d = c_s / (2 f_1) = 25.0 Β± 5.0 m after APS detrending. SAR measures surface micro-motion; depth is a model inference.
Loading...
Supports JSON (.json), ESA SNAP XML (.dim, .xml), Matrix (.csv), or MiniSEED
{
"title": "Mt. Etna Northeast Crater Volcanic Void",
"satellite": "COSMO-SkyMed CSK2 X-Band",
"site_name": "Mt. Etna Caldera, Sicily, Italy",
"wgs84_coords": { "lat": 37.7510, "lon": 14.9934, "elevation_m": 3326.0 },
"utm_zone": "33S (WGS84)",
"orbit_pass": "Ascending (Track 112)",
"look_angle_deg": 36.5,
"heading_deg": 350.2,
"terrain_type": "Basaltic Lava & Volcanic Conduit",
"geological_target": "Subsurface Magma Reservoir (42.5 m)",
"detected_f0_hz": 5.0,
"shear_velocity_mps": 850.0,
"amplitude_um": 180.0
}
Drop this snippet directly into your team's Python SAR processing pipeline:
from connectors.cosmo_skymed import COSMOSkyMedConnector
from plugins.tded_plugin import TDEDPlugin
# 1. Connect to SLC SAR Product (synthetic calibration target shown;
# pass your own complex SLC matrix + azimuth timeline for real data)
connector = COSMOSkyMedConnector(satellite_id="CSK1")
slc_data = connector.create_synthetic_slc_product(
n_azimuth_lines=1000,
n_range_samples=400,
target_vibration_hz=8.5,
target_displacement_m=152.7e-6
)
# 2. Full TDED run: demodulation, ladder classification, integrity seal
tded = TDEDPlugin(default_shear_velocity_mps=850.0)
result = tded.process_cosmo_skymed_slc(slc_data)
inv = result["elastodynamic_inversion"]
print(f"Peaks: {result['vibrometry']['detected_peaks_hz']} Hz")
print(f"Model: {inv['boundary_model']} | d = {inv['estimated_depth_m']} Β± {inv['depth_uncertainty_m']} m")
print(f"Void-consistent: {inv['is_void_consistent']} | Seal: {result['scientific_integrity_audit']['coherence_seal_sha256'][:23]}β¦")
Stream real continuous seismic waveforms via ObsPy (EarthScope FDSN):
from obspy.clients.fdsn import Client
from obspy import UTCDateTime
import numpy as np
client = Client("IRIS")
t0 = UTCDateTime("2026-08-28T00:00:00")
st = client.get_waveforms("IU", "ANMO", "00", "BHZ", t0, t0 + 60)
st.detrend("demean")
st.filter("bandpass", freqmin=1.0, freqmax=15.0)
# Extract ground displacement in micrometers
velocity_m_s = st[0].data * 1e-9
displacement_um = np.cumsum(velocity_m_s) * (1.0 / st[0].stats.sampling_rate) * 1e6
print(f"β
Ingested {len(displacement_um)} real GSN samples from ANMO station")
Download real Copernicus Sentinel-1 Level-1 SLC products via ASF Search API:
import asf_search as asf
# Search Sentinel-1 SLC bursts over Campi Flegrei / Norcia
wkt = "POINT(14.1394 40.8272)"
results = asf.geo_search(
platform=asf.PLATFORM.SENTINEL1,
processingLevel=asf.PRODUCT_TYPE.SLC,
intersectsWith=wkt,
maxResults=5
)
print(f"Found {len(results)} Sentinel-1 SLC scenes. Downloading burst...")
# results.download(path="./data/sentinel1_slc")
Execute this graph directly via ESA SNAP gpt command-line:
gpt TDED_SNAP_Workflow.xml -PinputSLC=CSK_RAW_SCS.h5 -PoutDim=TDED_INVERSION.dim
Live REST Endpoint for remote worker batch ingestion:
curl -X POST https://state.catalyst-brain.strategic-innovations.ai/api/v1/tded/ingest \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $CATALYST_API_KEY" \
-d '{"mission":"csk", "target_f0_hz":8.5, "shear_velocity_mps":850.0}'
Cryptographically signed deep-link snapshot for Dr. Biondi's research team:
Satellites such as COSMO-SkyMed (X-band 9.6 GHz, Ξ»=3.12 cm) and Sentinel-1 (C-band 5.4 GHz, Ξ»=5.55 cm) illuminate the Earth from 619β693 km in orbit.
Radar microwaves do not penetrate solid rock kilometers deep directly (skin depth in conductive rock is limited to centimeters). Instead, this technique measures surface micro-vibrations:
d = c_s / (2 fβ); a layer over stiff bedrock β quarter-wave odd ladder (fβ, 3fβ, 5fβ) and d = c_s / (4 fβ) β a stiff interface, never a void.Why Scientists Are Skeptical: Classical InSAR and raw Doppler FFT suffer from Atmospheric Phase Screen (APS). Turbulent water vapor in clouds causes massive phase delays (10Γ larger than real micro-vibrations), creating catastrophic false positives that look like fake underground pyramids or giant cavities.
The TDED Solution:
Choose an orbital preset (Campi Flegrei, Norcia, Albuquerque) or drag-and-drop your custom SAR product or seismic record via π€ Upload Dataset.
Click β‘ Run Inversion to execute 16,384-pt chirp demodulation, filter tropospheric APS, and invert the depth plane ($Z$).
Use π― Coordinate Scan to read live WGS84 Lat/Lon and radar backscatter $\sigma_0$. Click π₯ Export SNAP XML or π Team Connector to export ready-to-run Python/SNAP pipelines for your team.
What is 100% Real Live Data in This Visualizer:
How to Ingest Raw Multi-GB Space Agency Products:
Full Copernicus Sentinel-1 `.SAFE` and COSMO-SkyMed `.h5` files are 4β8 GB archives. Use our provided ObsPy and ASF Search connector scripts (in π Team Connector) to stream raw bursts directly into this engine without manual conversion!
Reflectivity Contrast (SCR): +24.5 dB
Matrix Centroid: Row 60, Col 80 (160Γ120 grid)
Detection Algorithm: Cell-Averaging Constant False Alarm Rate (CA-CFAR) with 12 guard cells & 24 background reference cells.
Displacement Amplitude ($u_0$): 118.4 ΞΌm
Microwave Phase Swing ($\Delta \phi$): 1.42 rad (X-Band)
Tropospheric APS Filter: 2nd-order spatial polynomial carrier removal detrended $3.8\,\text{rad}$ atmospheric turbulence.
Fundamental Eigenmode ($f_0$): 22.40 Hz
Harmonic Ladder Detected: [22.40, 44.80, 67.20] Hz
Boundary Model: half_wave_free_base (Integer 1:2:3)
Inverted Overburden Depth: 32.37 Β± 3.2 m (via $d = \frac{c_s}{2 f_1}$ in limestone)
Causal Coherence ($\Gamma^*$): 1.0000 ($\ge 4.808\sigma$ Bonferroni Barrier)
False-Positive Risk: LOW (< 0.1%)
Cryptographic Seal: sha256:736b7b49...
In-Situ Archaeological / Geological Correlation: