Assessment of Extreme Rainfall Using Gumbel Distribution for Estimation of Peak Flood Discharge for Ungauged Catchments
- February 13, 2018
- Posted by: RSIS
- Categories: Applied Science, Environmental Science
International Journal of Research and Innovation in Social Science (IJRISS) | Volume I, Issue VI, June 2017 | ISSN 2454-6186
N. Vivekanandan
Scientist-B, Central Water and Power Research Station, Pune 411024, Maharashtra, India
Abstract—Estimation of Peak Flood Discharge (PFD) at a desired location on a river is important for planning, design and management of hydraulic structures. For ungauged catchments, rainfall depth becomes an important input in derivation of PFD. So, rainfall depth can be estimated through statistical analysis by fitting probability distribution to the rainfall data. In this paper, the series of annual maximum 1-day rainfall derived from the daily rainfall data observed at Dhaulakuan rain gauge station is used for estimation of 1-day extreme rainfall adopting Gumbel distribution. Maximum likelihood method is used for determination of parameters of the distribution. Anderson Darling test is applied for checking the adequacy of fitting of the distribution to the observed rainfall data. The estimated 1-day extreme rainfall obtained from Gumbel distribution is used to compute the 1-hour maximum value of distributed rainfall that is considered as an input to estimate the PFD by rational formula adopting CWC guidelines. The study suggests the estimated PFD could be used for design of flood protection works for different ungauged catchments of river Yamuna.
Index Terms—Anderson-Darling test, Gumbel, Rainfall, Peak flood discharge, Maximum likelihood method
I. INTRODUCTION
Estimation of Peak Flood Discharge (PFD) at a desired location on a river is important for planning, design and management of hydraulic structures such as dams, bridges, barrages and design of storm water drainage systems. These include different types of flood such as standard project flood, probable maximum flood and design basis flood. In case of large river basins, the hydrological and stream flow series of a significant duration are generally available. However, for ungauged catchments, more data is not available other than rainfall. The rainfall data is also of shorter duration and may becomes an important input in derivation of PFD [1]. For arriving at such design values, statistical analysis by fitting probability distribution to the rainfall data is carried out.
