Journal - April 2012

7/28/2019 Journal - April 2012

1/71

Vol.: XXXXV, No. 02, April 2012 ISSN 1800-1122


2/71

ENGINEER

JOURNAL OF THE INSTITUTIONOF ENGINEERS,SRI LANKA

EDITORIAL BOARD

Eng. (Prof.)A. K. W. Jayawardane

From the Editor..

SECTIONI

CONTENTSVol.: XXXXV, No. 02,April 2012

ISSN 1800-1122

III

Eng.Priyal

De Silva

Eng. W. T. R. De Silva

Eng. (Prof.) K. P. P. Pathirana - Editor Transactions

Eng. (Prof.) T. M. Pallewatta - Editor ENGINEER

Eng. (Prof.) (Mrs.) N. RathnayakaEng. (Dr.) D. A. R. Dolage

Eng. E W Karunarathna

Eng. (Miss.) Arundathi Wimalasuriya

Eng. M. L. Weerasinghe - Editor SLEN

Eng. (Dr.) K. S. Wanniarachchi

The Institution ofEngineers, Sri Lanka

120/15,Wijerama Mawatha,

Colombo - 00700

SriLanka.

Telephone: 94-11-2698426, 2685490, 2699210

Fax: 94-11-2699202

E-mail:[email protected]

E-mail(Publications):[email protected]:http://www.iesl.lk

The statements made or opinions expressed in the

Engineer do not necessarily reflect the views of the

Council or a Committee of the Institution of

Engineers SriLanka, unless expresslystated.

COVER PAGE

National Performance Theatre

Constructed along the lines of the historic Lotus Pond or

Nelum Pokuna(inset - middle left) of the Polonnaruwa

era, the National Performance Theatre, a colossal andspectacular auditorium structure built in the heart of

Colombo was presented to the public recently.

Constructed with aid from China, at a cost of overRs. 3

billion, this theatre has many modern amenitiesincluding

a movable stage, state of art lighting and sound controlsystems. With a seating capacity of 1288 in the main

auditorium, coupled with 500 vehicle parking spaces,

rehearsalhalls and a library, thisindeedis a nationallevelamenity.

Key Issues ofData and Data Checking for 1Hydrological AnalysesCase Study of

Rainfall Data in the Attanagalu Oya Basin of

Sri Lanka

by : Eng. (Prof.) N T S Wijesekera and Eng. L R

H Perera

Performance of tall buildings with and 13without transfer plates under

earthquake loading

by : Eng. (Prof.) M T R Jayasinghe,Eng. D S

Hettiarachchiand Eng. D S R TN

Gunawardena

Flood InundationMapping alongthe Lower 23Reach ofKelani RiverBasinundertheImpact ofClimate Change

by : Eng. (Miss.) M M G T De Silva,Eng. (Prof.)

S B Weerakoon, Prof. SrikanthaHerath, Eng.

(Dr.) U R Ratnayake and Dr. Sarith Mahanama

An Application ofDistributed Hydrological 31Model, YHyM/BTOPMC to Gin Ganga

Watershed, Sri Lanka

by : Eng. (Mrs.) T N Wickramaarachchi, Eng.

(Dr.) H Ishidaira and Eng. (Dr.) T MNWijayaratna

SECTIONII

Evaluation and Selection ofTools forData 41Migration from Non-Spatial to Spatially

Referenced SoftwareA Case Study

Migration from MySQL to PostgreSQL

by : Miss. EA G Chandramaliand Eng. (Prof.)

N T S Wijesekera

Mortar Consumption Characteristics of 49'Brickwork' and a Framework for ManagingBrick and Mortar Walls in Chaotic

Environments

by : Eng. (Dr.) VasanthaAbeysekera

Notes:

ENGINEER, established in 1973, is a Quarterly

Journal, published in the months of January,

April, July & Octoberofthe year.

All published articles have been refereed in

anonymity by at least two subject specialists.

Section I containsarticlesbased on Engineering

Research while Section II contains articles of

ProfessionalInterest.
mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/http://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%206.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%205.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%204.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%203.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%202.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/Engineers%20Journal/April%202012/Paper%201.pdfhttp://www.iesl.lk/mailto:[email protected]:[email protected]


3/71


4/71

FROM THE EDITOR..

It is a well established fact that Engineers are a Left Brained set ofpeople. In

other words, individuals quite good in mathematics and logical reasoning. Though we

could be proud of this status, this also could well imply that our Right Brain functions

may not be up to the standard of a common person. Incidentally, it so happens that

people talk, to put it mildly, about the shortcomings of Engineers in general, when it

comes to inter-personal skills, artistic appreciation, etc., which are supposed to be in the

right brain domain. I say this with caution so as not to offend those Engineers who have

excellent artistic creativity coupled with capacity for appreciation of art as well as other

numerous soft skills.

In fact, as a friend of mine who is a Consultant Psychiatrist has told me several

times that a developed left brain naturally stimulates the right counterpart positively.

This is to say that the potential for art and other soft skills are at least latently lying

within us for the want of a stimulus. Why most of us do not get this breakthrough could

be due to many reasons, but mostly to be attributed to being engrossed during the better

part of our lives in our drabmathematical and technical world. It is a well known fact

that when a being gets used to a certain way oflife, it is ratherdifficult to break out ofit.

Whatever the minor and majorreasons could be, we Engineers should strive to

be more outward bound in our artistic and inter personal approaches, to be a successful

professional in the present society. However, one should be cautious too, not to overdo

it through ones over enthusiasm.

Eng. (Prof.) T. M. Pallewatta, Int. PEng (SL), C. Eng, FIE(SL), FIAE(SL)

Editor, ENGINEER, Journal of The Institution ofEngineers.

III


5/71


6/71

ENGINEER - Vol. XXXXV, No. 02, pp. [page range], 2012 The Institution of Engineers, Sri Lanka

ENGINEER 1

Eng. (Prof.) N. T. S. Wijesekera, C. Eng., FIE(Sri Lanka),B.Sc. Eng. (Sri Lanka), M. Eng. (Tokyo), D. Eng. (Tokyo),Senior Professor of Civil Engineering, Department of CivilEngineering, University of Moratuwa, Sri Lanka

Eng. L. R. H. Perera, B.Sc (Eng) (Moratuwa), M.Sc (Delft),C. Eng, MIE (Sri Lanka), MIAHS (UK), Chief IrrigationEngineer, Irrigation Department, Colombo, Sri Lanka

Key Issues of Data and Data Checking forHydrological Analyses - Case Study of Rainfall Data in

the Attanagalu Oya Basin of Sri Lanka

N. T. S. Wijeserkera and L. R. H. Perera

Abstract: Inconsistencies and non-homogeneities in the hydrological and meteorological timeseries could be identified by incorporating statistical tests that detect trends and change points.Inconsistency which reflects systematic errors during recording and the non homogeneity that arisesfrom either natural or man made changes to the gauging environment are both important foradequate time series analysis. It has also been identified that statistical tests together with physical orhistorical evidence and justifications from metadata need to be incorporated for a very detailed study.A case study was carried out for the rainfall data of Attanagalu Oya basin in the western province ofSri Lanka with a data set consisting of six stations having daily rainfall data for 30 years. According toPettitt test, a significant change around 1977 & 1985 at Karasnagala and Pasyala could be found.

However Pasyala is the most significant station for the change of rainfall pattern, which wasconfirmed by t-test. Knowledge of Meta data was found very important in order to make necessarycorrections to shifts identified through Double Mass Analysis. This paper shows that statistical testsand rational judgements would enable suitable corrections even though it is common to find that mostof the hydrological and meteorological data are either flagged for quality or poorly documented.

Keywords: Issues, Data Checking, Hydrological Analyses, Rainfall, Sri Lanka

1. IntroductionWater resources development and managementis heavily dependent on hydrological andmeteorological data. In order to make sure thatthe results obtained from these data are reliablefor practical applications, such data should be,homogeneous and consistent either to carryoutfrequency analyses or to simulate ahydrological system [1]. In hydrologic analysisit is customary to search for long datasets sincesuch data ensures that the sample takenrepresents the system performance. However,longer the time series the greater are thechances that the data series is neitherstationary, consistent nor homogeneous. It isalso necessary to identify the spatialrepresentation of the data used in an analysis.In case of precipitation, spatial distribution ofrain gauges is often non-representative sincethey are mostly located in the valleys whereeasy access is the main criteria. It has also beenidentified that in many mountainouscatchments, the higher elevations receive moreprecipitation than the regions in the valley [2].As such, prior to a responsible hydrologicalanalysis, a suitable spatial and temporalanalysis of data needs to be carried out throughan efficient screening procedure.

