The flood risk prevention and the effective water resources management and planning require flows of a river to be recorded continuously and analysed at considered barrages or dams. For a given barrage and dam the flow temporal variation over a certain span of time can graphically be represented by a hydrograph. The hydrograph provides information which is essential for determining the frequencies and severity of flood events. The major characteristics of flood hydrograph are duration (d), peak (p) and volume (v) which are obtained from the components of the hydrograph. Traditional approaches for frequency analysis of hydrology considered these features individually in a context of the univariate study. These features were considered jointly in a multivariate study so that the structure of their dependency could be encountered. Hence, the univariate approaches are not reliable enough to represent the phenomenon and cause reducing the accuracy of risk estimation whereas multivariate approaches contribute in improving the accuracy of risk estimation by considering joint features and dependency structure between characteristics of a flood. However, all the preceding approaches carried the aim of analysing the number of flood characteristics and ignored the comprehensive information provided by the hydrograph. Present research carries an object which contains a framework to introduce the hydrographs as a functional data curve for frequency analysis. This context uses the whole hydrograph as an infinite dimensional observation so that more efficient and effective estimates regarding the risk of associated extreme events can be obtained. The contribution of the proposed approach is to address the problem usually encountered in hydrology regarding incomplete data by exhaustively employing complete information produced by the hydrographs. A number of tools for the analysis of functional data (AFD) are introduced and applied to frequency analysis of flood (FAF) in order to execute exploratory analysis which is an initial step towards comprehensive functional FAF. The methods employed as a first step of exploratory analysis are a visualization of data, measures of location and scale, principal component analysis and detection of outliers are illustrated using real-world flood data of barrages and dams on the River Indus of Pakistan.