Analysis and comparison of GloFAS and Google FloodHub river discharge data

This repository contains scripts that together perform an analysis of GloFAS's and Google FloodHub's predictive performance of river discharge data. For both, the data is downloaded, processed to a uniform type, and, subsequently, their accuracies (e.g. by probability of detection (POD) and false alarm ratio (FAR)) are assessed using both impact data and observational data.

As of now, late 2024, the scripts focus on floods in Mali, but later, the goal is to generalize to more countries. For questions, contact [email protected] and [email protected].

Overview

An analysis of historical forecasts versus a ground truth, i.e. impact- or observational data, can be divided into different parts: (1) preprocessing forecasts to events; (2) preprocessing ground truth to events; (3) comparing them. Here, "events" flexibly denote periods of consecutive flooding in a predefined spatial area. Something is considered a flood when either, in case of quantitative discharge data, a certain return period threshold or percentile treshold is passed, or, in case of qualitative event data, according to the source's flood interpretation.

The repository contains four main folders:

1. comparison: contains scripts to directly compare results of different forecasting methods, including plotting, et cetera.
2. GloFAS: contains scripts to preprocess GloFAS forecasts into events, and also scripts to preprocess the impact and observational data to these same uniform events.
3. GoogleFloodHub: contains scripts to download both real-time forecasts through an API, and download historical forecasts (the GRRR dataset) through an online environment. The latter are transformed to events together with impact- and observational data in the GRRR.ipynb file.
4. PTM: contains scripts to analyse the status quo; the so-called "propagation trigger model" (PTM).

Data

Multiple data sources were utilised.

GloFAS:

retrieved from the early warning data system of copernicus https://ewds.climate.copernicus.eu/ The folder GloFAS_data_extractor contains information for extraction of these different datasets (this is also initialized in the probability calculator, just make sure your folder structure follows a similar structure as in the probability_calculator.py)

Impact data

Impact data is not stored online and can be shared upon requested, either through the contact persona above, or through [email protected].

Observation data

source: DNH, not publicly available.

Preparation of data

GloFAS:

comparison to observational data:

for comparison to observational data or needing any timeseries of a single cell, run Q_timeseries_station.py

comparison to impact data:

this is initialized in the probability calculator

Analysis

GloFAS

run and inspect scriptsi in the following order

1. probability_calculator.py
2. flood_definer.py
3. performance_calculator.py

scripts accommodate differences in: (and therefore need information on these variables)

• area of interest
• leadtimes
• trigger probability
• threshold type (return period or percentiles)
• threshold value
• comparison type (observational or impact data)
• timerange

Data is also retrieved in probability_calculator (though you need to run 'Q_timeseries_station.py' for comparison to obs data seperately)

Post processing

in the comparison folder - visualization folder, visualizations are suggested in the plot.py, which contains a visualization class