrom available rainfall records of D.I.D.rainfall stations withinornear the catchment. ii) Ifnorainfallrecordsareavailable,Rcanbeestimatedfromthe1:1,000,000 MeanAnnualRainfalllsohyetalMap(1950-1975) for PeninsularMalaysia(D.l.D, 1976). Note: The unitforR usedintheMAF regression analysis is inmetres. Determine the MAF regionofthe catchment from Map 2. Compute the MAF fromthe appropriare regional MAF equation. Determine the floodfrequency(FF region) region ofthe catchment from Map1. Obtainthedimensionlessordinates&/MAFfromtheappropriateregionalfloodfre- quency curves for the return periods required. Determine QTforthe variousreturnperiodsbymultiplyingtheQ,/MAFfactor by the MAF obtain in Step 4. 3.2WorkedExamples ExampleI - Determinethe30-yeardesigndischarge foranungaugedsiteon Sg.BatangKali located at~ a t . 3 ~ 23' N,Long. 101'38'E. Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Example 2- Catchment area A - 88 km2. Mean annual catchmentrainfallR = 2.S50m (2550mm). From Map 2, the site is locatedin mean annual flood region MAF 3. The MAF equationfor region MAF 3 is MAF=0.1192 A 0 ' 6 1 7 s R ~ ' ~ ~ ~ ~MAF = 0.1 192 (88 0 . 6 1 7 5) ( ~ . 5 5 0 3 ~ 0 5 7 11 MAF = 33.10 m3/s. From MapI ,the site islocated in floodfrequencyregionFF2. TheQT/MAFvalueforthe30-yearreturnperiodofregionFF2isobtainedfromthe floodfrequencycurve inFigureI :- QJ0/MAF =1.94. Q30 = Q30/MAF xMAF QjO =1.94 x 33.10 3 Q30 = 64.21 m1s. DerivethemeanannualfloodandfloodfrequencycurveforSg.LangatatKajang (DID Stn.no.2917442)withacatchmentareaof380kmandameanannualcatchment rainfall of2675mm (2.6751~1). Results:FromMap1 andMap2 ,thesiteislocatedinmeanannualfloodregionMAF3 andflood frequencyregionFF3.Followingthestepsinsection3.1. theresultsobtainedfromregional analysisarepresentedinTable3. Forcomparisonpurposes,theresultsofthesinglestation analysis are also presentedin Table 3.Table3:Resultsofdesignpeakdischarges derivedfromregionalanalysisandsinglestationanalysis for stn. no. 2917442 4.ACCURACYOF PROCEDURE Method ofMAF (m3/s) Analysis Regional94.59 Single Station99.61 Three d~fferentcomparisons are made to assess the accuracy of the procedure. The first comparisongives an indication ofhowthe10-year design peak discharges estimatedfrom regional analysis (using this procedure) varywith.the10-yearpeakdischargesrecordedbyriverstationsthroughoutPeninsularMalaysia.The secondcomparisonshows thedifferencesbetweendesignpeak discharges derived fromthis procedureHP4 (1987) anddesignpeakdischarges derived from the old procedure, HPq1974). The last comparison gauges theaccuracyofregionalanalysisusedin theprocedureascomparedtoothermethodsofdesignflood estimation used inflood studies for rivers in Peninsular Malaysia. 3 Q,for T-year recurrenceinterval (rn1s) T=jO 210.00 208.18 T=2 88.91 94.62 T=1 00 233.64 23 1 . I 0 T=5 127.20 131.49 T=IO 153.24 155.39 T=20 177.83 178.30 4.1Comparison with Observed Data Inthiscomparison,theobserved10-year peakdischargesfromallthesixtyoneriverstationsandthe 10-year peakdischargesestimatedusingregionalanalysisinthisprocedureareplottedintheformofa scatterdiagram inFigure 2. Itcanbeseenthatnomajordiscrepancies existbetweenthepeakdischarges derivedbyregionalanalysis andthatderivedfromsinglestationfrequencyanalysis. However,thescatter diagramshowsthat54.1percentofthefloodestimatesmadeusingthisproceduretends to beoveresti- mated.77.0percentofthe10-yearpeakdischargesestimatedfromthlsprocedurearewithintherange of 0.67t o 1.50 times the recorded values. Thepercentagebreakdownofthe10-year peakdischargesforalltheriverstations derivedfromregional analysisas comparedto the10-year peakdischarges derivedfromsinglestationfrequencyanalysis ispre- sentedinFigure3. F i g u r e 2:S c a t t e r d i a g r a mc o mp a r i n g Qlov a l u e s obt ai nedf r omt hi s pr ocedur eHP4 (19871,Qproand[rtoval uesobt ai nedf r o m si ngl es t a t i onf r equencyanal ysi sofobser veddat a(Qob, ) .Rat i oof PredictedQIO(regionalmalysis ) /observedQl o Figure3:Frequency diagramshowing thepercentage breakdown of theratios of Qlo values from regional analysis t oQ1, valuesfromobservedrecords. 