Skip to content

canvod-vod

Purpose

The canvod-vod package implements vegetation optical depth (VOD) estimation from GNSS signal-to-noise ratio (SNR) data. It provides the core scientific algorithms for the canVODpy analysis pipeline.

Physical Background

Tau-Omega Radiative Transfer Model

canvod-vod implements the zeroth-order tau-omega model for GNSS-Transmissometry, following Humphrey & Frankenberg (2022).

A reference receiver in open sky and a canopy receiver beneath vegetation observe the same satellite at polar angle θ. The SNR difference encodes the two-way canopy attenuation:

\[\Delta\text{SNR} = \text{SNR}_\text{canopy} - \text{SNR}_\text{reference}\]
\[T = 10^{\Delta\text{SNR} / 10} \qquad \text{(linear transmissivity)}\]
\[\text{VOD} = -\ln(T) \cdot \cos\theta\]

Assumptions: - Single-scattering approximation (no multiple canopy reflections) - Plane-parallel canopy layer - Attenuation proportional to path length through canopy

Usage

from canvod.vod import TauOmegaZerothOrder

calculator = TauOmegaZerothOrder(canopy_ds=canopy_ds, sky_ds=sky_ds)
vod_result = calculator.calculate_vod()

from canvod.vod import TauOmegaZerothOrder

vod_result = TauOmegaZerothOrder.from_datasets(
    canopy_ds=canopy_ds,
    sky_ds=sky_ds,
    align=True,  # auto-align epochs before differencing
)

from canvod.vod import TauOmegaZerothOrder

vod_result = TauOmegaZerothOrder.from_icechunkstore(
    icechunk_store_pth="path/to/store",
    canopy_group="canopy_01",
    sky_group="reference_01",
)

Input requirements

Both datasets must contain an SNR data variable. Both should be augmented with spherical coordinates (θ, φ) from canvod-auxiliary and assigned to grid cells by canvod-grids.

Output

calculate_vod() returns an xr.Dataset containing:

Variable Description
VOD Vegetation optical depth
phi Azimuth angles (from canopy dataset)
theta Polar angles (from canopy dataset)

Multi-Receiver SCS Expansion

When a single reference receiver serves multiple canopy positions, satellite geometry must be recomputed relative to each canopy location. The scs_from configuration in sites.yaml drives this expansion.

flowchart TD
    subgraph FIELD["Field Setup"]
        C1["`**Canopy 1**
        NE position`"]
        C2["`**Canopy 2**
        SW position`"]
        R1["`**Reference**
        open sky`"]
    end

    subgraph STORE["RINEX Store Groups"]
        SG1["canopy_01"]
        SG2["canopy_02"]
        SG3["`**reference_01_canopy_01**
        geometry at canopy_01 pos.`"]
        SG4["`**reference_01_canopy_02**
        geometry at canopy_02 pos.`"]
    end

    subgraph VOD["VOD Analysis Pairs"]
        VP1["canopy_01 vs reference_01_canopy_01"]
        VP2["canopy_02 vs reference_01_canopy_02"]
    end

    C1 --> SG1
    C2 --> SG2
    R1 --> SG3
    R1 --> SG4
    SG1 --> VP1
    SG3 --> VP1
    SG2 --> VP2
    SG4 --> VP2
    VP1 --> V1["VOD @ canopy_01"]
    VP2 --> V2["VOD @ canopy_02"]

Configuration

# config/sites.yaml
receivers:
  reference_01:
    type: reference
    scs_from:
      - canopy_01    # recompute geometry at canopy_01 position
      - canopy_02    # recompute geometry at canopy_02 position

References

Humphrey, V. and Frankenberg, C. (2022). GNSS-transmissometry: A new approach for vegetation optical depth estimation. Remote Sensing of Environment.