As there are many organizations havingdifferent objectives perform data collection,there is also a necessity to check suchobservation data series for consistency and

homogeneity. It is common to use statisticaltests, either parametric or non-parametric, inorder to detect the non-homogeneity in a timeseries. The choice between the two families oftests is based on the expected distribution ofdata involved. If data set is normallydistributed, parametric tests are usuallyselected. If data set is expected to be non-normally distributed, non-parametric tests arepreferred. Also it has been identified that somehomogeneity tests depend on meta-data whilethe others are purely statistical. The presence

of a single significant test result is considered asa weak evidence of change. In case of moreresults that are significant and not very similar,then they need to be taken as stronger evidenceof change [3].However it should be emphasized thatapplication of more than one test to data may


7/71

ENGINEER 2

make interpretation of the results rathercomplex. Due to differences in assumptionspertaining to each test, along with possibleinfluence with the change of catchmentcondition, it has been identified that it isusually difficult to compare the results ofdifferent tests. Since it is particularly difficult

to combine the results of different tests,distribution free testing methods arerecommended for hydrological data which areoften non-normally distributed [3].

2. Study Area & Data AvailabilityAttanagalu Oya basin which drains to thewestern coast of Sri Lanka (between 79 50' &80 7'E and 6 59' & 7 17' N) is having acatchment area of 727km2. The spatial coverageof the basin shows that it spreads over twoprovinces namely the Western andSabaragamuwa and flows through theGampaha and Kegalle administrative districts.The basin has an elevation of about 300m MSLas its highest. There are several large streamsthat combine to drain Attanagalu-Oya and theyare namely, Kimbulapitiya Oya, Mapalan Oya,Dee-eli Oya and Uruwal Oya (Figure 1).

There are 18 rainfall gauging stations locatedeither within the basin boundary or just outsidethe boundary. The rain gauging networkmaintained by Department of Meteorologyconsists of 17 stations, of which 16 do notpossess automatic recording facilities butmaintain daily records. The other one atKatunayaka in the vicinity of the catchment is arecording type. There is a recording type raingauge at Karasnagala, maintained by IrrigationDepartment..Based on the data availability and spatialcoverage, daily data of six stations wereselected for the study. This study considereddata from 1970 to 2001. Station names anddetails of missing rainfall data during the saidperiod are shown in Table 1.

3. MethodologyThe following tests were carried out in thisstudy with the use of the SPELL-Stat software[4].

1) Visual examination of Data2)

Outlier Testing3) Homogeneity Testing with,

Test for serial Correlation Test for Pre-Whitening

Test for Normality Spearmans rank correlation test Standard Normal Homogeneity

test (SNHT)

Change point test (Pettitt test) Test for stability of variance (F-test) Test for stability of mean (t-test) Double Mass Analysis Method of Cumulative Residuals

(Ellipse test)

Pattern of observed time series data wasanalysed [5] in order to: (1) Identify the natureof the phenomenon represented by thesequence of observations, and (2) Predict futurevalues of the time series variables.

At the inception, Data were plotted for visualexamination in order to identify any abruptchanges in the time series. Testing of high andlow outliers was done using the equation

ynH sKyy andynL sKyy

whereyH, yL are high and low outlier thresholds in

log and y is the mean, n is the sample size, syis the standard deviation and Kn is theparameter given in Chow et al.(1988)[6], forsample sizes varying from 10 to 140.

The serial correlation coefficient verifies the

independence of a time series which in turnhelps to ensure that each of the data have anequal probability of occurrence. If a time seriesis completely random, the population auto-correlation function will be zero for all lagsother than zero. If all the data sets are perfectlycorrelated to each other then its value is unity.Sample serial correlation coefficients willdeviate slightly from zero only because ofsampling effects. In case of hydrologicalanalysis, it is usually sufficient to compute thefirst lag serial correlation coefficient, i.e. the

correlation between adjacent observations in atime series [1]. A confidence level of 95% wasused for calculations.

Presence of serial correlation may alsocomplicate the detection and evaluation oftrends in hydrological time series. When a dataset shows a drift towards higher (or lowervalues) over the period of record, the drift maybe an indication of an underlying change orlong term persistence. It could probably be thatthe data are dependent on some processes

which are serially correlated. Severalapproaches have been suggested for removingthe serial correlation from a data set prior to


8/71

ENGINEER 3

applying the non-parametric tests. One of themost common approaches is the Pre-Whiteningof the time series. The Pre-Whitening approachinvolves in the calculation of serial correlationand the removal of correlation if the calculatedserial correlation is found significant at a levelof 5% [3].

It is important to make sure that there is nocorrelation with the order in which the datahave been collected and with an increase or adecrease in the magnitude of those data. It isalso important that the selected testing periodsare of sufficient length for test to be reliable [1].In a study of rainfall trends in Sri Lanka [7],which chose both Mann-Kendall rank statisticand the Spearman rank statistic, concluded thatboth tests have similar power in detecting atrend. In the present work, Spearmans rank

correlation method is used to verify the absenceof trend at a significance level of 5%.

Standard Normal Homogeneity Test (SNHT) &Pettitt test were chosen to identify any suddenshifts in the mean of the data sets therebyenabling the identification of change points. Acritical probability level of 80% was chosen foracceptance of significant change points in thePettitt test whereas critical confidence level of90% was used in the SNHT [3].

Instability of the variance was tested to identifythe existence of a non-stationarity of the timeseries. Ratio of the variances of two split, non-overlapping, sub sets of time series wasselected as the test statistic. The region for test

statistic oft

F was taken as, F {v1, v2, 2.5%} < Ft

< F {v1, v2, 97.5%}; where, v1 = n1-1 (the numberof degrees of freedom for the numerator), v2 =n2-1 (the number of degrees of freedom for thedenominator), and n1, n2 equals the number ofdata in each sub set [1].

The t-test for stability of the mean wasconducted after carrying out the F-test usingsame two non overlapping time series subsets.The test statistic tt [1] was taken to be boundedas, t {v, 2.5%} < tt < F {v, 97.5%} where, v = (n1-1)+ (n2-1) (the degrees of freedom) including n1and n2 data in each sub set.

In order to identify the employability of theparametric test procedure, the time series wastested for normality by computing probability

of exceedence based on the Blom equation [8].Estimation of the data Xest with standardvariates was used to determine the variabilityof the quantile with 95% confidence limit.

Homogeneity of the time series was inspectedwith the method of cumulative residuals. Theestimated cumulative residuals and the ellipsethat relate with the probability level wereplotted against years to find whether thecumulative residuals fall within the ellipse [9]

Double mass analysis was performed usingplots of cumulative values of a station underinvestigation against the cumulative values ofthe particular station or cumulative values ofthe average of other stations over the sameperiod of time. To identify the RelativeConsistency of time series, detection of non-homogeneities was performed by identifyinginflection points in the double mass plot. Incase of significant changes, the annual values ofan earlier portion of the record were adjusted tobe consistent with the latter portion [10].

4. Results and DiscussionThe present work conducted for rainfall datasets of six stations indicated the variation ofresults from different statistical tests which arecommonly used for hydrological data testing.Annual rainfall data were plotted in order tofind the presence of any abrupt changes.During the considered period of 30 years,abrupt changes or any dubious data were notapparent for all six stations. Results ofstatistical tests pertaining to each station areshown in Table 2. Missing data were filled withthe use of single & multiple regression analysis.Computed best fit coefficients of determination(Table 3) were considered for data filling.Generation of missing data was carried outrelative to a common data period (Table 3) inwhich the data were assumed as homogeneousfor the computations.

The co-efficient of determination withregression analyses is relatively good for thestations at Pasyala (0.91) and Vincit (0.88),whereas other stations showed to haverelatively low values (Table 3). It was assumedthat the period considered for regression (i.e.01.05.1981 31.03.1982) is homogeneous.Selecting a homogeneous period is entirelydependant on the available metadata. In thisstudy, the considered homogeneous period forregression is less than one year. Therefore, itwas felt reasonable to assume that a minimumor no changes could occur to the station duringthe selected period. Based on these facts, theabove assumption could be treated as realistic.


