This repository contains the source codes to estimate GLOF peak discharges from the Himalayan lake-size distribution, and to estimate return periods of GLOF volumes and peak discharges in the entire Himalayas and seven subregions. A mandatory prerequisite to run the scripts is to install R (https://www.r-project.org/) and RStudio (https://rstudio.com) on your machine. The comments within the scripts provide further details on model dependencies and usage of functions.
Bayesian piecewise regression model to learn peak discharge Qp from eta, the product of flood volume and the breach rate. Written in R.
This model calls a script to learn the posterior distributions of the model parameters, written in STAN.
Robust regression model to estimate the maximum depth of Himalayan glacier lakes from glacier lake area.
Script to estimate regional GLOF hazard from empirical GLOF rates and predicted GLOF volumes and discharges. Contains all commands to reproduce the figures from the associated manuscript.
Parts of our codes are based on the HDI function from Kruschke, J. K. (2014). Doing Bayesian Data Analysis: A Tutorial with R, JAGS, and Stan. 2nd Edition. Academic Press / Elsevier. We redistribute this function for ease of use, though the original function can be found here: https://sites.google.com/site/doingbayesiandataanalysis/software-installation
... can be found here: https://doi.org/10.5281/zenodo.3523213
Veh, G., Korup, O., & Walz, A. (2020). Hazard from Himalayan glacier lake outburst floods. Proceedings of the National Academy of Sciences, 117(2), 907-912.
Georg Veh
Working group on natural hazards
University of Potsdam
https://www.uni-potsdam.de/de/umwelt/forschung/ag-naturgefahren.html