Running an Radiometric Analysis

Here is a detailed description on how to execute a Radiometric Analysis with SCT in three ways:

1. From the Command Line Interface (CLI) → user-facing execution
2. From the Python API → programmatic/scripting/orchestration integration
3. From the Testing Interface → automated validation

From CLI

Several radiometric analyses have been implemented with a dedicated command that can be specified to perform the operation of choice. They have been grouped under the radiometry CLI group.

NESZ

NESZ analysis can be performed using the following command just specifying the input product and the output directory.

sct [--config <path_to_config>] radiometry nesz -p <product_path> -out <output_dir> [-g]

where -g is used to enable graphs generation (if added to the command).

Average Elevation Profiles

Average Elevation Profiles analysis can be performed using the following command just specifying the input product, the output directory and the desired output radiometric quantity.

sct [--config <path_to_config>] radiometry elevation-profiles -p <product_path> -out <output_dir> -r <radiometric_quantity> [-g]

where -g is used to enable graphs generation (if added to the command) and <radiometric_quantity> is the desired output radiometric quantity (sigma, gamma or beta).

Rain Forest

Rain Forest analysis can be performed using the following command just specifying the input product and the output directory. This analysis is just a shortcut for radiometry elevation-profiles -r gamma where the output radiometric quantity is automatically set to gamma.

sct [--config <path_to_config>] radiometry rain-forest -p <product_path> -out <output_dir> [-g]

Scalloping

Scalloping analysis can be performed using the following command just specifying the input product and the output directory.

sct [--config <path_to_config>] radiometry scalloping -p <product_path> -out <output_dir> [-g]

where -g is used to enable graphs generation (if added to the command).

From Python API

These analyses is also available directly from Python without using the CLI, and can be imported in any python script and executed as follows:

from pathlib import Path
from perseo_quality.core.generic_dataclasses import SARRadiometricQuantity
from sct.analyses.radiometry.main import (
    full_nesz_analysis,
    full_average_elevation_profiles_analysis,
    full_scalloping_analysis
)
from sct.analyses.radiometry.config import SCTRadiometricAnalysisConfig

config = SCTRadiometricAnalysisConfig()  # this is the default, but parameters can be set here

output_netcdf_file, output_kpi_file = full_nesz_analysis(
    product=Path("path/to/product"),
    output_directory=Path("path/to/output_directory"),
    config=config,  # optional, can be None
    graphs=True,  # optional, can be False
)

output_netcdf_file, output_kpi_file = full_average_elevation_profiles_analysis(
    product=Path("path/to/product"),
    output_radiometric_quantity=SARRadiometricQuantity.SIGMA,  # GAMMA for Rain Forest
    output_directory=Path("path/to/output_directory"),
    config=config,  # optional, can be None
    graphs=True,  # optional, can be False
)

output_netcdf_file, output_kpi_file = full_scalloping_analysis(
    product=Path("path/to/product"),
    output_directory=Path("path/to/output_directory"),
    config=config,  # optional, can be None
    graphs=True,  # optional, can be False
)

From Testing Interface

The radiometric analysis implementation also provides a testing interface, which can be used to run the analysis from the SCT Testing Framework. To do so, a registry .json file containing test cases must be created and provided to the testing CLI interface.

{
    "product-format-type": {
        "test-case-1": {
            "analysis": "radiometry-nesz",
            "product": "path/to/product/1",
            "config": "path/to/config/if/needed",
            "reference_output": [
                "path/to/output/reference/results/kpi_stats.csv",
                "path/to/output/reference/results/NESZ_profiles.nc"
            ]
        },
        "test-case-2": {
            "analysis": "radiometry-rain-forest",
            "product": "path/to/product/2",
            "config": "path/to/config/if/needed",
            "reference_output": [
                "path/to/output/reference/results/kpi_stats.csv",
                "path/to/output/reference/results/RAIN_FOREST_profiles.nc"
            ]
        },
        "test-case-3": {
            "analysis": "radiometry-scalloping",
            "product": "path/to/product/3",
            "config": "path/to/config/if/needed",
            "reference_output": [
                "path/to/output/reference/results/kpi_stats.csv",
                "path/to/output/reference/results/SCALLOPING_profiles.nc"
            ]
        }
    }
}

Then the testing CLI interface can be used as follows:

sct testing test -r <path_to_registry_file> -out <output_dir> [-g]

When a reference_output is provided, the testing CLI interface will compare the results with the reference output and report any discrepancy.