9/71

ENGINEER 4

Minimum values of annual rainfall which werelower than low outlier were corrected with thelow outlier. High outliers showed highervalues in case of the maximum values of annualrainfall (Table 4). Tabular comparison ofannual rainfall showed that the minimumvalues should be filled with the low outlier

except for the minimum value of Pasyala.Annual rainfall comparisons with valuesshown in Figure 4, were used to identify anyabrupt changes. From this data set, annualrainfall at Pasyala station which has moreissues than other stations, and with t-testconfirmed change points, is selected to discussthe issues related to rainfall data. Statisticalanalysis & homogeneity test results of Pasyalaannual rainfall are shown in Figures 2 (a-f) and3 (a-f). In these Figures, graphs before DoubleMass analysis correction are shown by letters a,

c and e, whereas the letters b, d and f, show theresults after the Double Mass Analysis.

The presence of significant changes around1977 & 1985 at both Karasnagala and Pasyalacould be identified (Table 2) from the Pettitttest. If there are no Meta data then it is difficultto conclude whether the changes around 1977&1985 are due to a situation as a result ofclimatic change, or due to some other natural orman made changes to the environment duringthe period of record or systematic errors

associated with the recording of the data forAttanagalu Oya Basin. Even though the t-testresults confirm the presence of a change inPasyala around 1977 and 1985, the non-availability of Meta data prevented fromincorporating Double Mass corrections withsufficient confidence.

In order to identify the possibility of data use,an alternative option was considered. Sincethere is no strong evidence that the mean stateof rainfall in Sri Lanka has changed

significantly over the past decades, it wasassumed that the effect of climate change hadnot significantly affected the rainfall ofAttanagalu Oya. Accordingly the applicationof Double Mass curve for correction of changepoints was considered realistic and the samewas utilized for Pasyala in order to correct thechange which was present at 1985. Reductionof the trend could be observed after DoubleMass correction. The change at Pasyala in 1977was insignificant after carrying out doublemass correction for 1985 (Table 2, Figure 2).

Homogeneity test shows that the annualrainfall at Henerathgoda, Karasnagala and

Vincit is homogeneous at 85% non-exceedenceprobability level whereas for Halgahapitiya,Katunayake and Pasyala it is at 90% non-exceedence probability level. Homogeneity testresults, before & after double mass correctionsfor Pasyala are shown in Figure 3. It could beobserved that Pasyala rainfall data set is

homogeneous at 90% non-exceedenceprobability level even after the double masscorrection. Homogeneity test showed that theacceptable probability level of Pasyala data setis 90% since all residuals were found to bewithin the 90% probability ellipse after DoubleMass correction (Figure 3). As the t-test resultsdid not confirm the results of the Pettitt test,rest of the stations were not subjected tocorrection.

Correlogram shows that 1st lag serial correlation

for all datasets had fallen within the 95%confidence limit. Therefore, all these annualrainfall time series are with a satisfactory levelof randomness and independence. As a result,pre-whitening of annual rainfall time series wasnot necessary prior to performing statisticaltests. In order to select the need of parametricor non-parametric testing, normality testingwas carried out and it was identified that allstations follow the normal distribution patternexcept Katunayake which exceeds the 95%confidence limits.

It can be observed that a decreasing rainfallpattern is prevailing in Attanagalu Oya basin.Also the change of rainfall pattern around 1977& 1985 is common for some stations. Masscurve shows that the change in the slope is notsignificant. Therefore, it suggests that the datafrom each station are satisfying consistency. Assuch, likely reasons for the changes aroundabove years are mainly due to man madechanges to the environment and most probablydue to change of instruments.

Some of the homogeneity tests depend onmeta-data while the other tests are purelystatistical. The presence of a single significanttest result could be identified as weak evidenceof change. If more tests not similar to oneanother lead to significant test results, then itprovides stronger evidence of change. Carryingout similar tests which would provide multiple-significance is not an extra proof of change.However, application of more than one test tothe data may make interpretation of results

complex. Since the differences in assumptionsof the tests and the possible influence of changein the catchment condition, it is usually difficult


10/71

ENGINEER 5

to compare and in particular to combine theresults of different tests.

Meta data plays a major role when making firmconclusions with regards to data checking. InSri Lanka most gauging stations are maintainedwithout proper documentation of meta data

and it is common knowledge that the most ofthe hydrological and meteorological data arepoorly documented and quality flagged. Thisis a big challenge faced by hydrologists whenattempts are taken to analyze rainfall data. It isknown that for situations where no meta dataare available, hydrologists need to considerregional and global changes to rainfall duringthat period. However it is difficult to addressmicro climatic changes without meta data. Inmany stations of Sri Lanka it is not a difficulttask to obtain a 30 year long rainfall dataset.

These data are bound to be with missing dataperiods, non-homogeneity and otherinconsistencies. Hydrologists need to identifythe purpose of data and perform checks toensure reliability of results produced with suchdata. The present study presents an attempttaken to identify measures that can be takenwhen data checking is carried out in case of aSri Lankan situation.

5. Conclusions1. The present work using daily rainfall data

identified the variation of results fromdifferent statistical tests indicating thenecessity to compare and rationalize theoutputs prior to practical use.

2. In the Attanagalu Oya basin, the co-efficient of determination from regressionanalyses showed relatively good values forthe Pasyala and Vincit stations with valuesof 0.91 and 0.88 respectively.

3. A decreasing rainfall pattern prevails inAttanagalu Oya Basin and a significantchange in rainfall pattern could beidentified around 1977 & 1985 for somestations.

4. Tests conducted confirm that rainfall dataof Pasyala station has an inhomogeneity,and rectification could be carried out toachieve a 90% confidence level.

5. A significant change in Pasyala wasidentified around 1977 & 1985 by the Pettitttest and confirmed by t tests whereas thechanges at other stations were identifiedonly by one test or none. Pasyala wasidentified as the most significant stationout of those used for the study.

6. Testing clearly identified the need ofsupporting tests for confirmation ofindications made by a particular test, whileraising the issue of testing carried out withthe use of several tests.

7. Data checking enabled identification ofconfidence limits for data use thereby

providing the most important informationto assess the validity of using the resultinghydrologic outputs for reliable conclusions.

References

1. Dahmen, E.R. and Hall, M.J., 1990. Screening ofHydrological Data, Tests for Stationarity andRelative Consistency, ILRL, The Netherlands.

2. Uhlenbrook, S., 2006.Catchment Hydrology withSatellites, Models and Rubber boots. UNESCO-IHE, Delft, The Netherlands.

3.

Tu, M., 2006. Assessment of the effect of climatevariability and Land use change on theHydrology of the Meuse River Basin. A.ABalkema Publishers, The Netherlands

4. Guzman, J.A. and Chu, M.L., 2003. SPELL-Statstatistical analysis program. UniversidadIndustrial de Santander, Colombia.

5. Ma, L.C.A, 2003. Assessment of the Long termRainfall Runoff relation of the Geul Catchment.M.Sc. Thesis (HH 444), UNESCO-IHE, Delft,The Netherlands (Unpublished).

6. Chow, V.T., Maidment, D.R., Mays, L.W., 1988.Applied Hydrology. McGraw Hill bookCompany, Singapore.

7. Jayawardene H.K.W.I., Sonnadara D.U.J., andJayewardene D.R., 2005 Trends of Rainfall in SriLanka over the Last Century, Sri Lankan Journalof Physics, Vol.6 (2005) 7-17.

8. Cunnane, C., 1978. Unbiased plotting position a review journal of Hydrology 37:205-222

9. Perera, L.R.H., 2007. Detecting the Impacts ofClimate Variability on MeteorologicalParameters and Evaporation. M.Sc. Thesis(WSE - HWR - 07.06), UNESCO-IHE, Delft, TheNetherlands (Unpublished).

10. Dingman, S.L., 2002. Physical Hydrology, 2ndedition, Prentice Hall, New Jersey, U.S.A.

Acknowledgement

This research was supported by University ofMoratuwa Senate Research Grant Number 202.Encouragement given by the University ofMoratuwa and the Senate Research Committeeis gratefully acknowledged.