4.2Comparison with HP4( 1974) Thisprocedure,HP;Y1987) isdevelopedtotallyindependentoftheoldprocedure,thefirst edition ofthe D.I.DHydrologicalProcedureNo.4"MagnitudeandFrequencyofFloods inPeninsularMalaysia"(Heiler andChew,1974).Theoldprocedurewasdevelopedbasedonstreamflowdatafromyear1948 to1970. Sincethen.anadditionaltenyearsormore data has beencollectedandmade availablefor analysis.Hence, thisprocedureutilizesalongerperiodofdata(upt oyear1982) inits analysis.Also, a differentapproach inthemethodof'analysis hasbeenadoptedt o improvetheaccuracyandperformanceoftheprocedure. Themajordifferenceof thisprocedureHPq1987)comparedt otheoldprocedureHPq1974)aresum- marizedbelow: (i)AdditionalStreamflowDataUsed Anadditionaltent otwelveyearsofstreamflowdata(upt o year1982) is usedt o derive the dimen- sionless flood frequency curves andthe meanannual floodregressionequations inthis procedure. (ii)Two Regional Maps Inthisprocedure,tworegionalmapsaredeveloped:oneforthefloodfrequencyregions (Map1) and the other for the meanannual flood regions (Map 2). The flood frequency regions are demarcated forareasinwhichthesetofdimensionlessfloodfrequencycurves(seeFigure1) applies whereas themeanannualfloodregionsaredemarcatedforareaswhichsharethesamemeanannualflood regressionequations coefficients (seeTable 2). (iii)AdditionalCharacteristicofMeanAnnual Cat chmentRainfall usedt oderive t he Mean AnnualFl oodRegression(MAF) Equat i ons Otherthancatchmentarea, anadditional characteristic,themeanannualcatchmentrainfallisused toderivetheregionalmeanannualfloodregressionequationsinthisprocedure.Unlketheold procedure(HP41974) whereonlycatchmentareas iscorrelatedwithmeanannualfloodstoderive theregressionequations coefficients,thisprocedureusedboththecharacteristics of catchment area and catchment meanannual rainfall to derive the meanannual floodregression equationtcoefficients. (iv)St reamfl ow dat a from di fferentgauging st at i ons used Streamflowdatafromgaugingstationsusedinthisprocedurearenotallfromthesamestations usedintheold procedure,HP4 (1974). Streamflow data from stations that do notfit into the Gumbel TypeIDistributionanddatafromstationsthatexhibitedgreatchangessince1970 (usuallyfrom catchments of non-homogeneous nature) are notincluded inthe anaiysis of this procedure. 4.2.1Comparison o fResults using HP4( 1987) and HP4(1974) ThemeandesignpeakdischargeestimatedforthreeungaugedcatchmentsatSg.BatangKali,Sg.Setiu and Sg. Lipis usingthis procedure HPq1987) and theoldprocedure HPq1974) are listed below: (i)Site: Sg. Batang KalatLat. 3"23' N,Long.101'38'E Catchmentarea A:88 km2 (34 mile2) Mean annual catchmentrainfall R = 2.550m (2550mm) HP4(1987):FloodFrequency Region FF2 Mean Annual FloodRegion MAF 5 MAF = 0.1 192 (88' 6' 75) (2.5503.05 1 = 33.10 m3/s. HP4( 1974): FloodFrequencyRegion F4 Q2.33=lo4 A 0 . 6 9 3= 104 (34.693, = 1197.70 ft3/sec = 33.94 m3/s. Table 4 :CornparisonforresultsobtainedfromHPq1987) andHPq1974) forSg.BatangKaliatLat. 3" 23' N,Long.101" 38'E 1IHydrological)Region1Qrfor T-year recurrence interval (m3/s) (ii)Site: Sg. Setiu at Lat. 5"31' N, Long.102'44'E CatchmentArea A:16 1 km2(62 mile2) Meanannual catchment rainfall R = 3490 mm Procedure Used HPq1987) HP4(1974) HP4 (1987): Flood FrequencyRegion FF5 Mean Annual FloodRegion MAF6 MAF= 0.4783 ~0.9066~0.9463 = 0.4783 ( 1 6 1 O . ~ ~ ~ ~ ) (3.49$.9463) = 156.35 m3/s. FF FF2 F4 M AF MAF3 - 2.33 33.10 33.94 5 43.36 39.03 10 51.31 43.44 100 77.45 55.32 20 62.3 1 46.84 50 59.58 5 1.S9 HP4(1974):FloodFrequencyRegion F4,Q2.33= 218 A 0.883 = 218(62.883) = 8339.47ft3/s = 236.33 m3js. (iii)Site: Sg. Lipis atLat. 4"00'N,Long.101" 40'E CatchmentareaA:130 ~ r n ~ (50 mile2) Mean annual catchmentrainfall R= 2.200m (2200mm) Table 5:Comparisonof results obtainedfromHPq1987)andHP.111974)forSg. Setiu aiLat. 5"31' N, Long.102'44'E H. 4 (1987):FloodFrequencyRegion FF 1 Mean Annual Flood Region MAF5 MAF = 0.0140 (130-79s4) (2.2Oos. O351 = 35.63 m3/s. Hydrological Procedure Used HP4(1987) HP4( 1974) HP4 (1 974):FloodFrequencyRegion F10 Qa.j3=1250A 0. 3860 =1 2 5 0 ( 5 0 ~ . ~ ~ ~ ~ 1 = 5658.64 ft3/sec =160.30 rn3/s. Table 6:Comparisonofresults obtainedfromHP4.1987)andHPq1974)forSg. Lipis at L.at.4"00' N, Long.101" 40'E Region Instatisticalstudies,erroranalyses areusuallycarried out to determinethe reliability ofdata and methods used.Hence,inthisprocedure,itisalsoessentialthatthestatisticalreliabilityoftheestimateddesign peakdischargesbeconsidered.Duetothecomplexnztureoftheerrorsinvolvedinthisprocedure,no theoreticalexpression forthestandarderrorisderived.Foruserstosubjectively evaluate theaccuracyof anydesignfloodsestimatedusingthisprocedure,thevarioussourcesandcauses oferrors are discussed below: QTfor T-year recurrence interval (m3/s) F F FF5 F8 Hydrological Procedure Used HPq1987) HPq1974) (i) .Lower peak discharge record from stick gauges MAF MAF6 - The presentautomaticrecorderstations (usedinthis procedure)thatwere established priort o 1960 were alloperatedbystickgauges beforebeingupgradedtotheautomaticrecordersystem.23% ofstreamflow stations presently usedin this procedure are still being operated ~ol el ybystick gauges. 