Local Level Lab

Airborne measurements. Open methods.

First time-tagged airborne IMU sample

This ILVIS0 file contains a real airborne IMU stream over Greenland on 14 April 2009: 132,444 time-tagged IMU packets at approximately 200 Hz, spanning 662.226 seconds (about 11 minutes), together with navigation and primary GPS receiver data. All 143,471 outer packet checksums pass. No unframed bytes or IMU timestamp gaps above 7.5 ms were found.

The instrument identifies AV-510 VER5, firmware04.60, ICD15.00 and IMU8. The public 2014 interface document matches observed outer packet layouts but describes the IMU payload as 24 bytes without physical scaling and axis definitions. It is a later version than this instrument. The initial inspector therefore preserved the payload exactly. A subsequent independent validation on 7 October 2026 reproduced six signed int32 increments, with delta-V scale 2^-14 m/s/count and delta-angle scale 2^-18 rad/count, and tested the axis order/signs against Group-1 fused navigation. See the separate validation record. The original inspector and artifacts below retain their historical opaque interpretation. This particular sample includes climb and maneuvers and is not a qualifying cruise stretch. The later archive analysis uses separately selected recordings.

The instrument belongs to professional airborne survey hardware, likely the LN200ROM fiber-optic IMU family used in POS AV 510 systems. This is an appropriate instrument class for investigating slow turning. Published system performance is available; the exact hardware variant and remaining unit-specific calibration are not identified by the type byte. The instrument summary explains the published capabilities and the measurements verified here.

Reassembling the primary GPS stream across packet boundaries recovers 660 checksum-valid GGA position messages, 660 VTG motion messages and 660 ZDA date/time messages. All GGA fix-quality values are 1. ZDA confirms the date. Under the compatible public header definition, the first timestamp is UTC seconds of the week and the second is instrument time since power-on. Receiver fix timing, IMU latency and lever arms still need verification.

The 662 fused navigation records describe a mainly southbound track from about 79.72°N to 78.86°N. Altitude rises from about 6.9 to 7.5 km, and roll reaches approximately −17.5°. This segment includes a climb and maneuvers; it is not an already qualified level-cruise window. Navigation-derived orientation/rates must not substitute for independent gyro data.

Preserved files:

Reproduce into a new output directory using the original downloaded file:

/home/sweisman/venv/bin/python analysis/tests/inspect_ilvis0_sample.py --input /home/sweisman/Downloads/ILVIS0_gyro_54935_atm_applanix_14Apr09.013 --output-dir /tmp/ilvis0-reinspection

The inspector fails on malformed outer framing/checksums and never repairs or interpolates. Two focused checks pass: whole-frame checksum convention/truncation, and receiver sentence reconstruction/checksum/signed coordinate conversion. No full suite, campaign or model fit ran.

Next check additional .013 logs for compatibility and useful level-flight windows, and investigate onboard corrections, mounting and latency. The geometry-only one-minute screen finds no qualifying interval in this reference sample. Its new bulk derivative files were discarded, while this historical corpus and the original source remain intact. The airborne inspection notes describe the remaining documentation questions.

Data citation: Hofton, M. & Blair, J. B. (2011). IceBridge LVIS L0 Raw Ranges, Version 1. NASA NSIDC DAAC. DOI10.5067/E6JPQ3QNW77R. Subset: ILVIS0_gyro_54935_atm_applanix_14Apr09.013, inspected6October2026.