11/71

ENGINEER 6

Figure 1 - Study Area, Stream Network and Rain Gauging Stations

Vincit

Katunayake

Henerathgoda

Halgahapitiya

Pasyala

Karasnagala

Uruwal Oya

Dee-eli Oya

Attanagalu Oya


12/71

ENGINEER 7

Statistical Analyses Results at Pasyala (Correction in 1985)

Before Double Mass Corrections(1970 2001)

After Double Mass Corrections(1970 2001)

Annual RF- Pasyala

3230282624222018161412108642

3,500

3,000

2,500

2,000

1,500

Figure 2(a) Time Series Behaviour (before)

Annual corrected rainfall - Pasyala

3230282624222018161412108642

3,400

3,200

3,000

2,800

2,600

2,400

2,200

2,000

1,800

1,600

Figure 2 (b) Time Series Behaviour (after)

SNHT FORA SINGLESHIFT - (95%CL = 7.7)

30282624222018161412108642

7

6

5

4

3

2

1

0

Figure 2 (c) Results of SNHT (before)

SNHT FORASINGLESHIFT - (95%CL = 7.7)

30282624222018161412108642

7

6

5

4

3

2

1

0

Figure 2 (d) Results of SNHT (after)

Trend-1.034

Trend-2.196

197719771985

1985

Threshold limit

Threshold limit

1977

1977

19851985


13/71

ENGINEER 8

CHANGEPOINT TEST

3230282624222018161412108642

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

Figure 2 (e) Results of Pettitt Test (before)

CHANGEPOINT TEST

3230282624222018161412108642

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

Figure 2 (f) Results of Pettitt Test (after)

Figure 2 - Statistical Analyses Results at Pasyala (Correction in 1985)

Threshold limitThreshol

d limit

19771977

1985

1985


14/71

ENGINEER 9

Before Double Mass corrections(1970 2001)

After Double Mass corrections(1970 2001)

Homogeneity test - Annual RF Pasyala

80% probability

-2000

-1500

-1000

-500

0

500

1000

1500

2000

2500

0 5 10 15 20 25 30

Years

CumulativeresidualsEi(%)

Figure 3(a)Variation of Residuals with 80% limit

Homogeneity test after correction - Annual RF Pasy ala

80% probability

-2000

-1500

-1000

-500

0

500

1000

1500

0 5 10 15 20 25 30

Years


Figure 3(b)Variation of Residuals with 80% limit

Homogeneity test - Annual RF Pasyala

85% probability

-2000

-1500

-1000

-500

0

500

1000

1500

2000

2500

0 5 10 15 20 25 30

Years


Figure 3(c)Variation of Residuals with 85% limit

Homogeneity test after correction - Annual RF Pasy ala

85% probability

-2000

-1500

-1000

-500

0

500

1000

1500

2000

0 5 10 15 20 25 30Years


Figure 3(d)Variation of Residuals with 85% limit

Homogeneity test - Annual RF Pasyala90% probability

-2500

-2000

-1500

-1000

-500

0

500

1000

1500

2000

2500

0 5 10 15 20 25 30Years


Figure 3(e)Variation of Residuals with 90% limit

Homogeneity test after correction - Annual RF Pasyala90% probability

-2500

-2000

-1500

-1000

-500

0

500

1000

1500

2000

2500

0 5 10 15 20 25 30Years


Figure 3(f)Variation of Residuals with 90% limit

Figure 3 - Homogeneity Test for Pasyala (Correction in 1985)

90% limit90% limit

80% limit

85% limit85% limit

80% limit


15/71

ENGINEER 10

Figure 4 - Annual Rainfall Comparison at Selected Stations

Comparision of Annual RF at different Stations

0

1000

2000

3000

4000

5000

1970

1971

1972

1973

1974

1975

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Year

AnnualRF-mm

Halgaha Henarath Karasna Katunaya Pasyala Vincit

Comparision of Annual RF at different Stations

0

1000

2000

3000

4000

5000

1970

1971

1972

1973

1974

1975

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Year

AnnualRF-mm

Halgaha Henarath Karasna Katunaya Pasyala Vincit


16/71

ENGINEER 11

Table 1 - Rain Gauging Stations & Missing Data

Station Name Missing Duration No. of Missing Days

Halgahapitiya1 Sep 83 30 Sep 83 30

1 Jul 88 31 March 89 274

Karasnagala None None

Henarathgoda

1 Oct 77 - 31 Dec 77 92

1 Jun 80 30 Jun 80 30

1 Jun 82 30 Jun 82 30

1 Dec 82 31 Dec 82 31

1 Sep 83 30 Sep 83 30

Katunayaka 7 Apr 87 20 Apr 87 14

Pasyala

1 Oct 73 31 Dec 73 92

1 Sep 79 30 Sep 79 30

1 Apr 89 30 Apr 89 30

1 Jun 89 31 Jul 89 61

Vincit1 Apr 81 30 Apr 81 30

1 Apr 82 30 Apr 82 30

Table 2 - Summery of the Statistical Analysis for Annual Rainfall

Station &Duration

Year ofchange(Pettitt

test)

Pettitt testprobability

(80%)

Spearmanlinear trend

t-value(95%)

F-testF-value(95%)

t-testt-value(95%)

SNHTT0

(95%)

Halgahapitiya1970-2001 1985 0.6605 -0.891 1.301 1.232 1.492

Henerathgoda1970-2001

1975 0.3132 -0.437 1.504 1.054 1.388

Karasnagala1970-2001

19771985

0.80530.7917

-1.6161.1051.172

1.7341.847

4.285

Katunayake1970-2001

1996 0.3471 -0.03 -* -* -

Pasyala1970-2001

19771985

0.80530.8922

-2.1961.4841.32

2.5332.267

6.923

Pasyala correctedfor 1985

1970-2001

1977 0.7917 -1.034 1.676 2.523 4.575

Vincit1970-2001

1988 0.7057 -1.19 1.151 1.243 2.713

Values in parenthesis are thresholds (Confidence Level) for each test.-* Not enough data for the split record test


17/71

ENGINEER 12

Table 3 - Summary of the Regression Analyses -Best Fit Coefficient of Determination for MissingData Estimation

Data MissingStation

RF StationsConsidered for

Regression

Period Considered forRegression

Coefficient ofDetermination

HalgahapitiyaKarasnagala,Katunayake,Pasyala

01.05.1981 31.03.19820.62

HenerathgodaKarasnagala,Pasyala,Katunayake

01.05.1981 31.03.19820.49

Katunayake

Halgahapitiya,Karasnagala,Pasyala,Henerathgoda

01.05.1981 31.03.19820.62

Pasyala Karasnagala 01.05.1981 31.03.1982 0.91

Vincit Karasnagala 01.05.1981 31.03.1982 0.88

Table 4 - Tabular Comparison of Annual RF Prior to Correction for Outliers

RF Station Halgahapitiya Henerathgoda Karasnagala Katunayake Pasyala Vincit

Mean 2400.1 2342.4 2908.5 2078.1 2596.1 3251.6

Max 3412.2 3020.3 3795.0 3223.7 3632.5 4494.1

Min 1122.6 1427.7 1686.8 1130.8 1486.5 1902.1High

Outlier4329.6 3719.3 4655.3 3361.3 4482.8 5199.9

LowOutlier

1263.9 1427.7 1758.7 1240.8 1438.5 1966.3

Skewness -0.91 -0.75 -0.76 -0.53 -0.76 -0.32

Minimum values shown in bold font were corrected with low outlier values


18/71

13ENGINEER

ENGINEER - Vol. XXXXV, No. 02, pp. [13-22], 2012 The Institution of Engineers, Sri Lanka

Performance of Tall Buildings with and withoutTransfer Plates under Earthquake Loading

M T R Jayasinghe, D S Hettiarachchi and D S R T N GunawardenaAbstract: The scarcity of valuable lands has created a leap in the tall building construction allover the world. Due to the mixed use of these buildings as commercial, residential, parking, etc., ithas become essential to have different column grids in the same building to ensure the efficient useof space and materials. Use of transfer plates is one of the widely used methods to transfer gravityloads among different column grids. With lateral loads governing the design of tall buildings, it isessential to consider their behaviour against earthquake loads. This paper compares the behaviourof tall buildings with and without transfer plates against earthquake loading.

Keywords: Tall buildings, Response spectrum analysis, Earthquakes

1. Introduction

The availability of the usable urban lands isdeclining at a rapid rate due to development ofcities. This has led to the construction of tallbuildings with mixed development in thevicinity of the city centres since they canprovide large office areas close to places wellserved by public transport. These tall buildingsoften require integrated parking facilities aswell for which the space requirements aredifferent from that of residential and officespaces in the rest of the building.

To ensure the effective usage of the tallbuilding, it is important to have two differentcolumn grids in parking areas and theresidential/ office areas. However, to transferthe gravity loads from one set of columns to theother, a load transfer system is required.Transfer beam systems and the transfer platesystems are the most widely used methods forthis purpose. When planned with dueconsideration for lateral load behaviour, thestructures with transfer plates can provideadequate behaviour with respect to wind

induced acceleration (Balasuriya et.al. 2007).

With the use of transfer plates, massdistribution of the building could becomehighly irregular due to the thickness of thetransfer plate.

Thus, it is advisable to analyse the behaviour ofthese buildings against earthquake loads underdynamic conditions, though many engineersstill may use static analysis methods due to thereluctance to use dynamic analysis unless acomplete 3D model of the structure is available.