100 537.84 385.22 5.RELIABILITY OF PROCEDURE 2.33 156.35 236.33 Region 20 383.06 326.14 . FF FFI F10 QTfor T-year recurrence interval (m3/s) 50 469.05 359.22 5 233.91 271.78 MAF MAF5 - 2.33 35.63 160.30 10 312.70 302.50 5 45.96 203.58 10 54.16 242.05 20 62,35 275.72 50 72,69 323.87 100 80.52 354.26 Forstickgauges,thewaterlevelisreadmanuallytwiceaday(8 a.m.and8p.m.daily).Duringflood times,thepeakwater-levelcouldhappenatanytimebetween8 a.m.and8 p.m.causingthepeaklevel tobemissedandnotrecorded.Alowerpeakwater-ievelwouldberecorded.This errorwilleventually resultinthe underestimationof the designflood estimatedusing this procedure. (ii) Poor quaIity rating curve and peakdischargerecords Rwercross-sectionsfrequentlychangeinurbanisedcatchmentswhererivershavebeencanalisedand widened.Otherchangesinrivercross-sectionalsohappened,especiallyinriversthatare. subseptibleto scouringandsilting.Whenariverhaschangeditscross-section,thentherecordswillberenderednon- homogenous. Theother factor that causes inacurratepeakdischargedata is the extrapolation of rating curves.For certain floodevents,thepeakwater-levelsrecordedarebeyondtherangeofthestage-dischargeratingcurves. Hence,theratingcurveswouldhavetobeextrapolatedinorderthatthepeakwater-levelsrecordedcan beconverted t o peakdischarge value. . Otherfactorslikeerrorsindatacollection,errorsindataanalysis, non-functioning ofwater-levelrecorders andinaccurategaugingmeasurementsalsoattributetothepoorqualityofpeakdischargerecordsofa streamflow station. (iii)Differentlengthof records Itisrecommendedin theU.S.G.S"FloodFrequencyAnalysis"(Dalyrmple,1960) that allperiodofrecords beadjustedt oa common baseperiod. The commonbaseperiod, derivedfromthe longestavailablerecords. isnecessarybecausestormsandfloodsofonep e r d timearedifferentinmagnitudeanddistribution tothoseofanotherperiod.Hence,fordifferentlengthofrecords,missingyears maybefilledinbycorro- lationtechniques.Theestimatedpeaksarenotuseddirectlybutservethepurpose of allowingthe correct recurrenceintervaltobeassignedtotherecordedpeakdischargedatainorderthattherecordscanbe compared or combined. Inthisprocedure,noadjustment of thelength of recordsto a commonbaseperiodwas carried outbecause thevariationinthelengthsofrecordsofstreamflowstationsusedinthisprocedureis nothighenoughto justifyfortheadjustmentofacommonbaseperiod.Theaveragelengthofrecordforthestreamtlow stationsusedis23. 2 years,theshortestperiodbeing8 ycrtrs andthelongestperiodis36 years.36'3 ofthe lengthof recordsi s between20to30 years.Hencetheerrorattributedbythedifferent lengthof recvrds isrelativelysmall. REFERENCES Beable,M.EandMckerchar,A.1(1982)"RegionalFloodEstimationinNewZealand",qa t e rand SoilTechnicalPublicationNo.20., WaterandSoilDivision, .Ministry.ofWorksandDevelopment, Wellington,New Zealand. Department of EngineeringHydrology,University College, Galway Ireland"FloodFrequencyEstima- tionfor UngaugedCatchments",University of Ireland, Galway, Ireland, U.K. Dalrymple,T (1960)"FloodFrequencyAnalyses",U.S.G.SWaterSupplypaper1543-A,U.S DepartmentofThe Interior, WashingtonD.C, U.S.A. DrainageandirrigationDepartment,Malaysia(1976)"HydrologicalData- RainfallRecordsfor PeninsularMalaysia1970 - 1975",Ministryof Agriculture, Malaysia. Goh,K.S(1974)"HydrologicalRegionsofPeninsularMalaysia"WaterResourcesPublicationNo. 2. , DrainageandIrrigationDepartment,MinistryofAgricultureandRuralDevelopment,Kuala Lumpur, Malaysia. Haan,C(1977)"StatisticalMethodsinHydrology",TheIowaStateUniversityPress,Ames,Iowa, U.S.A. Heiler,T.DandChew,H.H(1974)"MagnitudeandFrequencyofFloods inPeninsularMalaysia", DrainageandIrrigationDegartmentHydrologicalProcedureNo.4,MinistryofAgricultureand Fisheries, KualaLumpur, Malaysia. Nash,J.EandShaw,B.L(1965)"FloodFrequencyasaFunctionofCatchmentCharacteristics", kve r FloodHydrologyProceedingsofSymposium,InstitutionofCivilEngineers,London, United Kingdom. NaturalEnvironmentalResearchCounci1 (1975)"FloodStudiesReport, VolumeI- Hydrological Studies",N.E.R.C,UnitedKingdom. Shahin,M. A(1980)"StatisticalAnalysisinHydrology.VolIandVolII",InternationalInstitute for Hydraulic and EnvironmentalEngineering,Delft, Netherlands. Teh, S.KandKelsom, A (1982)"AverageAnnual and Monthly Surface WaterResources of Peninsular Malaysia"'WaterResourcesPublicationNo.12,Drainage andIrrigation Department, KualaLumpur, Malaysia. Toong,A.T(1985)"MagnitudeandFrequencyofLowFlowsinPeninsularMalaysia",Drainageand IrrigationDepartmentHydrologicalProcedureSo. 