Skip to content

Full Parameter Analysis — WI937

Filter → Echo Parameters → 3-D Drift Velocity

End-to-end diagnostic workflow: load a WI937 RIQ sounding, apply the six-stage IonogramFilter, then compute and compare all Dynasonde parameters (amplitude, EP, PP, V*) and 3-D drift velocity [Vx, Vy, Vz] between raw and filtered echo clouds.

Script: examples/vipir/ionogram_full_analysis_wi937.py

Pipeline overview

WI937_2022233235902.RIQ
        │
        ▼
  EchoExtractor          →  df_raw  (all echoes)
        │
        ▼
  IonogramFilter         →  df_filt (noise-rejected echoes)
  (stages 1–5, temporal skipped)
        │
        ├──► fit_drift_from_df(df_raw)   →  df_vel_raw
        └──► fit_drift_from_df(df_filt)  →  df_vel_filt
                                                │
                                                ▼
                                     3×3 diagnostic figure

Steps

1 — Load and extract echoes

from pynasonde.vipir.riq.echo import EchoExtractor
from pynasonde.vipir.riq.parsers.read_riq import VIPIR_VERSION_MAP, RiqDataset

riq = RiqDataset.create_from_file(
    "examples/data/WI937_2022233235902.RIQ",
    unicode="latin-1",
    vipir_config=VIPIR_VERSION_MAP.configs[1],  # vipir_version=0 / data_type=1
)

extractor = EchoExtractor(
    sct=riq.sct, pulsets=riq.pulsets,
    snr_threshold_db=3.0,
    min_height_km=60.0,
    max_height_km=1000.0,
    min_rx_for_direction=3,
    max_echoes_per_pulset=5,
)
extractor.extract()
df_raw = extractor.to_dataframe()

2 — Filter (stages 1–5)

from pynasonde.vipir.riq.parsers.filter import IonogramFilter

filt = IonogramFilter(
    rfi_enabled=True,       rfi_height_iqr_km=300.0,  rfi_min_echoes=3,
    ep_filter_enabled=True, ep_max_deg=90.0,
    multihop_enabled=True,  multihop_orders=(2, 3),
    multihop_height_tol_km=50.0, multihop_snr_margin_db=6.0,
    dbscan_enabled=True,    dbscan_eps=1.0, dbscan_min_samples=5,
    dbscan_features=(
        "frequency_khz", "height_km",
        "velocity_mps", "amplitude_db", "residual_deg",
    ),
    ransac_enabled=True,    ransac_residual_km=100.0,
    ransac_min_samples=10,  ransac_n_iter=200,
    ransac_poly_degree=3,   ransac_min_inlier_fraction=0.3,
    temporal_enabled=False,
)

df_filt = filt.filter(extractor)
print(filt.summary())

3 — Height-binned drift velocity (raw and filtered)

The script includes a standalone fit_drift_from_df() helper that operates directly on a DataFrame rather than an EchoExtractor object — useful when comparing pre- and post-filter echo sets without re-extracting:

df_vel_raw  = fit_drift_from_df(df_raw,  height_bin_km=50.0)
df_vel_filt = fit_drift_from_df(df_filt, height_bin_km=50.0)

fit_drift_from_df() parameters:

Parameter Default Description
height_bin_km 50.0 Bin width (km)
min_echoes 6 Minimum echoes required per bin
snr_weight True SNR-weighted least-squares
n_sigma 2.5 Sigma-clipping threshold

4 — Output figure (3×3)

Saved to docs/examples/figures/ionogram_full_analysis_wi937.png:

Panel Contents
(A) Raw ionogram Frequency vs height, amplitude colourmap
(B) Filtered ionogram Same axes after noise rejection
(C) Amplitude vs height Grey = raw, blue = filtered
(D) EP vs height Grey = raw, orange = filtered; dashed vertical at EP threshold
(E) PP (polarization) vs height Grey = raw, green = filtered
(F) V* (LOS velocity) vs height Grey = raw, red = filtered
(G) Vx/Vy/Vz vs height — raw Three-component drift from unfiltered echoes
(H) Vx/Vy/Vz vs height — filtered Three-component drift from filtered echoes
(I) Fit quality: raw vs filtered RMS LOS residual and echo count per bin, dual x-axes

Full analysis WI937

WI937 2022-233 23:59 UT. 3×3 diagnostic comparing raw (grey) and filtered (coloured) echo parameters and the resulting 3-D drift velocity.

Run

cd /home/chakras4/Research/CodeBase/pynasonde
python examples/vipir/ionogram_full_analysis_wi937.py

Key observations

  • Panels A vs B: the filtered ionogram retains the coherent F-layer trace while removing diffuse scatter above the MUF and low-height clutter.
  • Panels D (EP): the EP filter (ep_max_deg=90°) removes only strongly non-planar returns; most oblique echoes (EP 50–80°) are preserved.
  • Panels G vs H: filtering tightens the height-resolved drift estimates — the RMS LOS residual (panel I) drops substantially in height ranges dominated by multi-hop or RFI contamination.