FLOOD FREQUENCY ANALYSIS OF KUMAON REGION RIVERS UTTARAKHAND, INDIA
Keywords:
Flood Frequency Analysis, River Kosi, Gola & Sarda, Annual Peak Flood discharge, Return Period, Goodness of fit TestAbstract
Estimates of flood frequency quantiles are important in planning and design of hydraulic structures, Hence there is aneed to seek for the most appropriate design estimator that would meet both safety and economic considerations of
such structures. Flood frequency analysis is a tool used to estimate the frequencies of likely an occurrence of futurefloods. The objective of study is to estimate flood parameters of Kumaon Region Rivers of Uttarakhand, for
different return period using statistical approaches, screening test using Anderson’s Correlogram, Chowtest for
Outlier Detection and Kendall’s Rank Correlation. Goodness of Fit tests Chi-square test, D-Index and K-S tests areapplied to all the chosen, six probability distributions namely Normal, Log normal, Pearson type III, Log Pearsontype III, Gumbel, Log Gumbel distributions using method of moments. By applying Model efficiencytest,
Coefficient of determination, root mean square errors for distributions and to suggest for best fitting distributionfor
Kumaon Region Rivers.In the present study, flood frequency analysis has been carried out for Kumaon RegionRivers namely Gola near Kathgodam barrage having Catchment area 600 km2
, Kosi near Ramnagar, Nainital district,
6753 km2 and Sarda near Banbassa barrage of Champawat district 15100 km2 catchment areas. The annual flooddata’s of all the river stations has been collected from the Irrigation departments. Goodness of fit tests shows that
best probability distributions for all the three river stations are Log normal and Log Gumbel, but it is recommendedform the literature review that to use Log Gumbel. Hence Log Gumbel is recommended for estimation of floodquantities. From model efficiency study it has been found that for river Kosi, Log Gumbel distribution, Log PearsonType III distribution are best fitting for other two rivers.From trend line equation, maximumcoefficient of
determination ( ) value and minimum root mean square error shown that for Kumaon Region Rivers for
predicting expected flow Log Gumbel is the best distribution.