12.,MinistryofAgriculture,KualaLumpur, Malaysia. UnitedStates WaterResourcesCouncil (1976)"Guidelinesfor determiningFIoodFlowFrequency", BulletinNo.17 oftheHydrologyCommittee, U.S WaterResources Council, WashingtonD.C,U.S.A. UnitedStates WaterResourcesCouncil( 1967)"AUniformTechniquefordeterminingFloodFlow Requency",BulletinNo.15. of the HydrologyCommittee, U.S WaterResources Council, Washington D.C.U.S.A. APPENDIXI Listof Catchments andCatchment Characteristics LengthCatchmentObservedPredictedRegion StationofCdrchrnentMeanAnnualMAEMAF No.CatchmentRecordsArea( km2) Rainthll(13331s)(17131s)FFMAF (Years)mm Sg Johor at RantauPanjang Sg Sembrong atBrizayBr~dge Sg Bekokat Jln. YongPeng- AirHitarn Sg Kcsang atChin Chin SgLenggorat Bt. 42 Kluang - Mersing Sg Melakaat Pantai Belimbing SgLinggiat Sua Betong SgPedas at Kg.Pilin Sg Gemencheh at Jln. Gemas - Rompin Sg Muarat BuluhKasap SgMuarat Bt. 57 Iln. Gemas - Rompin SgLinggi atRahang SgMuaratKuah Pilah SgLangatat Dengkil S_e Langatat Kajang Sg Semenylh at Semenyih Sg Triang at Juntai Sg Combak at Jln. Tun Razak Sg Batuat Sentul Sg Triang at Jln. Keretapi Sg Selanpora tRantauPanjang Sg Sernantan.4t Jam. Keretapi SgPahang at Ternerloh Sp Bentongat Jam.K.Marong Sg Bernamat Tanjung Malim Sg Bernamat Jam. SKC Sg Trolakat Trolak Sp BilatJln. Tg. Malim- Slim Sg SlimatPekanSlim Sg Sungkai atSungkai Sg Bidor atBt. 9 Jln. Anson SgBidoratBt.18 Jln.Anson SgLipis at Benta Sg BatangPadangat T. Keramat Sg Batang Padang at Tapah Sg Gedong atBidor Sg J ehiatJam. Bunggor Sg Cherul atBanHo SF Ternbeling at Merting Sg KemamanatRan. Panjang Sg Kampar at Kg.Lanjut Sg Kinta atBt. Gajah Sg Kinta atKellas Sg PerakatJam. Iskandar Sg Dungun atJam. Jerangau Sg Plus atKg.Lintang Sg Ara atBt.20 Jh. Taiping Sg Krian atDusun Rimau Sg IjokatT,iti Ijok Sg Terengganu atKg.Tanggol Sg Krian atSelama ~g Galas atDabong 5405421Sg Kulirn atAra Kuda 5505412Sg Muda atVictoria Estate 5506413Sg Muda atBatu Pekitka 136.76I'F4MAF4 42.45FF3MAF4 59.32FF4MAF4 151.03FF3MAF3 94.59FF3MAF3 57.68FF3MAF3 88.86FF3MAF3 42.99FF3MAF3 49.25FF3MAF3 184.26FFIMAFS 217.50FF2MAF3 418.71FF1MAFS 1489.43FFIMAFJ 85.80FFIMAPS 66.21FF2MAF3 201.73FF2MAF3 34.33FF2MAFI 24.16FI-2MAF2 ll6.!)9FF?MAF? 85.61FF2MAF2 112.70FF2MAF? 105.87FF2MAF2 333.02FF1MAFS 123.07FF2MAF2 116.34FF2MAFZ 51.45FF2MAFZ 1053.27FF1MAFS 366.75FF5MAF6 652.05FFI MAFS 485.83FF5MAF6 107.91FF2MAF2 179.64FF2MAF2 71.99FF2MAF2 620.12FF2MAF2 956.33FF5MAF6 183.31FF2MAF2 59.37FF2MAFZ 143.83FF2MAFl 71.36FF2MAF2 1925.59FF5MAF6 131.27FF2MAFl 3684.33FF6MAF6 37.54FF2MAFl 545.64FF2MAFl 468.11FF2MAFl 5506416Sg Sedim atMerbauPulas264402835101.9799.22FF2MhFl 550641 7Sg Karangan at Titi Karangan1083310024.7527.04FF2MAFl 5721 442Sg Kelantan at Jam. Guillermard251 190024305236.865486.45FF6MAF6 5806414Sg Muda at Jeniang361710218.5310.26275.44FF2MAFl 6022421Sg Kemasin at Peringat2048287549.6943.43FF5MAF6 6204421Sg Padang Terap atLengkuas2 212701830185.69210.23FF2MAFl HYDROLIGAL PROCEDURES PUBLISHED Price H.P.No .1-EstimationoftheDesignRainstorminPeninsularMalaysia (Revisedand updated,1982). .. .. .. .. .. .. .. .MS 10.00 H.P. No .2-Water Quality Sampling for Surface Water (1973). .. .. .'M$3.00 H.P.No .3-A General Purpose Event WaterLevel Recorder ( 1973). . . . . .M$3.00 H.P.No .4-MagnitudeandFrequency ofFloods inPeninsular Malaysia (1 974). .M$6.00 H.P. No .5-RationalMethodofFloodEstimationforRuralCatchments inPeninsuiar Malaysia (1974). . . . . . .. .. .. .. .. .M$3.00 H.P.No .6-HydrologicalStation Numbering System (1 974). .. .. .. .M% 3.00 H.P. No .7-Hydrological Station Registers (1974). .... .... .MS5.00 H.P.No .8-FieldInstallation andMaintenance of Capricorder1599 (1971). .. .M% 5.00 H.P. No .9-Field111stallationand Maintenance of Capricorder1598 (1974). .. .M$5.00 H.P. No . I0-Stage-Discharge Curves (1977). .. .. .. ...MS5.00 H.P. No . ll-DesignFloodHydrographEstimationforRuralCatchmentsinPeninsular Malaysia ( 1 976). .. .. .. .. .. .... .M$6.00 1PYo . 1- MagnitudeandFrequencyofLowFlowsinPeninsularMalaysia(Revised and updated1985). .. .. .. .. .. .. .MS20.00 H.P. No . 13-TheEstimationofStorage-DraftRateCharacteristicsforRiversinPenin- sular Malaysla (1976)... .. .. .. .. .. .MS5.00 H.P. No .14-GraphicalRecorders- InstructionforChfirtChangingandAnnotation (1976). . . .. .. .... .. .. ...M$5.00 . . . . H.P.?io . 15-River DischargeMeasurementbyCurrent Meter ( 1 976)..MS5.00 H.P. 30.16-FloodEstimationforUrbanAreas inPeninsular Malaysia ( 1 976). .MS5.00 H.P. No .I7-EstimatingPotentialEvapotranspirationusingThePenmanProcedures (1977). . ... .. .. .. .. .. .. .MS5.00 H.P. No . 18-HydrologicalDesignof Agriculture DrainageSystems (1 977). .. .M$5.00 H.P. No . 19-The Determinationof Suspended Sediment Discharge (1977). .. .M$ 5.00 H.P.No . 20-HydrologicalAspectsRelatedtoAgriculturalPlanningandIrrigation Design(1 977). .. .. .. .. .. .. .. .M$10.00 H.P.No . 21-EvaporationDataCollectionusingU.S.Class"A"AluminiumPan (1981)MS 5.00 H.P.No..22-River Water QualitySampling ( I 98 1). .. .. .. .. .M$ 5.00 H.P.No . 23.-Operation and Maintenance of CablewayInstallations (1982). .. .MS5.00 H.P.No . 24-Establishmentof AgrohydroIogical Stations (1982). .. .. .M$3.00 H.P. No . 25-Standard Stick Gauge for River Station (1982). .. .. .. .M$5.00 H.P. No . 