# Introduction

lood affected many of the engineering structures such as bridge, embankment, barrage, levees, reservoirs, etc. while designing the proper safeguards must be made for the safe passage of the maximum expected flood. The structure must be sound not only for its own safety but also for the life and property which might be in danger by its failure. The valley then becomes ''flooded". A flood is commonly considered to be an unusually high stage of a river. It is often the stage at which the stream channel becomes filled and starts overflowing its banks. In Webster's new international dictionary, a ''flood" is "a great flow of water especially, a body of water, rising, swelling and overflowing and not usually thus covered a deluge, a freshet, an inundation". A flood problem in Bangladesh is gigantic and becomes more complicated with the passage of time. Every year a large area of this country is more or less affected by the flood. For the unique geographical situation of Bangladesh flood cannot be protected. But damages caused by the flood are lowered by the proper and timely forecasting about flood. Most of the flood studies are made for the flood controlling. Here flood forecasting system is very poor. So attempts have been taken to develop appropriate flood forecasting model. Flood is a serious problem in our country. Every year a large number of hydraulic structures, crops and properties are damaged by the flood.

Author ? ? ?: Department of Civil Engineering, Rajshahi University of Engineering & Technology, Rajshahi, Bangladesh. e-mails: mnz_ruet@yahoo.com, nirjhor.ruet.ce@gmail.com, sharid.shahnewaz@gmail.com

The necessity of the study is given below: 1. To ensure the safety of the hydraulic structure (like-Barrage, levees etc.).

2. To take measure for the safety of the crops and properties of the adjacent land. Where, Y is a dependent variable and is the water level at the forecasting station at time, t(MSL), X is a independent variable and is the water level at the base station at time (t-T) with T as the travel time between this station and the forecasting station, a, b are multiple correlations co-efficient.

It should be noted that the advanced time for the forecast at the forecasting station is the least of the travel times. The procedures involved in the development of the model are:

? Identifying flood forecasting stations;

? Identifying potential base stations;

? Preparation of data base;

? Estimation of travel time; and In this method the travel time is considered as the time difference between the peak water level of the base station and forecasting station.
?
ii. Mutreja's Method This method consists in collecting the water level data of flood at base station for the Nth hour and at the forecasting station for the (N+T)-th hour in a tabular form. By taking different values of T different data tables are prepared such that each data table corresponding to one of assumed T. To compute the cross correlation of the water level data of these two stations at different legs the cross correlation of the data on each table is computed. The value of T corresponding to the data table the maximum correlation is travel time of the reach.


# c) Necessary Data

In this study the following three types of data have been collected:

? Daily water level data ? Daily discharge data


# ? Danger level data

All these data used in this study were collected from the surface water hydrology -II of Bangladesh Water Development Board (BWDB).


# i. Discharge Level Data

The BWDB in this the primary source of discharge data. The mean daily discharge data during a water year is published by hydrology directorate of the BWDB. Data sheet for daily discharge also contains the annual maximum and minimum discharge. Daily discharge data of Hardinge Bridge and GoalundoTransi station of the river Padma are collected for the purpose of this study.


# ii. Water Level Data

The BWDB is also the primary source of water level data. The water level of the river is measured 5 times a day, at 6.00, 9.00, 12.00, 15.00 and 18.00 hour on stuff gauges. The mean of the 5 measurements is published as mean daily water level by the hydrology directorate of BWDB of Dhaka. Data sheets containing the mean daily water level at Hardinge Bridge and Goalundo Transi station of the river Padma during a water year (July to October) are given in Appendix-A. The data sheet also contains the annual maximum water level data of the monsoon period has been used for this study.

iii. Danger Level Danger level data for two stations Hardinge Bridge and GoalundoTransi has been collected from the BWDB, Dhaka. The danger level of Padma at the selected river at the selected stations is given below: 


# Results and Discussion


# a) Travel Time From Base to Forecasting Station

The travel time from the base station to the forecasting station is given in the following table calculated by two separate methods. The following graphs show the correlation of Nth hour stage of base station with (N+T)-th hour stage of forecasting station. Putting the value of daily water level data at X axis (base station) and daily water level data at Y axis (forecasting station) after the travel time T and finally get a linear equation. 


