Final Manning's N and Infiltration Analysis¶
This notebook demonstrates how to extract final Manning's N and preprocessed infiltration values from a reproducible example project. It also includes a regression check for multi-ring infiltration override polygons.
Python
from pathlib import Path
import logging
import importlib
import sys
import types
repo_root = (
Path.cwd().resolve().parent
if Path.cwd().resolve().name == 'examples'
else Path.cwd().resolve()
)
package = types.ModuleType('ras_commander')
package.__path__ = [str(repo_root / 'ras_commander')]
sys.modules.pop('ras_commander', None)
sys.modules['ras_commander'] = package
logging_config = importlib.import_module('ras_commander.LoggingConfig')
package.get_logger = logging_config.get_logger
package.setup_logging = logging_config.setup_logging
decorators = importlib.import_module('ras_commander.Decorators')
package.log_call = decorators.log_call
package.standardize_input = decorators.standardize_input
for logger_name in [
'ras_commander.RasExamples',
'ras_commander.hdf.HdfLandCover',
'ras_commander.hdf.HdfInfiltration',
'ras_commander.hdf.HdfBase',
]:
logging.getLogger(logger_name).setLevel(logging.CRITICAL + 1)
import h5py
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
RasExamples = importlib.import_module('ras_commander.RasExamples').RasExamples
HdfLandCover = importlib.import_module('ras_commander.hdf.HdfLandCover').HdfLandCover
HdfInfiltration = importlib.import_module('ras_commander.hdf.HdfInfiltration').HdfInfiltration
print('Imports successful')
Text Only
Imports successful
Python
project_path = RasExamples.extract_project('BaldEagleCrkMulti2D')
geom_hdf = None
mesh_names = []
for candidate in sorted(project_path.glob('*.g*.hdf')):
with h5py.File(candidate, 'r') as hdf_file:
areas_group = hdf_file.get('Geometry/2D Flow Areas')
if areas_group is None:
continue
current_mesh_names = [
area_name
for area_name, area in areas_group.items()
if isinstance(area, h5py.Group) and 'Infiltration' in area
]
if current_mesh_names:
geom_hdf = candidate
mesh_names = current_mesh_names
break
assert geom_hdf is not None, 'No geometry HDF with preprocessed infiltration data found.'
print(f'Project: {project_path.name}')
print(f'Geometry HDF: {geom_hdf.name}')
print(f'Meshes with infiltration data: {mesh_names}')
Text Only
Project: BaldEagleCrkMulti2D
Geometry HDF: BaldEagleDamBrk.g09.hdf
Meshes with infiltration data: ['BaldEagleCr']
Preprocessed Per-Cell Manning's N¶
Python
mann_df = HdfLandCover.get_preprocessed_mannings_n(geom_hdf)
assert len(mann_df) > 0, "Expected preprocessed Manning's n values."
print(f'Cells: {len(mann_df):,}')
print(
f"N range: {mann_df['mannings_n'].min():.4f} to {mann_df['mannings_n'].max():.4f}"
)
print(f"Mean: {mann_df['mannings_n'].mean():.4f}")
stats = HdfLandCover.get_preprocessed_mannings_stats(geom_hdf)
stats
Text Only
Cells: 19,597
N range: 0.0300 to 0.1500
Mean: 0.0674
| mesh_name | n_cells | min_n | max_n | mean_n | median_n | std_n | |
|---|---|---|---|---|---|---|---|
| 0 | BaldEagleCr | 19597 | 0.03 | 0.15 | 0.067352 | 0.06 | 0.025607 |
Python
fig, ax = plt.subplots(figsize=(10, 5))
ax.hist(
mann_df['mannings_n'],
bins=50,
edgecolor='black',
alpha=0.7,
color='steelblue',
)
ax.set_xlabel("Manning's N")
ax.set_ylabel('Cell Count')
ax.set_title("Distribution of Manning's N Values")
ax.axvline(
mann_df['mannings_n'].mean(),
color='red',
linestyle='--',
label=f"Mean={mann_df['mannings_n'].mean():.4f}",
)
ax.legend()
ax.grid(True, alpha=0.3)
plt.tight_layout()
plt.show()
plt.close(fig)
Calibration Table and Base vs Calibrated Comparison¶
Python
cal = HdfLandCover.get_mannings_calibration_table(geom_hdf)
assert cal is not None and len(cal) > 0, "Expected Manning's calibration data."
print(f'Land cover classes: {len(cal)}, Calibration regions: {len(cal.columns) - 2}')
print()
print('Base values:')
print(cal.iloc[:, :2].to_string(index=False))
Text Only
Land cover classes: 17, Calibration regions: 1
Base values:
Land Cover Name Base Manning's n Value
NaN
NoData 0.060
Barren Land Rock/Sand/Clay 0.040
Cultivated Crops 0.060
Deciduous Forest 0.100
Developed, High Intensity 0.150
Developed, Low Intensity 0.100
Developed, Medium Intensity 0.080
Developed, Open Space 0.040
Emergent Herbaceous Wetlands 0.080
Evergreen Forest 0.120
Grassland/Herbaceous 0.045
Mixed Forest 0.080
Open Water 0.035
Pasture/Hay 0.060
Shrub/Scrub 0.080
Woody Wetlands 0.120
Python
comp = HdfLandCover.compare_base_vs_calibrated(geom_hdf)
if comp is not None and len(comp) > 0:
print(
f"Override entries: {len(comp)}, Classes: {comp['land_cover_class'].nunique()}, Regions: {comp['region_name'].nunique()}"
)
print()
print('Largest N increases:')
print(
comp[comp['delta_n'] > 0]
.nlargest(5, 'delta_n')[['land_cover_class', 'region_name', 'base_n', 'calibrated_n', 'delta_n']]
.to_string(index=False)
)
else:
print('No calibrated overrides found.')