26-Estimationof Design Rainstorm in Sabah and Sarawak (1983). .. .M$10.00 HYDROLOGICAL PROCEDURE NO. 5 RATIONAL METHOD OF FLOOD ESTIMATION FOR RURAL CATCHMENTS IN PENINSULAR MALAYSIA JABATAN PENGAIRAN DAN SALIRAN KEMENTERIAN PERTANIAN MALAYSIA Hydrological Procedure No. 5 RATIONAL METHOD OF FLOOD ESTIMATION FOR RURAL CATCHMENTS IN PENINSULAR MALAYSIA (REVISED ANDUPDATED) 1989 Bahagian Pengairan dan Saliran Kementerian Pertanian Malaysia RATIONAL METHOD FLOOD ESTIMATION FORRURAL CATCHMENTS IN PENINSULAR MALAYSIA (REVISED ANDUPDATED) CONTRIBUTORS: AZMlMD. JAFRl ZAHARIOTHMAN Price: DRAINAGE ANDIRRIGATION DIVISION MINISTRY OF AGRICULTURE MALAYSIA First Published in 1974 Revised and Updated in1989 CONTENTS Page .... . . ..................1.INTRODUCTION... 2 .THERATIONAL METHOD ASA FLOOD ESTIMATION PROCEDURE . . ..... . . . . . . . . ............2.1General...... ....... . . ...2.2Essential Features oftheRational Method 3 .THEINVESTIGATIONS................ . . .... . .............. . . .........3.1General...... . . . .........3.2Methodology ofthe investigation. . . ... . . . ...............3.2.1Design Sequence...... ....... . . ...3.2.2Estimation ofTime ofConcentration... . . . 3.2.3Estimation ofthe Average Intensity ofthe Design Storm. . .............3.2.4Estimation ofRunoff Coefficient.... . .3.2.5Regional Runoff Coefficient C based in Flood Frequency Regions ...............4 .ACCURACYOF THE PROCEDURE... . . . .........4.1Comparison w~t hDID HP5 (1972). . . . . . .. . . 4.2Comparison with the Regional Flood Frequency Analysis... .... . . 5 .LIMITATIONS OF THEPROCEDURE. . . ... .... . . 6 .USE OF THEPROCEDURE. . . . . ..... . . . . . ....... . . .... . . REFERENCES. . . ... APPENDIX A .Multiplying Factor totake Account ofCatchment Development ...APPENDIX8 .Check on Estimation ofTcUsing Empirical Relationships TheRational Method offlood estimationis widelyused throughoutthe world.Recent studieshave shown - thatit is most useful i f regarded as a statistical link between the frequency distributions of the design rainstorm and thedesign flood.In thisprocedure, thestatisticalapproachwasadopted andused toprepare a flood estimation forsmall rural catchments up to100 square kilometres in Peninsular Malaysia. With the establish- ment of afewrepresentative andexperimentalcatchmentsand theincreaseinrecord lengthofsome of the catchmentsused in the previous study, more dataare available forreviewing and bpdating the old prc- cedure. In this study, records from 20 small rural catchments with 5 years ormore ofcontinuous data were analysed to provide new design data. Since the methodology adopted in the new edition was basically similar to the previous edition, the presentation and arrangement are kept as similar as possible for the convenience oftheusers. 1 .INTRODUCTION. Thisprocedureistheresult of a studyintotheapplicabilityoftheRational Method offloodestimationfor smallruralcatchmentsin PeninsularMalaysia. Theuse ofthe Rational Method forurban environment has workedreasonably well in many countries. Forrural catchments the use ofRational Method hasreceivedmuch criticism. Few overseasresearchers who have studied the method as a deterministicmodel and tested it with observed data tqund that the method gavelow accuracy when individual storms and resulting peak discharges were considered. However, studies byFrench etal. (1 974) whoexamined the validityofthemethod have shownthatstatisticallythemethod served the purpose of engineering practice wherepeak discharges of a given frequencyare linked withthe rainfall intensities ofthesamefrequency. Since the annual total expenditure on many small hydraulic structures such as bridges, culverts, diver- sion worksand so on amounted from a hClndred thousand dollars totens of million, the need tohave a pro- ceduretoguidethepractitioners toarrive atamorereasonabledesignofsuchstructuresisbothurgent and important. For this purpose DID HP NOS(Heiler.1974) has been published and used asa basis for the designofabovestructuresbymostpractitioners. 2 .THERATIONALMETHOD ASAFLOODESTIMATION PROCEDURE 2.1General The StatisticalRationalMethod forpeakdischargescanbewrittenas: whereQ, isthepeak dischargeofadesignfloodin m3/s withreturn period Tyearsselectedusingany recommended designreturnperiodsapplicableinMalaysia. i T istheaverageintensityofthedesignrainstorm ofdurationnormallytakenasbe~ngequalto time ofconcentration,Tc and ofreturnperiod Tyearsinmmlhr. Aisthecatchmentareainkm2. C-is a dimensionlessrunoff coefficientnormallyconsidered tobe a function ofthe catchment and designstormcharacteristics andofreturnperiod Tyears. and Tc, theso-called timeofconcentrationreferred aboveisdefinedasbeingthetimetakenfor adropofwatertotravel fromthemostremotepartofthecatchmenttotheoutletdesignpoint. 2. 