In this study, the dynamic behaviour of abuilding with a transfer plate was comparedwith a similar building without a transfer plate,

when subjected to earthquake loading using

finite element models (FEM) developed withSAP 2000 software. The parameters used for thebuildings are generally applicable to Sri Lankanconditions.

2. Methodology

The research was carried out using finiteelement models of two 35 storied apartmentbuildings of which one building consists of atransfer plate whereas the other building isuniform throughout.

Since the research was carried out in the SriLankan context, it was required to findearthquake data which are relevant to SriLankan conditions. Thus AS1170.4, AustralianStandard for Earthquake Loads was used toobtain earthquake loads for the buildings.

AS 1170.4 was chosen as the best representationof Sri Lankan conditions due to followingfactors;

x The type of earthquakes that Sri Lanka mayexperience is intra-plate type which wouldbe similar to that experienced by Australiaas well.

x Sri Lanka and Australia lie on the sametectonic plate; i.e. Indo-Australian plate.

Eng. (Prof). M T R Jayasinghe, B.Sc. Eng(Moratuwa), Ph.D. (Cambridge), C.Eng., MIE(SriLanka), Senior Professor, Department of CivilEngineering, University of Moratuwa

Eng. D S Hettiarachchi, B.Sc. Eng (Moratuwa),

Structural Engineer, Design Consortium Ltd.Eng. D S R T N Gunawardena, B.Sc. Eng(Moratuwa), Structural Engineer, DesignConsortium Ltd.


19/71

ENGINEER 14

x The types of tall buildings found inAustralia are quite similar to those thatwould come up in Sri Lanka now; i.e. in 35to 50 storey range with reinforced concreteframes.

Static and dynamic responses for the buildings

were obtained in accordance with AS1170.4 forcomparison purposes.

Figure 1aFEM of the Building with TransferPlate

Figure 1bFEM of the Building withoutTransfer Plate

3. Case Study

3.1 General

A rectangular layout with two axes ofsymmetry was chosen for the building models.The symmetry in turn would minimise effectsof shape and mass distribution in the transversedirection. The transfer plate was taken to be1.5m thick. Section sizes were chosen to matchthe structural requirements.

3.2 Finite Element Model

Two 3D models were developed with SAP2000, one with a transfer plate and 5 parkingfloors (with a different column arrangement)and the other without a transfer plate. These areshown in Figures1a and 1b respectively.

The transfer plate was modelled with thickplate (Mindlin/Reissner) formulation toinclude the transverse shear deformation.When modelling the transfer plate, it is ofutmost importance to make sure that all thecolumn nodes coincide with the nodes in thefinite element mesh of the transfer plate toensure proper shear transfer through nodes.This was achieved by meshing the transferplate manually.

However, the accuracy of the thick-plateformulation is more sensitive to large aspectratios and mesh distortion than the thin-plateformulation (SAP2000 Analysis Reference,2002). Thus the mesh was further refined tomake the aspect ratios of the mesh as close to1.0 as possible. The mesh achieved by thismethod is shown in Figure 2.

Figure 2 - Meshing of Transfer Plate

26.5 m

20.3 m

Transfer Plate Level


20/71

15ENGINEER

3.3 Analysis

The static analysis was carried out inaccordance with the Quasi-Static analysisprocedure in AS1170.4. Response SpectrumAnalysis was chosen as the dynamic method ofanalysis. The earthquake response spectrum in

AS1170.4 was used.

Peak Ground Acceleration at rock sites for a10% probability of exceedance in 50 year or 475year return period is around 0.026g forColombo (Peiris, 2008). Further, Sri Lanka hasnot experienced significant earthquake

activities in the recent past except for a fewtremors. Since the emphasis is on the variationpattern of earthquake loads rather than on themagnitude, the use of an exact value ofacceleration coefficient and its applicability to aparticular region would be of less significanceto this study. Therefore, purely to assess the

performance of tall buildings with transferplates slightly higher acceleration values in therange of 0.10g - 0.15g were selected.Acceleration values in this range are wellaccommodated in the Australian standard AS1170.4.

Figure 3 Acceleration Coefficient vs. Base Shear Building with Transfer Plate

Figure 4 Acceleration Coefficient vs. Base Shear Building without Transfer Plate


21/71

ENGINEER 16

Figures 3 and 4 show the variation of the staticbase shear and dynamic base shear (accordingto AS1170.4) with the acceleration coefficientfor the building with transfer plate and the onewithout a transfer plate, respectively. Theresulting base shear values in the X-directionare given for the response spectrum in the X-

direction (SpecX) and for the static earthquakeforces that are applied in the X-direction(EQX). SpecX includes a 30% component fromthe response spectrum in the Y-direction(SpecY) as per the requirements of AS1170.4.

It can be seen that the base shear valuesincrease linearly with the accelerationcoefficient when it is varied from 0.10 to 0.15.For comparison purposes, an accelerationcoefficient was chosen as 0.12 (See Appendix Afor calculations).

When Figures 3 and 4 are compared, it isevident that calculated static base shear islower than the dynamic base shear for bothbuildings. This gives an indication thatdynamic effects can be significant foracceleration values for the range considered.

3.4 Results

All results presented here are for theacceleration coefficient of 0.12. All base shear

and drift values are for the X-direction andunder the EQX and SpecX load cases.

The vertical distribution of static base shear isshown in Figure 5. (See Appendix A forcalculations)

It can be seen that the static base shear value ofthe building with transfer plate is greater thanthat for the building without transfer plate.This is due to the higher mass of the transferplate and the parking areas with a thicker floor

slab in comparison with the residential floorsin the building without transfer plate.

While the building without transfer platefollows a typical pattern of continuouslyincreasing base shear components with height,the building with transfer plate has a suddenincrease of the base shear component at Level5, i.e. the transfer plate level, and then a dropin the next level. This is due to the higher massconcentration at the transfer plate level.

However it should be noted that these valuesare mere fractions of base shear applicable ateach level. What is important is the aggregate

of these values at each level, i.e. the totaldistribution of base shear which is shown inFigure 6.

Figure 5 Vertical Distribution of Static BaseShear

Figure 6 Vertical Distribution of TotalStatic Base Shear

(kN)

(kN)


22/71

17ENGINEER

The total static base shear distribution showsthat although there is a sudden variation ofindividual storey forces at the transfer platelevel, it is only a small variation in the totalbase shear distribution.

The total base shear distribution shows that

the difference of the total shear applicable atthe transfer plate level and just above that isonly about 36kN which is not that significantin comparison with the total shear forceapplicable at that level which is about 3400kN.

Thus, it can be seen that the building withtransfer plate follows more or less the samepattern as the building without transfer platewhen it comes to the total static sheardistribution.

The vertical distribution of dynamic base shearand the total dynamic base shear are shown inFigures 7 and 8, respectively.

Figure 7 Vertical Distribution of DynamicBase Shear

The total dynamic shear distribution issignificant here since it clearly shows a suddenand sizable increase at the transfer plate level.The difference between the total base shearvalue just before the transfer plate and that atthe transfer plate level is about 1000 kN.Compared with the component of base shearat transfer plate level of 5700 kN, this is quite

significant and thus it is clearly seen that thedynamic effects could be significant for thebuilding with transfer plate.

Figure 8 Vertical Distribution of TotalDynamic Base Shear

The total dynamic shear distribution for thebuilding without transfer plate follows asimilar pattern to its total static sheardistribution. However, looking at the results, itis evident that dynamic base shear is criticalfor both the buildings.

Another critical factor that needs attention isthe top deflection and lateral drifts for a

building under lateral forces such asearthquakes. The top deflections and thelateral drifts are given in Table 1.

It is interesting to see what an optimum levelwould be for the transfer plate to bepositioned. For this, the dynamic analysis wascarried out for the building with transfer plate,keeping the transfer plate at different floors.As described earlier the dynamic effects wereconsidered with an acceleration coefficient of0.12, and the resulting shear values in the X-

direction are given here against the responsespectrum case SpecX.

(kN)

(kN)


23/71

ENGINEER 18

Table 1 Top Deflection & Drift Index with Equivalent Static and Dynamic Analysis

The critical parameters taken into account herewere the total dynamic base shear componentat the transfer plate level and the dynamic base

shear. The results are shown in Table 2.

Table 2 Variation of Dynamic Base Shearwith Transfer Plate Level

It is evident that there is an increase in the baseshear as the transfer plate location is movedup. This indicates that it is advisable to havethe transfer plate at a lower level. If it isessential to have it at a higher level, thenadequate attention requires to be paid for theductility needs of the columns andconnections.

