A Python CLI tool to simplify hydrologic model calibration for NextGen In A Box (NGIAB) workflows.
Project description
ngiab-cal
A Python CLI tool to simplify hydrologic model calibration for NextGen In A Box (NGIAB) workflows.
Table of Contents
- What is this?
- Installation
- Requirements
- Basic Usage
- Advanced Options
- Calibration Process
- Calibration Configuration File
- Example: Calibrating CAMELS Basins
- How It Works
- How is ngen-cal running?
- Development
- License
- Acknowledgments
What is this?
ngiab-cal is a utility that works with the NGIAB folder structure. It automates the creation of a calibration directory with all necessary configuration files to run a modified version of ngen-cal.
The tool simplifies these key tasks:
- Creating calibration configurations
- Running the calibration process using Docker
- Copying calibrated parameters back to your model configuration
Installation
Several installation options are available:
Using uvx (recommended: for its speed,and efficient environment management)
# Run directly without installation
uvx ngiab-cal --help
# Or install as a tool
uv tool install ngiab-cal
ngiab-cal --help
Using pipx
pipx install ngiab-cal
ngiab-cal --help
Traditional pip installation
pip install ngiab-cal
python -m ngiab_cal --help
Requirements
- Python 3.10+
- Docker (for running calibrations)
- Internet connection (for downloading USGS data)
Basic Usage
# Create calibration configuration
ngiab-cal /path/to/ngiab/data/folder -g USGS_GAGE_ID
# Create and run calibration (200 iterations)
ngiab-cal /path/to/ngiab/data/folder -g USGS_GAGE_ID --run -i 200
# Force recreation of calibration configuration
ngiab-cal /path/to/ngiab/data/folder -g USGS_GAGE_ID -f
Advanced Options
usage: ngiab-cal [-h] [-g GAGE] [-f] [--run] [-i ITERATIONS] [--debug] [-w WARMUP] [--calibration_ratio CALIBRATION_RATIO]
data_folder
Create a calibration config for ngen-cal
positional arguments:
data_folder Path to the folder you wish to calibrate
options:
-h, --help show this help message and exit
-g GAGE, --gage GAGE Gage ID to use for calibration
-f, --force Overwrite existing configuration
--run Try to automatically run the calibration, this may be unstable
-i ITERATIONS, --iterations ITERATIONS
Default:100 number of iterations to calibrate for
--debug enable debug logging
-w WARMUP, --warmup WARMUP
Default:365
Number of days at the beginning of the simulation
to exclude from calibration objective metric calculation
--calibration_ratio CALIBRATION_RATIO, --cr CALIBRATION_RATIO
Default:0.5
How to split time after warmup into calibration and validation.
1 == 100% calibration, 0 == 100% validation, 0.8 == 80% calibration 20% validation
Calibration Process
The tool applies a standard hydrological modeling workflow, which involves warmup, calibration, and validation periods. The --warmup period is crucial for allowing the model to reach a stable state before its performance is evaluated against observed data. Following the warmup, the remaining period is typically divided into calibration (where model parameters are adjusted to match observations) and validation (where the model's performance with the calibrated parameters is tested on an independent dataset). The tool facilitates this split, as detailed in the diagram and options below.
Default calibration settings on a 5 year period
| year 1 | year 2 | year 3 | year 4 | year 5 |
|<- warmup ->|<- calibration ->| |
|<- warmup ->|<- validation ->|
Example: Calibrating CAMELS Basins
Here's a script to calibrate all CAMELS basins:
#!/bin/bash
# Download a list of CAMELS gage ids
wget https://raw.githubusercontent.com/peckhams/nextgen_basin_repo/5e1317256a9365ae3a24a250358314e1e9ffc339/CAMELS/Data/camels_name.txt ./camels_name.txt
output_folder=$(cat ~/.ngiab/preprocessor)
while read line
do
gage=$(echo "$line" | cut -d ';' -f 1)
echo $gage
# subset the hydrofabric, calculate mean-average area forcings, generate model config files
uvx --from ngiab_data_preprocess cli -i gage-"$gage" -sfr --start 2007-10-01 --end 2013-09-30
# calibrate gage for 200 iterations
uvx ngiab-cal "$output_folder"/gage-"$gage" -g "$gage" --run -i 200
done < <(tail -n +2 ./camels_name.txt)
Calibration Configuration File
ngiab-cal generates a configuration file at calibration/ngen_cal_conf.yaml that controls the calibration process. This file is created from a template with the following key sections:
General Configuration
general:
strategy:
type: estimation
algorithm: dds # Uses Dynamically Dimensioned Search algorithm
name: calib # Don't modify this
log: true # Enable logging
workdir: /ngen/ngen/data/calibration # Don't modify this working directory in the Docker container
yaml_file: /ngen/ngen/data/calibration/ngen_cal_conf.yaml # Don't modify this either
iterations: 100 # Number of calibration iterations (customizable with -i flag)
restart: 0 # Start from beginning (0) or resume from iteration
Model Parameters
The file defines parameters for both CFE (Conceptual Functional Equivalent) and Noah-OWP-Modular (an extended, refactored version of the Noah-MP land surface model) hydrologic models:
CFE:
- name: b # CFE parameter name
min: 2.0 # Minimum allowed value
max: 15.0 # Maximum allowed value
init: 4.05 # Initial value
- name: satpsi
min: 0.03
max: 0.955
init: 0.355
# Additional parameters...
For each parameter, the configuration specifies:
name: Parameter identifiermin: Minimum allowed value during calibrationmax: Maximum allowed value during calibrationinit: Initial value
Evaluation Configuration
eval_params:
objective: kge # Kling-Gupta Efficiency as objective function
evaluation_start: "..." # Start time for calibration period
evaluation_stop: "..." # End time for calibration period
valid_start_time: "..." # Start time including warmup
valid_end_time: "..." # End time of simulation
# Additional time parameters...
basinID: 01646500 # USGS gage ID
site_name: "USGS 01646500: " # Label for plots
These time periods follow the calibration process diagram shown earlier, with separate periods for warmup, calibration, and validation.
How It Works
- Input Validation - Checks that all required files are present
- USGS Data Download - Retrieves observed streamflow for calibration
- Configuration Creation - Generates necessary files for ngen-cal
- Docker Execution - Runs the calibration in a containerized environment
- Parameter Extraction - Copies calibrated parameters back to your configuration
How is ngen-cal running?
The tool uses a custom branch of CIROH-UA/ngen-cal called ngiab_cal. This branch contains:
- A modified version of ngen-cal
- A Dockerfile that builds on top of the Next Gen In-A-Box container
- Installation of ngen-cal inside the container
When you provide the --run argument, it downloads a pre-built Docker image (joshcu/ngiab-cal:demo) to run the calibration.
Limitations
ngiab-cal has several important limitations to be aware of:
-
Limited Calibration Algorithms: Currently only configured to use the Dynamically Dimensioned Search (DDS) algorithm. Other algorithms available in the main ngen-cal branch are not supported.
-
Limited Model Support: Only calibrates parameters for CFE (Conceptual Functional Equivalent) and NoahOWP-Modular hydrologic models. Other models in the NextGen framework are not currently supported.
-
Single-Gage Calibration: Designed for calibrating single-basin models using one USGS streamgage. Multi-gage or multi-objective calibration is not supported.
-
Custom ngen-cal Branch: Uses a modified version of ngen-cal from the
ngiab_calbranch, which differs from the main branch. Features from newer releases of ngen-cal may not be available. -
Compatibility Concerns: The customized ngen-cal implementation may not be compatible with future changes to either ngen-cal or the NGIAB framework.
Development
Contributions welcome! Comprehensive development instructions coming soon
Visualisation
During calibration, plots will be created inside ./calibration/Output/Calibration_Run/ngen_*******_worker/Plot_Iteration/
After calibration, the tool will copy the validated parameters back into the root config folder which allows you to run ngiab with the guide script and use the teehr / tethys visualiser.
More streamlined workflow coming soon.
License
"Software code created by U.S. Government employees is not subject to copyright in the United States (17 U.S.C. §105). The United States/Department of Commerce reserve all rights to seek and obtain copyright protection in countries other than the United States for Software authored in its entirety by the Department of Commerce. To this end, the Department of Commerce hereby grants to Recipient a royalty-free, nonexclusive license to use, copy, and create derivative works of the Software outside of the United States."
Acknowledgments
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file ngiab_cal-0.3.4.tar.gz.
File metadata
- Download URL: ngiab_cal-0.3.4.tar.gz
- Upload date:
- Size: 61.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: uv/0.7.16
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
84a66bca2020e9b04514084d47c47066e30a76f798d7993be578c10e8c9c814c
|
|
| MD5 |
3ab200859af8ef1c074a85c78f0eb4a4
|
|
| BLAKE2b-256 |
d8bdcc37381ee656e2ba8be9233ed6dffc693ed47da30e9ebe5f3421c43c5142
|
File details
Details for the file ngiab_cal-0.3.4-py3-none-any.whl.
File metadata
- Download URL: ngiab_cal-0.3.4-py3-none-any.whl
- Upload date:
- Size: 13.8 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: uv/0.7.16
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
c24a500ad09ddf554e5fe4e117077112f382ac37a8782f1a78400e855bc2eaaa
|
|
| MD5 |
80a4b1aaa1c3cd0f5e8849b54d3772c4
|
|
| BLAKE2b-256 |
8fc84ea83dba6b2d4fcc302d0b69a8df358f4914dbbbd984343250f110ffc94b
|