The Duration-Over-Threshold Model

Overview

dotmodel is the python library for applying the Duration-Over-Threshold model (Lawson et al., 2024). The Duration-Over-Threshold model is a hierarchical statistical model that characterizes the return period of a given flowrate being exceeded for a given duration. The current functionality of this library includes:

• Querying site and flowrate data from USGS National Water Information Service (NWIS) REST API.
• Cleaning flowrate records.
• Extracting and attributing flood events from a flowrate record.
• Estimating parameters of the Duration-Over-Threshold model from flood event series.
• Plotting of the Duration-Over-Model fit.
• Calculating relationships between flowrate, duration, and recurrence interval.

For a detailed explanation of this model, please see the open-access article from Lawson et al. (2024).

Example

The example below downloads relevant data for USGS gage 01638500 (Potomac River at Point of Rocks, MD), cleans the timeseries, and estimates the Duration-Over-Threshold model parameters.

from dotmodel import Gage

db_path = r'01638500.db'
potomac = Gage(db_path)
potomac.download_nwis('01638500', '10-01')
potomac.clean_timeseries()
potomac.flow_frequency_analysis(truncation_arrival_rate=2)
potomac.duration_analysis()

The model fit may then be plotted.

potomac.plot_flow_frequency('01638500_ffa.png')

potomac.plot_duration_power_law('01638500_duration.png')

potomac.plot_model_summary('01638500.png')

Installation

dotmodel can be installed using pip:

$ python3 -m pip install dotmodel

or conda:

$ conda install -c conda-forge dotmodel

Issue tracker

Any bugs, suggestions, or questions may be logged using the issues tab

https://github.com/sclaw/duration-over-threshold/issues