pybaseflow

A comprehensive Python toolkit for baseflow separation from streamflow hydrographs.

pybaseflow implements 17 baseflow separation methods spanning four paradigms: recursive digital filters, graphical/interval methods, recession-based methods, and tracer-based methods. It also provides parameter estimation, USGS data retrieval, and a CMB-to-Eckhardt calibration bridge.

This project builds on the baseflow package by Xie et al. (2020), which implemented methods described in "Evaluation of typical methods for baseflow separation in the contiguous United States" (Journal of Hydrology, 583, 124628). pybaseflow extends that work with new methods (PART, CMB, BFlow, IHACRES), a unified filter architecture, modern packaging, and USGS data integration.

This project is funded by CIROH.

Install

pip install pybaseflow

Quick Start

import pybaseflow

# Load bundled sample data (USGS 01013500, Fish River, Maine)
data = pybaseflow.load_sample_data()
Q = data['Q']

# Run a single method
b = pybaseflow.eckhardt(Q, a=0.98, BFImax=0.8)

# Or fetch your own data from USGS NWIS
from pybaseflow.io import fetch_usgs
data = fetch_usgs('01013500', '2019-01-01', '2020-12-31')

Methods

Recursive Digital Filters

All recursive digital filters share a generalized core: b[t] = alpha * b[t-1] + beta * (Q[t] + gamma * Q[t-1])

gamma=0 family (linear reservoir based):

Method	Function	Reference
Boughton	boughton(Q, a, C)	Boughton, 1993
Chapman-Maxwell	chapman_maxwell(Q, a)	Chapman & Maxwell, 1996
Eckhardt	eckhardt(Q, a, BFImax)	Eckhardt, 2005
EWMA	ewma(Q, e)	Tularam & Ilahee, 2008
Furey-Gupta	furey(Q, a, A)	Furey & Gupta, 2001
WHAT	what(Q, BFImax, a)	Lim et al., 2005 (alias for Eckhardt)

gamma=1 family (signal processing based):

Method	Function	Reference
Lyne-Hollick	lh(Q) / lh_multi(Q, num_pass=3)	Lyne & Hollick, 1979; Nathan & McMahon, 1990
Chapman	chapman(Q, a)	Chapman, 1991
Willems	willems(Q, a, w)	Willems, 2009

Variable gamma (hybrid):

Method	Function	Reference
IHACRES	ihacres(Q, a, C, alpha_s)	Jakeman & Hornberger, 1993

Graphical / Recession-Based Methods

Method	Function	Reference
UKIH	ukih(Q, b_LH)	UKIH, 1980
Local minimum	local(Q, b_LH, area)	Sloto & Crouse, 1996
Fixed interval	fixed(Q, area)	Sloto & Crouse, 1996
Sliding interval	slide(Q, area)	Sloto & Crouse, 1996
PART	part(Q, area)	Rutledge, 1998

Tracer-Based Methods

Method	Function	Reference
Conductivity Mass Balance	cmb(Q, SC)	Stewart et al., 2007

Recession Analysis

Function	Description
bflow(Q)	BFlow 3-pass filter + recession analysis (Arnold & Allen, 1999)
bn77(Q, ...)	Brutsaert-Nieber drought flow identification (Cheng et al., 2016)

Parameter Estimation

import pybaseflow

data = pybaseflow.load_sample_data()
Q = data['Q']

# Estimate recession coefficient from the hydrograph
strict = pybaseflow.strict_baseflow(Q)
a = pybaseflow.recession_coefficient(Q, strict)

# Use the estimated recession coefficient with any filter
b = pybaseflow.eckhardt(Q, a, BFImax=0.8)
b = pybaseflow.chapman_maxwell(Q, a)
b = pybaseflow.willems(Q, a, w=0.5)

CMB Calibration Bridge

Use specific conductance data to calibrate Eckhardt's BFImax:

from pybaseflow.tracer import calibrate_eckhardt_from_cmb

cal = calibrate_eckhardt_from_cmb(Q, SC)
b = pybaseflow.eckhardt(Q, cal['a'], cal['BFImax'])

BFlow / SWAT Integration

result = pybaseflow.bflow(Q)
print(f"ALPHA_BF = {result['alpha_factor']:.4f}")  # for SWAT calibration
print(f"BFI = {result['BFI']:.3f}")
print(f"Baseflow days = {result['baseflow_days']:.1f}")

USGS Data Retrieval

from pybaseflow.io import fetch_usgs

# Fetch daily discharge
data = fetch_usgs('01013500', '2015-01-01', '2020-12-31')
Q = data['values']

# Fetch specific conductance (for CMB)
sc_data = fetch_usgs('01013500', '2015-01-01', '2020-12-31', parameter='sc')

License

MIT