#geomorphology #radiocarbon #charcoal


A library for simulating Bolin & Rodhes reservoir models

7 releases

new 0.1.6 May 3, 2021
0.1.5 Feb 24, 2021

#71 in Science

Download history 84/week @ 2021-02-18 19/week @ 2021-02-25 12/week @ 2021-03-04 2/week @ 2021-03-11 24/week @ 2021-03-25 6/week @ 2021-04-01 6/week @ 2021-04-08 18/week @ 2021-04-15 6/week @ 2021-04-22 23/week @ 2021-04-29

59 downloads per month



Reservoirs - A library for modeling Bolin & Rodhe reservoirs.

Bolin & Rodhe (1973) describe methods for characterizing the turnover rate of mass accumulating in a reservoir. The distribution of ages of particles in the reservoir constrain the possible input and output rates that could produce the observed record. The functions in this crate allow the user to compare synthetic accumulation records to observed records using the K-S and Kuiper statistics, to determine the best-fitting input/output pair for an observed record.

In my research at Oregon State University, I estimate the transit times of stream sediments moving through headwater valleys of the Coast Range by fitting reservoir models to a record of charcoal ages sampled from stream bank deposits. Inherited age refers to the age of charcoal when it enters a stream deposit. If we do not account for inherited age in the model, then transit times become artificially inflated. I have added an inherited age capacity to reservoirs, and while this is not a traditional feature of Bolin & Rodhe reservoirs, it is useful for dealing with charcoal ages.

This library includes the full code base used to estimate transit times for my ongoing dissertation, published here in the interest of academic transparency.

Quick Start

To use reservoirs, add it to your Cargo.toml

reservoirs = "^0.1.6"

Let's load the stream bank charcoal data from the Oregon Coast that I use in my research. Many of the functions serve to compare a synthetic distribution against an observed record, and stream bank charcoal makes a handy foil.

use reservoirs::prelude::*;

fn main() -> Result<(), ResError>{
    // mean expected deposit age and inherited age by facies
    let dep = Sample::read("https://github.com/crumplecup/reservoirs/blob/master/examples/dep.csv")?;
    let iat = Sample::read("https://github.com/crumplecup/reservoirs/blob/master/examples/iat.csv")?;

    // subset mean ages of debris flows
    let df: Vec<f64> = dep.iter()
        .filter(|x| x.facies == "DF")
        .map(|x| x.age)
    // subset inherited ages
    let ia: Vec<f64> = iat.iter()
        .map(|x| x.age)

    let mut debris_flows = Reservoir::new()

    // model parameters
    let period = 30000.0; // run simulations for 30000 years
    let runs = 1000; // run 1000 simulated accumulations per candidate pair for goodness-of-fit

    // create reservoir model using builder pattern
    let mut model = Model::new(debris_flows)

    // sample a stereotypical record from 1000 runs of 30000 years
    let eg = model.stereotype(500);
    // compare the CDF of the synthetic example to the observed debris-flow deposit record
    plot::comp_cdf(&eg, &df, "examples/df_cdf.png");



~212K SLoC