4. Conclusions

It should be noted that previous studies haveindicated that the buildings with transferplates can perform better under wind inducedaccelerations. Therefore, it is recommended toconsider transfer plates as a member withfavourable structural actions since it is anessential component of mixed developments.

Buildings without transfer plates performbetter against earthquakes. This appears to bedue to the regularity of its mass distribution asagainst that of buildings with transfer plates. Itwas observed that the building with a transfer

plate could indicate higher top deflection and agreater base shear.

It was also clear through the findings of theresearch that a dynamic analysis needs to becarried out for buildings with irregular massdistributions such as those with transfer plates,since dynamic effects may prove to be morecritical. This would need the aid of a 3D modelof the structure.

It is advisable to consider transfer plateswherever they are needed. They should neverbe considered as a problematic feature thoughthey may have some negative effects on the

actual earthquake behaviour. It can be clearlyseen that the presence of transfer plates cangive rise to significant shear values at transferplates. Thus, the columns located immediatelyunder the transfer plates are the areas that maysuffer the worst damage. Any formation ofplastic hinges and associated crushing ofconcrete may have severe adverse effects.Therefore, it is advisable to pay very specialattention to improve confinement to thecolumns as much as possible using thedetailing practices (such as links and hooks

details for reinforced concrete columns)recommended for enhanced earthquakeresistance.

References

1. Abayakoon, S. B. S. (1995) Seismic Risk Analysisof SriLanka,Journal of the Geological Society of SriLanka, 6:65-72

2. AS1170.1:1989, Minimum Design Loads onStructures- Part 1: Dead and live loads and load

combinations, Standards Australia, New SouthWales.

Transfer Plate Building(Transfer Plate at 5th Floor Level)

Building Without Transfer Plate

Deflection(mm)

DriftIndex

Deflection(mm)

DriftIndex

Static a=0.12 44.8 0.00231 41.5 0.00207

Dynamic

a=0.10 28.3 0.00175 27.8 0.00164a=0.11 31.1 0.00192 30.6 0.00180

a=0.12 32.9 0.00210 33.4 0.00196

a=0.13 36.8 0.00228 36.2 0.00213

a=0.14 39.6 0.00245 39.0 0.00230

a=0.15 42.5 0.00262 41.8 0.00246

Transfer PlateLevel

Base Shear(kN)

ShearComponentat TransferPlate Level

Second Floor 5314 351

Third Floor 5710 606

Fourth Floor 6060 854

Fifth Floor 6261 1014


24/71

19ENGINEER

3. AS1170.4:1993, Minimum Design Loads on

Structures- Part 4: Earthquake loads, StandardsAustralia, New South Wales.

4. Balasuriya S.S., Bandara K.M.K., Ekanayake S.D.,Jayasinghe M.T.R. (2007), The Influence ofTransfer Plates on the Lateral Behaviour ofApartment Buildings, Engineer, Journal ofInstitution of Engineers, Sri Lanka , Volume XXXX,

No: 4, October, pp 22-305. CHOPRA, A.K. (2001), Dynamics of Structures

Theory and Application to Earthquake Engineering.2nd edition Prentice Hall

6. COMARTIN C. D., NIEWIAROWSKI R.W.,ROJAHN C. (1996), Seismic Evaluation andRetrofit of Concrete Buildings. Volume 1, AppliedTechnology Council, USA

7. Peiris L.M.N., (2008), Seismic HazardAssessment of Sri Lanka and Seismic Risk inColombo, Risk Management Solutions, London,UK

8. SAP2000 Analysis Reference, (2002), Computersand Structures, Inc., Berkeley, California, USA,Version 8.0

9. SMITH, B. S., COULL, A.(1991), TallBuildingStructures, John Wiley, USA

10. WILKINSON, S. (1997).Seismicity and GroundMotion. The Structural Branch of EngineersAustralia

11. WILSON J., HUTCHINSON G. (1997), DynamicCharacteristics and Response. The StructuralBranch of Engineers Australia

12. WILSON, L. EDWARD.(1995),ThreeDimensional Static and Dynamic Analysis ofStructures, Computers and structures, Inc., USA


25/71

ENGINEER 20

Appendix A

Earthquake Load Calculations

All calculations here are made in accordance with AS 1170.4 guidelines.

The depth of bed rock in Sri Lanka can be assumed to be in a range of 15m to 20m.Thus site factor = 1.0 (Table 2.4 a)

Building is classified under Type II (Clause 2.2.3 b)

Acceleration coefficients are taken in a range from 0.10 to 0.15.For the specimen calculations, the acceleration coefficient has been taken as 0.12.

Importance factor, I = 1.00 (Table 2.5)

Static Analysis

For the building with the transfer plate;

(Clause 6.2.4)

(Clause 6.2.3)

Rf= 6.0

Kd = 5.0 (Table 6.2.6 b)

(Clause 6.2.2)

V0.01 Gg

Fx = CvxV (Clause 6.3)

Cvx = (Ggxhxk)(Ggihik)

The calculations have been carried out to find V (base shear) using a spreadsheet with the aid of the

equations given above.


26/71

21ENGINEER

Resulting Base shear (V) = 3390 kN

LevelGgx

(kN)hx

(m)hxk Ggx.hxk Cvx

Fx(kN)

35 4399.86 106.3 2382.40 10482235 0.0594 201.4

34 5395.62 103.3 2271.42 12255700 0.0695 235.4

33 5395.62 100.3 2162.56 11668344 0.0661 224.232 5395.62 97.3 2055.85 11092581 0.0629 213.1

31 5395.62 94.3 1951.31 10528528 0.0597 202.3

30 5395.62 91.3 1848.96 9976308 0.0565 191.7

29 5395.62 88.3 1748.84 9436051 0.0535 181.3

28 5395.62 85.3 1650.95 8907890 0.0505 171.1

27 5395.62 82.3 1555.33 8391965 0.0476 161.2

26 5395.62 79.3 1462.00 7888423 0.0447 151.5

25 5395.62 76.3 1371.00 7397419 0.0419 142.1

24 5395.62 73.3 1282.36 6919116 0.0392 132.9

23 5395.62 70.3 1196.10 6453684 0.0366 124.0

22 5395.62 67.3 1112.25 6001304 0.0340 115.3

21 5395.62 64.3 1030.87 5562168 0.0315 106.9

20 5395.62 61.3 951.97 5136479 0.0291 98.7

19 5395.62 58.3 875.61 4724452 0.0268 90.8

18 5395.62 55.3 801.82 4326319 0.0245 83.1

17 5395.62 52.3 730.65 3942326 0.0223 75.7

16 5395.62 49.3 662.16 3572740 0.0202 68.6

15 5395.62 46.3 596.38 3217848 0.0182 61.8

14 5395.62 43.3 533.39 2877960 0.0163 55.3

13 5395.62 40.3 473.24 2553417 0.0145 49.1

12 5395.62 37.3 416.00 2244590 0.0127 43.1

11 5395.62 34.3 361.75 1951891 0.0111 37.5

10 5395.62 31.3 310.58 1675778 0.0095 32.2

9 5395.62 28.3 262.58 1416766 0.0080 27.2

8 5395.62 25.3 217.85 1175443 0.0067 22.6

7 5395.62 22.3 176.53 952482 0.0054 18.3

6 5395.62 19.3 138.76 748676 0.0042 14.4

5 21495.38 16.3 104.71 2250788 0.0128 43.2

4 5757.06 12 62.86 361866 0.0021 7.0

3 5369.87 9 38.92 208984 0.0012 4.0

2 5369.87 6 19.80 106335 0.0006 2.0

1 4476.35 3 6.24 27925 0.0002 0.5

Total 203341.37 176434781 3389.6


27/71

ENGINEER 22

For the building without the transfer plate;

Same values are used as for the building without the transfer plate. Only value that will differ is thedead weight of the building.