# Conclusions

The following conclusions can be drawn from the above analysis:

? The accepted value of travel time from Hardinge Bridge to Goalundo Transi is 2 days. 
3123![Improvement of the existing channel section for the computed discharge. Within the Padma basin in Bangladesh the important tributaries are the Punarbhaba and Mahananda from the left which drain Panchagar, Dinajpur and Chapainawabgong districts. They enter southeast zone of Bangladesh covering greater districts of Kustia, Jessore, Faridpur, Khulna and Barishal served by the important distributaries of the Padmaviz the Mathavanga and the Arialkhan are the most important. The lower gigantic delta in Bangladesh has a large area subjected to the tides from the Bay of Bengal. Catchment area of this basin is 53706 km2. The general objectives of this study are given below: Determine the travel time of flood wave from base station to forecasting station. To determine the correlation between the N-th hour stage of base station and (N+T)-th hour stage of forecasting station. To develop a flood forecasting model for the river Padma. II. Background Rahman, M.M., Goel, N.K. and Arya, D.S. (2012) developed flood forecasting system by using MIKE11 river-modeling software modules rainfall-runoff (RR) [or Nedbor-Afstromnings model (NAM)], hydrodynamic (HD), and flood forecasting (FF) for the Jamuneswari river catchment of the northwestern part of Bangladesh. A Chowdhury, M. R. , and Ward, N. (2004) worked on Hydro-meteorological variabilty in the greater Ganges-Brahmaputra-Meghna basins. Rahman, M. M. ,Arya, D. S. , Goel, N. K. , and Dhamy, A. P. (2011a) have carried out their research for design flow and stage computation in the Teesta river.The statistical model uses the multiple correlation technique. Basically, only gauge of base stations and forecasting stations are utilized in different forms in developing these models. III. Methodology Statistical method has been used in this paper to develop the forecasting model. a) Outline of Statistical Method Daily water level of base station X has been used to develop a multiple correlation model for predicting water level Y, at the forecasting station. The model is Y = a + bX.](image-2.png "3 . 1 . 2 . 3 .")
![Development of flood forecasting model b) Estimation of Travel Time To estimate the travel time the approaches used in this paper are:](image-3.png "")
12::8:10![Figure : Correlation of Nth hour stage of base station with (N+T)th hour stage of forecasting station for 2004 and 2005](image-4.png "1 2 Figure : Figure : 8 9Figure : Figure 10 :")
1
2RiverStationsDangerlevel (m)PadmaHardinge Bridge(Base Station)14.25GoalundoTransi(Forecasting Station) 8.65Hydrologic yearTravel time (day) Historical method Mutreja's methods200412200512200622200712200822200912201022201112201212b) Correlations Between Nth hour stage of Base Station and (N+T)th hour Stage of Forecasting Station
10Y = 1.283X -8.3517.5 8 8.5 9 9.5 HOUR GAUGE OF FORECASTING STATION (m)(N+T)-TH712.612.813 N-TH HOUR GAUGE OF BASE STATION (m) 13.2 13.4 13.6 13.814EHydrological yearFlood Period (day)Forecasting model Y= a+bXNature of the curveRemarks200423Y = 1.262X-7.450Linear VariationFlood Occurred200522Y = 0.827X-2.340Linear VariationFlood Occurred200619Y = 0.850X-3.164Linear VariationNo Flood200720Y = 1.558X-11.91Linear VariationFlood Occurred200816Y = 0.952X+1.131Linear VariationFlood Occurred200919Y = 0.736X-1.047Linear VariationNo Flood201028Y = 0.452X+1.529Linear VariationFlood Occurred201122Y = 0.311X+3.52Linear VariationNo Flood201224Y = 0.99X-4.732Linear VariationFlood Occurred? Combined co-relation of N-th hour stage of base station with (N+T)th hour stage of forecasting station has been established as a linear equation. ? The general equation for the flood forecasting for the GoalundoTransi station is Y = 1.283X -8.351.
			© 2014 Global Journals Inc. (US)
			Year 2014
		
		
			

				


* 
	
		Development of the Jamuneswari Flood Forecasting System: Case Study in Bangladesh
		
			MMRahman
		
		
			NKGoel
		
		
			DSArya
		
	
	
		Journal of Hydrologic Engineering
		
			17
			10
			
			2012
		
	




* 
	
		Hydrometeorological variabilty in the greater Ganges-Brahmaputra-Meghna basins
		
			MRChowdhury
		
		
			NWard
		
	
	
		International Journal of Climatology
		
			24
			12
			
			2004
		
	




* 
	
		Design flow and stage computation in the Teesta river
		
			MMRahman
		
		
			DSArya
		
		
			NKGoel
		
		
			APDhamy
		
	
	
		J. Hydrol. Eng
		
			16
			2
			
			2011a
		
	




* 
	
		Applied Hydrology
		
			KNMutreja
		
		
			1986
			Tata McGra-Hill Publishing Company Limited
		
	




* 
	
		A Textbook of Hydrology
		
			PJ RReddy
		
		
			2005
			Laxmi Publications