explore extreme values of data - Githubissues

ai4er-cdt / flood_risk_shipston

The purpose of this project is to investigate whether we can establish the effectiveness of natural flood management (NFM) interventions undertaken in the British town of Shipston-on-Stour during 2017 to 2020 from publicly available meteorological data and private data from the river gauge in Shipston.

1 stars 0 forks source link

explore extreme values of data #1

Closed sdat2 closed 3 years ago

sdat2 commented 4 years ago

there aren't many python packages for doing this

sdat2 commented 4 years ago

I will probably initially use https://github.com/sdat2/scikit-extremes to do the analysis. This only has Block-Maxima currently implemented in it (i.e I haven't managed to add Peak-over-Threshold).

JoycelynLongdon commented 4 years ago

Work Stream Goals

[x] Understanding the seasonality of extreme flood events
[ ] Building an extreme value distribution
[ ] Determine whether there are flood basins (gauged) and their thresholds - so we can know
[ ] Create a multivariate extreme distribution between rainfall and the river height (generally in R)

sdat2 commented 4 years ago

bm-fit

Block maxima fit for river height is terrible!

sdat2 commented 4 years ago

useful looking package: https://github.com/iagolemos1/thresholdmodeling/tree/0.0.1

sdat2 commented 4 years ago

return-level-3m-threshold Used 3m threshold alongside stage 60 min data (filtered). The non-independence of different values contributes to quite a bad fit; the top values all come from the large flood in 2007. news article on floods

sdat2 commented 4 years ago

peaks-diff-thresholds

Looking at the different peaks that come out after using scipy.find_peaks, more reasonable as it doesn't pick up as many from the 2007 flood.

Croydon-Brixton commented 3 years ago

Project closed.