Resulting Base shear (V) = 3115 kN

LevelGgx

(kN)hx

(m)hxk Ggx.hxk Cvx

Fx(kN)

35 4399.86 105 2334.05 10269486 0.06065 188.9

34 5395.62 102 2223.98 11999763 0.07087 220.7

33 5395.62 99 2116.05 11417416 0.06743 210.0

32 5395.62 96 2010.28 10846713 0.06406 199.5

31 5395.62 93 1906.69 10287772 0.06076 189.2

30 5395.62 90 1805.30 9740720 0.05753 179.2

29 5395.62 87 1706.14 9205688 0.05437 169.3

28 5395.62 84 1609.23 8682811 0.05128 159.7

27 5395.62 81 1514.61 8172234 0.04826 150.3

26 5395.62 78 1422.28 7674106 0.04532 141.2

25 5395.62 75 1332.30 7188586 0.04245 132.2

24 5395.62 72 1244.68 6715838 0.03966 123.5

23 5395.62 69 1159.47 6256039 0.03695 115.1

22 5395.62 66 1076.68 5809374 0.03431 106.9

21 5395.62 63 996.37 5376039 0.03175 98.9

20 5395.62 60 918.57 4956243 0.02927 91.2

19 5395.62 57 843.32 4550208 0.02687 83.7

18 5395.62 54 770.66 4158171 0.02456 76.5

17 5395.62 51 700.64 3780387 0.02233 69.516 5395.62 48 633.32 3417132 0.02018 62.9

15 5395.62 45 568.74 3068702 0.01812 56.4

14 5395.62 42 506.97 2735420 0.01615 50.3

13 5395.62 39 448.07 2417639 0.01428 44.5

12 5395.62 36 392.12 2115748 0.01250 38.9

11 5395.62 33 339.20 1830177 0.01081 33.7

10 5395.62 30 289.38 1561407 0.00922 28.7

9 5395.62 27 242.79 1309984 0.00774 24.1

8 5395.62 24 199.52 1076528 0.00636 19.8

7 5395.62 21 159.72 861762 0.00509 15.96 5395.62 18 123.53 666542 0.00394 12.3

5 5395.62 15 91.17 491903 0.00291 9.0

4 5395.62 12 62.86 339148 0.00200 6.2

3 5395.62 9 38.92 209986 0.00124 3.9

2 5395.62 6 19.80 106845 0.00063 2.0

1 4399.86 3 6.24 27448 0.00016 0.5

Total 186855.18 169323966 3114.8


28/71

23ENGINEER

ENGINEER - Vol. XXXXV, No. 02, pp. [23-29], 2012 The Institution of Engineers, Sri Lanka

Flood Inundation Mapping along the Lower Reach ofKelani River Basin under the Impact of Climatic

Change

M. M. G. T. De Silva, S. B. Weerakoon, Srikantha Herath, U. R. Ratnayake andSarith Mahanama

Abstract: The downstream low lying region of the Kelani River including the Colombo suburbs,experience severe inundation due to heavy rainfalls in the upper basin of the Kelani River. Occurrenceof heavy rainfalls is expected to be more frequent in the tropics with the impact of climatic change(IPCC, 2007). Therefore, understanding future rainfall intensity in the river basin and inundation inthe low lying region along the lower reach of the Kelani River is extremely important as this is aregion with a high population density and economic activities in the suburbs of the capital.

The present study analyses the potential extreme rainfalls and resulting flood inundation along the

lower Kelani River. Coarse grid atmospheric parameters provided by Global Climate Model (GCM)models for A2 (high emission scenario) and B2 (low emission scenario) scenarios of IntergovernmentalPanel on Climate Change (IPCC, 2007) were downscaled to local scale by applying StatisticalDownscaling Model (SDSM). Flood discharge and inundation along the Kelani River reach belowHanwella were analyzed by applying two-dimensional flood simulation model (FLO-2D). Inflow tothe model at Hanwella, is estimated by the Hydrologic Engineering Center Hydrologic ModelingSystem (HEC-HMS) model under future extreme rainfall events. Areas vulnerable to inundationunder the above climatic change scenarios are presented.

Keywords: Extreme rainfall, Flood inundation, Climatic change

1. Introduction

Climate change is defined as statisticallysignificant variation in either mean state of theclimate or in its variability, persisting for anextended period (typically decades or longer).It may be due to natural internal processes orexternal forcing or to persistent anthropogenicchanges in the composition of the atmosphereor in land use [15]. Climate change directlyaffects precipitation and temperature. Highintense rainfalls, frequent and prolonged

droughts are some examples. It has beenpredicted that the implications of climaticchange on Sri Lanka are variations in rainfallpatterns, sea level rise, and temperature [3].

Two monsoon systems, the Southwest (May-Sep) and the Northeast (Nov-Feb) anddevelopment of extreme low pressureconditions in the Bay of Bengal also have directimpacts on the rainfall patterns in Sri Lanka.Anomalously high seasonal precipitationtypically associated with La Nina phenomenonand cyclonic storms which originate from theBay of Bengal are usually the main reasons fordevastating floods in the island.

This study discussed the frequencies of highintensity precipitation with their inundationextents in the Kelani River basin, producedunder IPCC Special Report on EmissionsScenarios (SRES) A2 and B2 scenarios [16]. BothA2 and B2 emphasize on rapidly growing self-reliant nations while B2 accounts for moreecologically friendly growth. If extremeprecipitation events are becoming morefrequent in the 21st century as manifested by

Eng. (Ms) M. M. G. T. De Silva AMIE(Sri Lanka), BScEng(Peradeniya), MPhil candidate, Dept. of Civil Engineering., Uni.of Peradeniya.

Eng. (Prof.) S. B. Weerakoon, FIE(Sri Lanka), Int. PE SL, C.Eng, BScEng (Peradeniya), MEng, PhD (Uni. of Tokyo),Professor of Civil Engineering, Dept. of Civil Engineering, Uni.of Peradeniya

Prof. Srikantha Herath, BScEng (Peradeniya), MEng (AIT,Thailand), DEng (Uni. of Tokyo), Senior Academic ProgrammeOfficer, UNU-ISP, Tokyo, Japan

Eng. (Dr.) U. R. Ratnayake, AMIE(Sri Lanka), C. Eng,BScEng (Peradeniya), MEng, DEng (AIT, Thailand), SeniorLecturer, Dept. of Civil Engineering, Uni. of Peradeniya.

Dr. Sarith Mahanama, BScEng (Peradeniya), PhD (Uni. ofHong Kong), Research Scientist, NASA Goddard Space FlightCenter, Greenbelt, Mar land, USA


29/71

ENGINEER 24

ever changing global climate system, the KelaniRiver basin is one of the most vulnerable riverbasins for floods and costly flood damagessince the river flows through the commercialcapital [12].

2. Study Area

The Kelani River basin is located in betweenNorthern latitudes 6 47' to 7 05' and Easternlongitudes 79 52' to 80 13'. The basin area isabout 2230 km2 with two distinct types, as theupper basin is mountainous and the lowerbasin which is below Hanwella, has plainfeatures.

The basin receives about 2400 mm of averageannual rainfall and carries a peak flow of about800-1500 m3/s during monsoon seasons, to the

Indian Ocean. The flood level gauge atNagalagam Street (Colombo) defines theseverity of the flood as; minor floods (levelbetween 5 ft / 1.5 m and 7 ft / 2.1 m), majorfloods (level is between 7 ft / 2.1 m and 9 ft /2.7 m), and severe flood (level exceeds 9 ft / 2.7m) [5].

With climatic change impacts, a properunderstanding of occurrence of rainfall, floodforecasting and inundation mapping in KelaniRiver basin are very important due to the

occurrence of frequent floods and the resultingsocial and economic loss. According to theDisaster Management Centre (DMC), morethan 38,000 families living in flood plains ofKelani River were affected during the 2008flood [18], while more than 78,000 families wereaffected during the 2010 flood [17]. Moreover,Irrigation Department records show that therewere two consecutive severe floods whichoccurred during the year 2008.

3. Methodology

3.1 Preparation of data

Topographic data of the basin were obtainedfrom Shuttle Radar Topographic Mission(SRTM) data as Digital Elevation Model (DEM).The SRTM-DEM data, with a horizontalresolution of approximately 90 m near theequator and a vertical resolution of 1 m,constitutes the finest resolution and mostaccurate topographic data available for most ofthe globe [11]. According to a study carried outin the Ruhuna basin, the most useful data werethe SRTM DEM at 90m resolution. The slope

map generated using the SRTM DEM was veryuseful to identify low lying areas [9]. SRTMDEM has also been used in the hydrologicalmodeling of upper Kotmale basin in SriLanka [6].

Inflow hydrograph at Hanwella was generated

by HEC-HMS and the rainfall of the lowerbasin was generated by Statistical DownScalingModel (SDSM). Land use data was from theDepartment of Survey, Sri Lanka at 1:50,000scale.

3.2 Rainfall Modeling

Statistical Downscaling Model (SDSM) which isdesigned to downscale GCM data into regionallevel, was applied to downscale precipitationfor the basin up to 2099 under A2 and B2

scenarios published by IPCC. The model wascalibrated for the period from 1961 to 1975 andvalidated from 1976 to 1990. SDSM uses linearregression techniques between predictor(observed large scale climate fields) andpredictand (local observed meteorologicalvariables) to produce multiple realizations ofsynthetic daily weather sequences. Thepredictor variables provide daily informationabout large scale atmosphere condition, whilethe predictand describes the condition at thesite level [2].