Text Only
Override entries: 14, Classes: 14, Regions: 1
Largest N increases:
land_cover_class region_name base_n calibrated_n delta_n
Open Water Main Channel 0.035 0.04 0.005
Preprocessed Infiltration Parameters¶
Python
for var in ['Curve Number', 'Abstraction Ratio', 'Minimum Infiltration Rate']:
df = HdfInfiltration.get_preprocessed_infiltration(geom_hdf, variable=var)
if len(df) > 0:
print(
f"{var}: {len(df):,} cells, range {df['value'].min():.3f}-{df['value'].max():.3f}, mean {df['value'].mean():.3f}"
)
else:
print(f'{var}: No data')
Text Only
Curve Number: 18,066 cells, range 36.000-100.000, mean 73.645
Abstraction Ratio: 18,066 cells, range 0.000-0.200, mean 0.137
Minimum Infiltration Rate: 18,066 cells, range 0.000-0.120, mean 0.115
Python
cn = HdfInfiltration.get_preprocessed_infiltration(geom_hdf, variable='Curve Number')
assert len(cn) > 0, 'Expected preprocessed Curve Number values.'
fig, ax = plt.subplots(figsize=(10, 5))
ax.hist(
cn['value'],
bins=50,
edgecolor='black',
alpha=0.7,
color='forestgreen',
)
ax.set_xlabel('SCS Curve Number')
ax.set_ylabel('Cell Count')
ax.set_title('Distribution of Curve Number Values')
ax.axvline(
cn['value'].mean(),
color='red',
linestyle='--',
label=f"Mean={cn['value'].mean():.1f}",
)
ax.legend()
ax.grid(True, alpha=0.3)
plt.tight_layout()
plt.show()
plt.close(fig)
Infiltration Region Polygon Regression¶
The bundled example project exposes preprocessed infiltration cell values, but it does not contain geometry-level infiltration override polygons. This regression cell builds a minimal geometry HDF fixture with one exterior ring and one interior ring so the notebook exercises HdfInfiltration.get_infiltration_region_polygons() on the multi-ring path.
Python
fixture_dir = repo_root / 'working' / 'notebook_runs'
fixture_dir.mkdir(parents=True, exist_ok=True)
fixture_hdf = fixture_dir / '211_infiltration_multipart_fixture.g01.hdf'
outer_ring = np.array(
[
[0.0, 0.0],
[4.0, 0.0],
[4.0, 4.0],
[0.0, 4.0],
[0.0, 0.0],
],
dtype=float,
)
inner_ring = np.array(
[
[1.0, 1.0],
[1.0, 3.0],
[3.0, 3.0],
[3.0, 1.0],
[1.0, 1.0],
],
dtype=float,
)
polygon_points = np.vstack([outer_ring, inner_ring])
polygon_parts = np.array(
[
[0, len(outer_ring)],
[len(outer_ring), len(inner_ring)],
],
dtype=np.int32,
)
polygon_info = np.array(
[[0, len(polygon_points), 0, len(polygon_parts)]],
dtype=np.int32,
)
attributes = np.array(
[(b'MultiRingRegion',)],
dtype=np.dtype([('Name', 'S32')]),
)
with h5py.File(fixture_hdf, 'w') as hdf_file:
geometry = hdf_file.create_group('Geometry')
infiltration = geometry.create_group('Infiltration')
infiltration.create_dataset('Attributes', data=attributes)
infiltration.create_dataset('Polygon Info', data=polygon_info)
infiltration.create_dataset('Polygon Parts', data=polygon_parts)
infiltration.create_dataset('Polygon Points', data=polygon_points)
regions_gdf = HdfInfiltration.get_infiltration_region_polygons(fixture_hdf)
assert len(regions_gdf) == 1, 'Expected one infiltration region from regression fixture.'
fixture_geom = regions_gdf.geometry.iloc[0]
assert fixture_geom.geom_type == 'Polygon'
assert len(fixture_geom.interiors) == 1, 'Expected one interior ring.'
assert np.isclose(fixture_geom.area, 12.0), f'Unexpected polygon area: {fixture_geom.area}'
print(f'Regression fixture: {fixture_hdf.name}')
print(f'Geometry type: {fixture_geom.geom_type}')
print(f'Interior rings: {len(fixture_geom.interiors)}')
print(f'Area: {fixture_geom.area:.2f}')
regions_gdf[['region_id', 'Name', 'geometry']]
Text Only
Regression fixture: 211_infiltration_multipart_fixture.g01.hdf
Geometry type: Polygon
Interior rings: 1
Area: 12.00
| region_id | Name | geometry | |
|---|---|---|---|
| 0 | 0 | MultiRingRegion | POLYGON ((0 0, 4 0, 4 4, 0 4, 0 0), (1 1, 1 3,... |
Key Takeaways¶
BaldEagleCrkMulti2Dprovides a reproducible real-project source for preprocessed Manning's N and infiltration values.- The notebook now includes an executable regression for multi-ring infiltration override polygons.
- Running this notebook end to end validates both the example workflow and the polygon-construction fix.