🦦 OTTAR

Ode To Transient Ancho de los Rivers

Transiently evolving river-channel width as a function of streambank properties, sediment in transport, and the hydrograph.

Purpose

This model is designed to compute the rates of river-channel widening and narrowing based on changing hydrological regimes. It is currently designed for rivers with cohesive banks, with a critical shear stress for particle detachment and an erosion-rate coefficient.

Installation

From PyPI:

pip install ottar

Locally, inside a clone of this git repository (the -e permits you to make local updates to the code and have them incorporated into the way that OTTAR runs):

pip install -e .

Structure

OTTAR contains:

• The RiverWidth class, which contains methods to evolve the width of an alluvial river.
• The FlowDepthDoubleManning class, which is used to estimate flow depth from discharge, even with an evolving river-channel geometry.

Examples

There's a folder for these!

Model inputs and outputs

Inputs

Key input parameters (RiverWidth)

Variable	Description	Typical value(s)
h_banks	Stream-bank height. This is the thickness of material that must be removed for the river to widen by one unit lateral distance.	1-5 m
S	Channel downstream-directed slope. This is used to compute shear stresses and (if necessary) flow depth from water discharge.	10−3
b0	Initial width. Starting width of a channel.	1–1000 m
tau_crit	Critical shear stress required to start eroding muddy banks. At this stress, the flow begins to be able to detach particles. When set up to perform an inversion using data on river widening and past flows, this is one of two key parameters to be estimated for rivers with detachment-limited banks.	1–10 Pa
tau_star_crit_sed	Critical shear stress required initiate sediment motion. This defaults to 0.0495 from the Wong & Parker (2006) rebuild of the Meyer-Peter & Müller (1948) sediment-transport equation.	0.03–0.06
k_d	Cohesive-detachment erosion-rate coefficient. This determines the rate of erosion as a function of shear stress above critical. When set up to perform an inversion using data on river widening and past flows, this is the other of two key parameters to be estimated.	~10−7 m / (Pa s)
k_E	Noncohesive erosion-rate (entrainment) coefficient. This relates theoretical sediment entrainment rate via near-bank Shields stress to bank-retreat rate via erosion.	~0.01–1
f_stickiness	Fraction of suspended-load particles contacting the bank that "stick" to it. This modulates the turbulence-driven lateral-transport term and its impact on channel-narrowing rate, and comprises the abillity of banks to trap sediment and to hold it.	0–1
k_n_noncohesive	Narrowing coefficient (noncohesive sediment). Trapping and holding efficiency in regards to noncohesive sediment; this may be due to deep pits between grains and/or other bank-rougness properties.	0–1
Parker_epsilon	Excess bed shear-stress factor. $\tau_b = (1+\epsilon) \tau_\beta$, where $\tau_b$ is bed shear stress and $\tau_\beta$ is bank shear stress.	0.2
intermittency	Intermittency. Fraction of the time that the discharge given is active. This is always equal to 1 for a full hydrograph, and is $\leq$ 1 when a characteristic "geomorphically effective" discharge is considered. It can be thought of as a time-dialation factor.	10−3–1
D	Sediment median grain size. This is the median size of the material both in transport and in the banks, and is required for bedload and/or noncohesive-sediment-dominated systems. It may also be specified for rivers dominated by susepended load and bank cohesion, though will likely play a more minor role in these.	10−4–1 m

Key input data sets and parameters (FlowDepthDoubleManning)

This step is used to compute flow depths from a discharge time series, and may be skipped if you already posess a time series of flow depth

• Discharge time series
• Manning's n (channel)
• Roughness / topogrpahy coefficient (floodplains)
• Depth / topography exponent (floodplains)

Outputs

This program outputs a time series of channel width, b(t). It organizes this within a Pandas DataFrame that can also be exported using the write_csv() function within the RiverWidth class.

Plots can also be made of just river width (plotb()) or of discharge and river width (plotQb).