Accordingly in order to ensure that thedownscaling method in reproducing the meanand variability of observed variable in thebasin, an uncertainty analysis was carried out.Table 1 shows the statistical parameters ofannual maximum daily rainfall for simulatedand observed data during the periods ofcalibration (1961-1975) and validation (1976-1990).

Table 1 - Statistical parameters affect to

uncertainty

Parameter1st

QuartileMedian

3rdQuartile

Variance

Obs 61-75 119 149 192 2101

A2 61-75 118 164 186 1723

B2 -61-75 111 153 183 1609

Obs 76-90 115 151 174 2136

A2 76-90 117 148 163 2089

B2 76-90 119 132 152 2199

According to the statistical parameters, 1st & 3rdquartiles, and median lie in the same range asobserved while the variance shows some


30/71

25ENGINEER

deviation from the observed values during thecalibration period. All the statistical parametersfrom simulations show a satisfactory agreementwith observed values during the validationperiod.

Figures 1(a) and 1(b) show the uncertainty of

forecasted annual maximum daily rainfall withobserved values for calibration and validationperiods respectively, for lower basin. Thesimulation results for validation period underA2 and B2 scenarios show good correlationwith observed data

Figure 1 Box plots to compare observed anddownscaled results for annual maximum dailyrainfall

Forecasted rainfall of the upper basin by SDSMwas used in HEC-HMS model to generate theflow at Hanwella. HEC-HMS, developed by theUS Army Corps of Engineers, is designed tosimulate the precipitationrunoff processes ofwatershed systems. Numerous past studieshave shown this model to provide accurate anduseful results in flood related studies [8].

The upper basin is about 1740 km2 and the timeof concentration from upper basin which isfrom the Samanala mountain to Hanwella isabout 2 days and at the edge of the lower basinit is about 2.5 days. This could be due tosubstantial losses because of infiltration anddepression storage, within the upper basin

which has a vegetation cover of about 70%.Frequency analysis of return period was carriedout for 3 day total rainfall of the upper basin byusing forecasted data from 2020 to 2099.

The lower basin has an area of about 500 km2with plain features and ample amount of built-

up areas. The time of concentration of lowerbasin at the edge of the basin is around 12hours. Therefore, one day rainfall in the lowerbasin can cause severe floods. Therefore, thedaily rainfall forecasted from 2020 to 2099 wasused for the frequency analysis within thelower basin.

Accordingly the flood and inundation analysisin the lower basin was carried out using 3 dayrainfall events of 50 year and 100 year returnperiods of upper basin and daily rainfall events

of 50 year and 100 year return periods of lowerbasin. Table 2 and 3 illustrate the summary ofresults of frequency analysis for upper andlower basins respectively by applying theGumbel distribution [4].

Table 2 3 day rainfall for upper basin

Return Period/ (yr)

3 day rainfall / (mm)

A 2 B 2

50 415 393

100 456 432

Table 3 Daily rainfall for lower basin

Return Period/ (yr)

Daily rainfall / (mm)

A 2 B 2

50 244 227

100 268 248

3.3 2D flow modeling

Two-dimensional flood simulation model, FLO-2D was utilized to analyse and map inundationareas. FLO-2D is two-dimensional flood routingmodel that distributes a flood hydrograph overa system of square grid elements. FLO-2Dnumerically routes a flood hydrograph whilepredicting the area of inundation andsimulating flood wave attenuation [13].

The FLO-2D system consists of processorprograms to facilitate graphical editing andmapping that simulate channel and floodplain

details. The Grid Developer System (GDS)generates a grid system that represents thetopography as a series of small tiles. The FLO-2D model has components for rainfall, channel

(a) Calibration period

(b) Validation period


31/71

ENGINEER 26

flow, overland flow, infiltration, levees andother physical features [14].

FLO-2D moves the flood volume around on aseries of tiles for overland flow or throughstream segments for channel routing. Floodwave progression over the flow domain is

controlled by topography and resistance toflow. Flood routing in two dimensions isaccomplished through a numerical integrationof the continuity equation and the equations ofmotion [14].

Flow depth Depth-averaged velocity Excess rainfall intensity Friction slope Bed slope

The differential form of the continuity andmomentum equations in the FLO-2D model issolved with a central, finite differencenumerical scheme as one dimensional flow inthe channel and two dimensional flow in theflood plain.

3.4 Model calibration

The GDS was used to generate 250 m x 250 mgrids covering the river basin. The topographyof the basin area was determined by using thedownloaded DEM from SRTM. Manningscoefficients over the basin were allocatedaccording to the land use patterns. The modelwas calibrated for the extreme event whichoccurred in November 2005, by comparing theflow at Nagalagam Street gauging station. The

validation was carried out for the discharge atNagalagam Street by using the events whichoccurred in April 2008, May 2008 and May2010. Figure 2 and 3 show the observed andsimulated hydrographs for 2005 flood eventand 2010 flood event respectively. The observeddata during 2010 flood event, the rising andfalling limbs show some oscillations and whichcould be due to external factors.

Normalized Objective Function , Nash-Sutcliffe efficiency

and percentage bias

given in equations (3) to (5) respectively,were used to evaluate the Goodness of Fitbetween the simulated and observed flows.

Figure 2 Time series of observed andsimulated flow at Nagalagam Street during2005 flood

Figure 3 Time series of observed andsimulated flow at Nagalagam Street during2010 flood

Where,

Observed discharge

Simulated discharge

Number of data points Mean of the observed dischargeTable 4 shows the Goodness of Fit of thesimulation results for selected flood events.

Table 4 Goodness of Fit

Event / (%)Nov. 2005 0.27 0.84 21

April 2008 0.33 0.65 25

May 2008 0.26 0.59 18

May 2010 0.32 0.68 23


32/71

27ENGINEER

The rsimugaugThecomflood

ALOinunManonlyThepubliandto isame4(c) iand s

Theconsiclassisucce([1],[follo

(cells

preditheexactvalueoverl

valMayrang

(a)

esults indiclating disching station.

simulatedared withevents in

/PALSARation mapgement Ceevents witbserved in

shed by thehaded regiundation.event is sh

llustrates thimulated in

inundationdering thefied correctssfully util] and [10])s.

and aor pixels)

cted as inueasure of

ly overlapis 100%

aps with obue is 71% fo2008 event.with comp

- Observed

te that thearge at the

inundatiothe observe

ay 2008 a

HH datas publishednter (DMC)

inundatioundation eDMC are sns indicatehe simulatwn in Figu

e combinatindation ex

extents wfraction

ly by theized in sand the e

e set of dbserved to

ndated resFit. If thewith the oand if noerved extenr May 2010These valueared to past

odel is capNagalagam

n extentsd extents fd May 201

set andby the Drespectivelmaps ava

tents in theown in Figuthe area sued result fre 4(b) andn of the obents.

re comparof the dodel, whic

me past suation use

main sub rbe inundate

ectively ansimulatedserved extsimulated

ts value isevent and 6s are in accestudies.

able ofStreet

wereor the

from

floodisastery (theilable).

basinre 4(a)jectedor theFigureserved

ed byomainh wastudies

is as

egionsd and

isxtentsntsextent0. The

9% forptable

Figure 4 in May 201The Gooddischargeand the inshow gooAccordinglfuture fore

4. Re

Flood inun50 year anand B2 sce2D model.

(b) - Sim

(c) Cobserved

Inundation0

ness of Ft Nagalagandation ex

agreemey, the calibrasts.

sults and

dation exte100 year r

narios were

lated

mbinationand simula

extents due

it parametStreet ga

tents of thet with thted model

iscussio

ts due to teturn periocomputed

ofed

to floodin

ers of theging statiosimulations

observed.as used fo

s

e rainfall os under A2y the FLO-


33/71

ENGINEER 28

Figure 5 rainfall ofbasin andperiod in t

(a) A2

(b) B

(a) A2 S

Inundatio50 year retdaily raine lower bas

cenario

Scenario

cenario

extents drn period i

fall of 50in

e to 3 daythe upper

year return

FigurrainfbasinperioFigurrainfinunSomeKoloDivisidentinunFigurto rai

to thexpaJayainunto eaTable

Tabl

Inude

1.5

2.0

>

TheidentiDepefloodpro

Journal - April 2012

Documents

Transcript of Journal - April 2012