EXTREME VALUE ESTIMATION OF HYDROMETEOROLOGICAL DATA

Abstract

Wind speed is of particular importance in hydrology as it controls the important component of evaporation.. It is also one of the sources of energy generation, specially in and regions where hydropower potential is very limited. Wind -speed exhibits wide variation not only from place to place but also shows strong diurnal variation at the same place. Wind flow and speeds are controlled by local pressure anomalies which in turn are controlled by the temperature. It may also be influenced by the local topographic features, which may funnel the wind and increase it above the aerial average, conversely some stations will have wind speeds reduced by shelter. Extraordinary wind speeds may be experienced in some parts through the incursion of tropical cyclones. Modern design of structures is based on the concept of probability. A structure is designed to provide a specific degree of safety against high winds, determined by the probability of occurrence of high winds exceeding design value. Therefore, extreme wind quantiles are needed to calculate design values of wind load-effect on structures. These quantiles are estimated by conducting frequency analysis of the available wind speed data. Usually, at the site.of interest either wind speed data are not available or are of very short record length (generally 20-40 years). The reliable estimation of extreme quantiles under such situation is of considerale theoretical and practical importance. A study has been undertaken to arrive at the extreme wind quantiles of various stations in India using pooled frequency analysis approach. In this approach, the data of individual stations are combined to form regional super stations to decrease sampling errors by effectively increasing the sample length. In this study, frequency analysis of 3-hourly AMS of wind speed of 144 stations spread throughout India has been carried out. The randomness of AMS of various stations has been tested using Turning point test and Anderson's correlogram test while analysis for presence of trend has been done using Kendall's rank correlation test and Mann Kendall's test. For most of the stations, all the test have given similar results. However, for few stations conflicting results have been obtained.