So in Peak over the threshold, you pick randomly and then check if its good with the Mean residual plot and stability plot, or is there a way to choose the threshold in peak over threshold
Well, you do not pick the threshold randomly. Assuming you want to create a series for flood frequency analysis, you want set your threshold little high and also make sure that have enough data points in your series. Hope this helps.
Thank you for the interesting presentation. Why did you call it the probability of failure if you already calculated the probability of exceeding the threshold? could you please explain. I didn't understand this part.
I have 30years data i need flood of 10years, 25years, 50years and 100years return period . do i use FFA using anual maximum series or log pearson type III (PDF) to dtermine the return period
Professor, I have a question. Say I have 3650 days( 10 years) of daily discharge. I found 365 days Q is larger than 100K cfs. Does that mean the exceedance probability for 100K cfs is 10%? Then, is my return period 10 days? or 10 years? Thanks a lot!
Thank you for this video! Very well done and helpful!
Brilliant! Thank you very much for an amazing lecture.
So in Peak over the threshold, you pick randomly and then check if its good with the Mean residual plot and stability plot, or is there a way to choose the threshold in peak over threshold
Well, you do not pick the threshold randomly. Assuming you want to create a series for flood frequency analysis, you want set your threshold little high and also make sure that have enough data points in your series. Hope this helps.
Thank you for the interesting presentation. Why did you call it the probability of failure if you already calculated the probability of exceeding the threshold? could you please explain. I didn't understand this part.
I have 30years data i need flood of 10years, 25years, 50years and 100years return period . do i use FFA using anual maximum series or log pearson type III (PDF) to dtermine the return period
your LP3 or any other suitable distribution that fits your data, espeicially for 25, 50 and 100 year flow
Professor, I have a question. Say I have 3650 days( 10 years) of daily discharge. I found 365 days Q is larger than 100K cfs. Does that mean the exceedance probability for 100K cfs is 10%? Then, is my return period 10 days? or 10 years? Thanks a lot!
10 years.
Thanks Sir.
So does the calculated probability of 93% also means, in every year there is 93% chance of exceeding 75,000 cfs?
The probability that the flow will exceed 75000 at least once in next 25 yrs is 93%.
Thank you Sir. one question. How many number of data is ideal should be used to do a flood frequency analysis?
Check the flood frequency analysis part 2 video. That will hopefully answer your question.
Thankyou :)
How can we define the threshold value for eg 120 years of RF data
Is it possible to generate a demensionless hydrograph using discharge data?
yes, but you still need to know the amount of rainfall that produced that discharge to get the dimensionless unit hydrograph.