2EssentialfeaturesoftheRationalMethod TheRationalMethod isgenerallyconsidered tobe oneofthebestavailablefloodestimation procedures forsmallurban andrural catchmentareas. However, much confusionhasresulted inthepastandisstill commontodaywithregard totheprinciplesunderlyingthemethod. TheRationalMethod hasbeen criticised becauseofitsinabilitytoreproduce particularfloodevents whenactual rainfall is used as theinput. Such criticism implies anassumption thatthe method is a "deter- ministic"one torepresent the physical operation on rainfall-runoff process. Thisis notthe intended use of themethod.Ithaslong beenrecognisedthemostrealistic waytouse the Rational Method istoconsider it as a "statistical"link between the frequency distributions ofrainfall and runoff. Assuch it provides a mean of estimating the design flood of a certain return period, and of a duration equal to the time of concentration. Rewriting equation (1)in aslightlydifferent form,thefollowingequationresults: whereC, isnow takentobe dimensionlessstatisticallinkbetween thefrequencyof peakdischarge, qin m3/sper km2and mean intensity ofthe designstorm i , in mmlhr. Subsequently, fora particularcatch- - menthaving adequateflood and intensity data,designvalues ofrunoffcoeficients,CT canbederived by theuse oftheformula where q, is a peakrunoffrate ofreturn period T years derivedfrom a frequencyanalysis of observed flood and i, is the mean design storm intensity of duration equal to Tc derived from a frequency ofstorms of dura- t ion equal to Tc. Using therelation intheabove equa'tiona consistentincreaseinC,withincreasingreturnperiod for 20smallruralcatchmentsin Peninsular Malaysiaasshownin Table1wasfound. Thisconformswiththe investigationbyFrenchetal. (1974) on37ruralcatchmentsinNew South Wales, Australia. 3. THEINVESTIGATION 3.1General It is the intention ofthis section to outline the general methodology employed fortheinterest ofthe general user, ratherthanto describein detail the developmen: of theprocedure. Thefrequencyanalysis ofthean- nual maximum flood data from the catchments are no:covered. Interested readers should refer to the revis- ed and updatededitionsofDID HydrologicalProcedureNo. 1andNo. 4fordetails 3 2Methodologyoftheinvestigation 3.2.1Design Sequence In using theRationalMethodforfloodestimationtheusual designsequenceisasfollows: (a) estimatethecritical durationofthedesignstorm(made equaltoTJ. (b) compute thevariousvaluesofmeanintensity(i,)forduration=T,(c) estimatethevalues ofC, fromstorm andfloodfrequencyregions, (d) computetheestimatesofpeakdischarge(aT)forvariousvaluesofTp usingequation(1). 3.2.2Estimationoftimeofconcentration An essential part ofthe Rational Method is the estimation ofthetime ofconcentration, Ti previously defin- ed in para 2.1. While thisis anunrealisticphysical conceptin a natural catchment, thereis little doubt that some "characteristictime"exists that is critical forparticular catchment. This cannot be defined precisely. and probably variesfrom season toseasonand from storm tostorm. Variouspractical methods werepro- posed bypreviousresearchers to estimate timeofconcentration. Attheinitial stage ofthestudy, average timelags between centroids of rainfall hyetographs to the peak ofcorrespondinghydrographs asproposed bySchaakeetal. (1967) wasattemptedtoestimateT-.Duetoinadequaterainfalldataforseveralcat- chmentsused inthestudy, derivationofTc bythismeihod wasabandoned. Theminimumtimeofriseof a flood hydrograph thatreflects total catchment contribution as proposed by Pilgrim et al. (1974) wasselected for this study. Theminimum time ofrise against duration of storm was studied for20 small rural catchments in Peninsular Malaysia. Atypical plot of therelationship is shown in Figure1wherethe typical time of rise wasadopted asthetime ofrise forthatparticular catchment. In ordertoestablish a relationshipbetween the time of concentration and the physiographiccharacteristics of the catchments, amultiplelinear regres- sion analysiswascarriedoutby presupposingarelation oftheform TJL=KAaSb....................(4) whereT, isthetimeofconcentrationinhours Listhelengthofthemainstreamin kilometres. Sis the slope in percent from the main stream catchmentboundry intersection to the design point, measured alongthemain stream. k,a and b areconstants. Transformingequation(4)intologarithmicformtheequationwhichresultedfromtheregression analysis was withamultiplecoefficientofcorrelation=0.87 between theobservedandestimated valuesofT-. The graphicalsolutiontothisequationispresentedinFigure 2. 3.2.3Estimation oftheaverageintensity ofthedesign storm, i, The method ofestimating the design rainstorm contained in therevised and updated 010Hydrological Pro- cedureNo.1(Fadhlillahetal.1982)hasbeenusedthroughoutthisinvestigationforcomputingthe characteristicsof the design storm for eai h of the study catchments. The only input for using this procedure i s the durationofthe storm, whichis made equal to Tc found fromequation (4), and the geographicalloca- tion of the design point. Note thatthe design intensity should be adjusted totake accountof thereduction in stormintensity with catchmentareaaccordingl o Table 6, page12 oftheabove procedureorFigure 6 page10 ofDID WaterResourcesPublicationNo17. 