HYDROLOGICAL APPRAISAL UNDER MISSING DATA CONDITIONS

Abstract

Continuous hydrological observations are required for assessment of water resources from a catchment for planning and design. However, gaps in the hydrological data series are common due to variety of reasons. For meaningful hydrological analysis, missing values in such data series are often reconstructed using statistical and deterministic methods. Depending upon the type of data and the information available, different technique are employed to find the missing values. It is common to use hydrological simulation models to compute runoff for periods when observations are not available. However, the statistical characteristics and matching of end point values between observed and model computed values pose some problem in making proper hydrological assessment. It is therefore important to select model suitable to a particular watershed. In present study, the TOPMODEL is chosen to simulated runoff response in predominantly forested watershed of Western Ghats situated in humid climate. Suitability of the TOPMODEL was adjudged by calibrating and validating model on independent data sets. Results of the application indicate that the TOPMODEL is suitable to simulate runoff response in Kokkarne catchment as the model efficiency is more that 87% for both calibration and validation. Sensitivity of model parameters was also carried out and it was found that the model is sensitive to transmissivity decay parameter (m), the surface transmissivity (1nTo), the root zone water available capacity (SRmax), initial moisture deficit in the root zone (SRo). The channel velocity (ChVel) parameter does not show any sensitivity indicating that it is a dummy parameter. The calibrated TOPMODEL is used to fill the missing values of different length in observed runoff series using. four different methods namely scaling, linear transformation, move by constant and adjustment of end values depending on gap characteristics. The choice of use of a particular method is governed by certain conditions discussed in study. Analysis of results obtained indicate that the adjustment of model generated series — by methods presented in this study — for places where data gaps exists, improve the fit between observed and computed series is there compared to filling the gap with model generated series alone. For all four cases of data filling methods discussed it is found that in all the methods, when we replace the TOPMODEL computed runoff values by the computed values with adjustment, efficiency increases. This indicated that the data'filling methods described in this study are useful and improve the matching between observed and computed series. Therefore, these methods can be successfully used to estimate the gaps in the discharge series in future application.