UM-5 : Rating curve analysis

Abstract

The water level and the volume of water passing a site in a river provides useful information about the flow, which can be used for many planning purposes. The water levels are easily observed and measured at shorter time interval at a given site, while the discharge measurements are time consuming and costly too. But the data of the discharge at shorter time intervals are needed in hydrological analysis. This is achieved by computing the discharge using the data of water level observations. To enable such computation, a relationship between the water level(stage) and the volume of water passing (discharge) is established. his relationship is naturally a complex one and depends on many factors like inundation of flood plain, unsteadiness of flow, obstruction to flow caused by dams, barrages etc., scouring and deposition and other related factors. If the unsteadiness is not pronounced, for a given water level there would be only one discharge value corresponding to this level.

Establishment of relation between stage and discharge assuming steady state conditions is a current practice. The present practice has some shortcomings as follows : (a) manual plot of stage and discharge and drawing a curve after eye judgement often lacks accuracy, (b) where rating (stage discharge relation) is developed using computer just a mathematical criteria ( Least square method) is used, the para7mters thus arrived may lack physical meaning with respect to particular river and cross section. Therefore, a 7ethadology which takes into account the information on physical features like cross-section, bedslope etc. for developing a rating curve is needed. In case when a rating curve is available and needed to be used for finding of discharge given the stage, the present practice is to read the graph. This practice inherits lot of errors. Sometimes rating tables are prepared and interpolations are made. Again this practice does not remove the data error present in the table or uses smooth interpolations. The methodology described in this user manual has both the features.

A smooth interpolation which accounts for the slope and curvature of the rating curve known as Hermitian interpolation is done. A computer programme for interpolation has been included in the manual. The essential features of the methodology for development of rating curve are: (1) initial data analysis using simple graph plot to remove obvious data errors, (2) double log plot for application of grouping of data and graphical estimation of parameters,(3) physical analysis, and (4) application of the least squares technique and (5) interpretation of results.

The user manual provides step by step method of analysis along with computer programmes. The flow charts for all these programmes are given. Input and output specifications alongwith a sample are also provided.