3.2.4Estimation ofrunoffcoefficient,C TherunoffcoefficientC in theRational Method whichis oftenregarded asasimpleparameteris complex and affected by various factors and processes. Several factors affecting runoff coefficients Care infiitrat~on losses, variation ofrainfall intensities, catchment storage, antecedentwetness and physical characteristic of the drainagearea. Various approacheshave been made available topresent designrunoff coefficient, Cintabularselectiontables,graphicalrelationsand simplerecommended valueswhichcanbefoundin variousreferencebooks.Mostoftheseapproacheshavebeenbasedonengineeringjudgementand experienceratherthanvaluesderivedfromobservedflooddata.Inthisprocedure,theapproachusedwas toderiveCforvariousreturnperiodsfromfrequencyanalysisofobservedflooddataanddesignrainfall intensities for20smallruralcatchmentsinPeninsularMalaysia. 3.2.5Regional Runoff coefficient, Cbased onFloodFrequency Regions Mean values ofrunoff coefficientforareturn period of10 years, C,, werecomputed foreachregion. For the application of this procedure, Peninsular Malaysia was divided into 4 regions (Figure 3). These are Flood Frequency Regions based onDID HP No. 4(Ong, 1987) where runoff coefficients C withinthe same region would not be much affected by the frequency distribution of flood peaks and the flood producingrainstorms. Mean valuesof C,IC,,whereC, arerunoffcoefficients ofreturn period T yearswerecomputed.Afamily of curvesrepresenting differentregions was established as shown in Figure 4. Knowing theregion in which aparticular project lies and using the appropriateregional runoffcoefficient curve, the factor C- I C. ,ISob- tainedforanyparticularreturnperiod and henceC, canbe obtained. 4.ACCURACYOF THEPROCEDURE Three methods were employed to assess the accuracy ofthenew procedure. Onemethod was the scatter diagram (as showninFigure5) whichcompared the10 years design peakdischargesestimatedusing the new procedurewiththe10 yearspeak dischargesofsinglestationfrequencyanalysis of observeddata. Anothermethod was the comparison between the design peak discharges derived using thenew pro- cedure and the designpeakdischargesderived using thepreviousprocedure(DIDHP5,1974). Thethird comparison checked the accuracy of the procedure by comparing it with the regional flood frequency analysis (DIDHP4,1987). 4.1Comparison withHP5 (1972) Table 1 compared the results obtained from DID HP5 (1988) and DID HP5 (1974) for two single stations whlch areSg.Durian TunggalatBatu 11, AirResam and Sg. ChalokatJambatan Chalok. Althoughthe values estimated using HP5 (1974) gave higher value, the accuracy of the new procedure (HP5, 1988) should have beenimprovedforthefollowingreasons: (i)Additionalstreamflowandrainfall datauptothelatestrecord (1986) havebeenusedwherever possible. - Table 1 - Details of Study Catchments TEM I STATION NO. - STATION NAME Parit Madirono Sg.Permandi atBt. 27 J.B./Mersing Sg.Mupor atBt. 32 Jalan J.B. Sg.Kahang at Ulu Kahang Sg.Durian Tunggal atBt.1 1AirResam Sg. Mantau atKg. Mantau Sg. Kepis at Jam. Kayu Lama Sg.Pontian at Jam. Balak Sg.Lui at Kg. Lui Sg.Batu at Kg.Sg.Tua Sg.Serai atKg.Unchang Sg. Tekam,Site A Sg. Tekam, Site C Sg. Gedong at Bidor Sg. Cendriang at Bt. 32 Jln. Tapah Sg. Chalok at Jam.Chalok Sg.Lanas at AirLanas Sg. Kemasin at Peringat Sg. Pelarit at Titi Konkrit Sg.Buloh atBatu Tangkup CATCHMENT AREA KIM" MAINSTREAM LENGTH KM 1.96 6.44 6.44 21.6 14.48 6.4 11.0 11.9 15.2 14.48 12.07 0.99 1.34 24.0 10.1 7.0 15.2 17.4 10.8 6.17 SLOPE AT SOURCE % 0.50 0.90 0.87 3.30 1.50 3.40 1.49 0.47 4.03 7.98 0.20 0.01 3 0.80 4.99 2.07 0.83 3.80 0.06 1.52 0.60 Tc (hrs.) 2.62 4.23 4.32 7.87 6.44 3.37 6.46 7.41 5.29 4.40 9.57 4.97 2.08 7.12 3.70 4.82 5.18 19.79 6.96 6 .O4 DERIVED RUNOFF COEFFICIENTS(C) H.PNo. HYDROLOGICALPROCEDURESPUBLISHED Title Estimation oftheDesignRainstorm inPeninsularMalaysia (Revised and updated, 1982) WaterQualitySamplingforSurfaceWater(1 973).... .. . . . . ........ .. ..-.. .... AGeneralPurposeEventWaterLevelRecorder~ a ~ r i c o r d e rM ode11598 (1973) Magnitude andFrequency ofFloodsinPeninsularMalaysia (1974) Rational Method ofFlood Estimation forRural Catchments i n PeninsularMalaysia (1 974) HydrologicalStationNumbering System(1974)........ .............. ........ Hydrological StationRegisteis (1974). . ........ .. .. .. .. .......... ...... ... . .. FieldInstallationand Maintenance ofCapricorder1599 (1974) FieldInstallation andMaintenance ofCapricorder1598 Digital Event WaterLevel Recorder (1974) Stage-Discharge Curves (1 977).. ..... ....... . . . .. .. .. .. ... ... .. ... . ... ... .. . Design FloodHydrograph Estimation forRural Catchments i n Peninsular Malaysia (1974) Magnitude andFrequency ofLowFlows inPeninsularMalaysia (1976) TheEstimation ofStorage-Draft Rate Characteristiesfor RiversinPeninsularMalaysia (1976) GraphicalRecordersInstructionsforChart Changing and Annotation(1 976) RiverDischargeMeasurementbyCurrentMeter(1976). . .. ..... . .... ........ FloodEstimation forUrban AreasinPeninsularMalaysia (1976) EstimatingPotential EvapotranspirationUsing thePenman Procedure (1 977) HydrologicalDesign ofAgricultureDrainage Systems (1977) TheDeterminationofSuspended SedimentDischarge (1977) Hydrological Aspects ofAgriculturePlan'ning and IrrigationDesign (1 977) Evaporation Data CollectionUsing U.S. Class"A" AluminiumPan (1981) RiverWaterQualitySampling(1 981)..... .. ..... ... ... ...... .. . .. ..... . ..... Operation andMaintenance ofCableways Installation (1981) Establishment ofAgro-hydrologicalStations(1982)........ ... . ....... . ...... Standard StickGaugeforRiverStation(1982)........ .... ........ .. .. .. .. .... Estimation ofDesign Rainstorm in Sabah and Sarawak (1983) Price $1 0.00 $3.00 $5.00 $6.00 $3.00 $3.00 $5.00 $5.00 $5.00 $5.00 $5.00 $5.00 $5.00 $5.00 $5.00 $5.00 $5.00 $5.00 $5.00 $10.00 $5.CO $5.00 $5.00 $3.00 $5.00 $1 0.00 (ii) (iii) (iv) Therainfallregionshave beendividedintofourascompared toonlytwointhepreviousstudy (HP5,1974). Further subdivision ofrainfallregions isnot possible duetothelack ofsmall catch- ment data. Streamflow stations other than the stations used in theprevious procedure (HP5, 1974) have been included. Most of the stations in this study (HP5, 1988) are recorder stations. These stations provided continuous recorded data which should be more accurate in term ofreliability than manual stations. Most of the streamflow stations in this study have more gentle slopes compared to the stations used in the previous study. This will be more representative ofmost ofthe practical river catchments in Malaysia and thesteeperslopesin theprevious studymay have contributed tohigherflows. Table 2 - Comparison ofResults by VariousMethods -- -- Method -- Q,,valuefor Sg.Durian Tunggal atBt.11 AirResam (m3/s) O,,value for !3g. Serai at Kg.Unchang (m3/s) --- - - - Ratronal Method (1 974) Rat~onalMethod (1 988) Frequency Analys~sOf Observed Data Regtonal Flood Frequency AnalysisHP4,1987 4.2Comparison withthe regional floodfrequency analysisDID (HP4,1987) The comparison ofDID HP5 (1988) toDIDHP4 (1987) in Table2showed thatDIDHP4 (1987) estimated a higher value. It should be noted that DID HP4 (1 987) used the extension ofrecords and other informations fromgauged sites ofclose proximity tocoveraregion togetaregional mean annual flood equation. 5.LIMITATIONS OF THE PROCEDURE From the theoretical basis oftheRational Method, two important factorsare neglected. These are the ef- fects ofchannel storage and temporal and spatial variations ofrainfall intensities. Due tosuch limitations and also that the procedure wasderived utilizing data fromrural catchments with areas ranging from 0.3 - 100 km2, the use ofthe procedure toestimate therunoff forlarger areas is not recommended. Mult~ply- ing factors to take account of catchment development (Appendix A)should serve as a general guide to arrive at a reasonable estimate although there has been no study tosubstantiate thisrecommendation. Thepro- cedure may not be reliable to estimate therunoff for,areaswithsteeperslope.Forgeneral guidence the limit is approximately from 0.01 - 8 percent; the slope values ofwhich this procedure was developed. Like anyotherflood estimation procedures, the design flood obtained from thisprocedure should bechecked with other available procedures and the decision to adopt the estimated design values should be complemented bya sound engineering judgement. 6.USE OF THE PROCEDURE 6.1Component ofthe procedure The following items arerequired touse thisflood estimation procedure: (i) Figure 3, showing thegeneral disposition offourregions proposed forapplication ofFigure 3, (ii)Figure 2,being a graphical solution tothe Tcformulapresented asequation (5), (iii)'Figure 4, showing therelationship ofmean frequency factor C,./C,, fordifferentregions and (iv).D.I.0 Hydrological Procedure No. 1 (1982) "Estimation of the Design Rainfall in Peninsular Malaysia". 024681012 Dur at i onof St or m(hours) F i g u r e 1- Re l a t i o n s h i p Be t we e n Ti meo f Hydr ogr aph Ri seandDur at i onof St or mFor Sungai Gedonga t Bi d o r .5 1 0 2030405060708090100110120130140 Ca. t chment A r e a I urn2) Fi gur e 2- Ti meofConc e nt r a t i onGr a p hGr aphi cal Sol ut i onofTc=1. 286L ~ 0 . 2 2 3 SO- 263 L e g e n dVDi schar geS t a t i o