{ "schema": "ausmt-station", "version": "0.1", "ausmt_id": "au.cicada97-1997-98.Solo", "station": "Solo", "survey": "CICADA97 induction array 1997-98", "survey_id": "cicada97-1997-98", "country": "Australia", "organisation": "Australian National University", "location": { "lat": -35.821667, "lon": 150.581667 }, "data": { "type": "GDS", "n_periods": 20, "period_min_s": 10.0, "period_max_s": 6826.870563 }, "diagnostics": { "median_relative_error": null, "remote_reference": false, "tipper_available": true, "completeness_smoothness_diagnostic": { "value": null, "basis": "s", "note": "not a quality or geological-value judgement" }, "method": "phase-tensor (Caldwell 2004)", "note": "screening diagnostic, not an interpretation product" }, "processing": { "software": null, "algorithm": null, "remote_reference": false, "remote_site": null, "file_written_by": { "name": null, "version": null }, "note": "Tipper transfer functions minted from Hitchman, A. P. (1999) 'Interactions\nbetween aeromagnetic data and electromagnetic induction in the earth', PhD\nthesis, The Australian National University, hdl:1885/147975, Table B.2\n(printed 169 to 172), with station positions and elevations from Table 6.2\n(printed 47). Transcription:\nGDS/inventories/hitchman1999_cicada97_table_b2_transfer_functions.csv and\nGDS/inventories/hitchman1999_cicada97_stations.csv, both read off 300 dpi\npage renders rather than off the scan's text layer.\n\nREFERENCE SITE, AND WHAT IT MEANS FOR MAGNITUDE. These are NOT local\ntippers. The Table B.2 caption and Chapter 6 (printed 50) state that the\nhorizontal variations recorded at CNB, the Canberra Magnetic Observatory,\nwere used to represent the source field at every site, so each value relates\nthe LOCAL vertical field to the horizontal field at CNB rather than to the\nhorizontal field beneath the instrument:\nz(t) at this site = P . x(t) at CNB + Q . y(t) at CNB\nA local tipper is bounded near unity by the physics of a layered earth. A\nremotely referenced transfer function is not, because the source-field ratio\nbetween two separated sites is itself part of the quantity. Magnitudes near\nand above 1.0 are therefore EXPECTED here and are not defects: the largest\nreal arrow in this campaign is 1.33, at 1280 s at the continental-slope site\nSolo, which holds above 1.0 at every period from 426.7 s upward. Do not\ncompare these values against a locally referenced tipper without carrying\nthe reference. CNB itself is in this survey and is self-referenced, so its\nown values are a measure of the observatory against itself and are the\nsmallest in the campaign.\n\nCONVENTION, AND THE TRAP IT AVOIDS. Every row of Table B.2 prints an\ninduction arrow (Lr, theta_r, Lq, theta_q) beside the transfer functions it\nderives from (Pr, Pq, Qr, Qq). The thesis symbol list (printed xv to xvi)\nand Equation 3.2 (printed 27) define P as the geographic-NORTH transfer\nfunction and Q as the geographic-EAST one, with subscript r the real part\nand q the quadrature part, in the frame with x geographic north, y\ngeographic east and z positive down. That is the EDI tipper's own frame, so\nthe mapping is direct and no sign is changed:\nTXR = Pr TXI = Pq TYR = Qr TYI = Qq\nSection 5.2.4 (printed 42) says the ARROWS were reversed 'in accordance with\nthe convention that Parkinson induction arrows point toward a nearby\nconductor'. The reversal therefore lives in the printed arrow azimuth and\nNOT in P and Q. Tested over all 129 rows of this table whose real arrow\nreaches 0.10, the printed theta_r equals atan2(Qr, Pr) plus 180 degrees in\nevery single case and equals it without the 180 in none, and the 90\nqualifying quadrature rows behave the same way. Deriving a tipper from the\narrow columns would publish every value in this survey rotated by 180\ndegrees. The arrow columns are not used here.\n\nThe result is checked against physics and against this collection. Taking\nthe printed P and Q as written, the Parkinson form of the real arrow points\nbetween 129 and 149 degrees true at every land site and at Solo, that is,\noffshore into the Tasman Sea, and its magnitude climbs from 0.18 inland to\n0.68 at the coast and 1.23 at the shelf edge before collapsing to 0.14 and\n0.05 at the two abyssal sites, which sit beneath the conductor rather than\nbeside it. That is the coast effect. Compared against the compilation\nregister's own arrows at the six re-occupied sites, over periods of an hour\nand longer, the two occupations agree to within 0.5 to 4.5 degrees at the\nfour sites nearest the coast, and to 8 and 18 degrees at the two inland\nsites, where the arrow is shortest and its azimuth least determined. A\nreversed convention would show as a disagreement near 180 degrees, and\nnothing in that comparison is anywhere near it.\n\nFRAME: geographic. Table B.2 was derived in the geomagnetic frame and\nrotated to geographic using the AGRF95 declinations of Table B.1, so the\nrotation is the thesis's own and this generator applies none. Arrow\nazimuths in the printed table are degrees clockwise from geographic north.\n\nUNCERTAINTIES. Table B.2 prints two error columns, both headed 'e', one\ninside the P group and one inside the Q group. The caption states that each\nis a STANDARD ERROR and that it applies to BOTH the real and the quadrature\npart of its own pair. TXVAR.EXP and TYVAR.EXP hold a variance, so this file\nwrites\nTXVAR.EXP = e_P squared TYVAR.EXP = e_Q squared\nand nothing else: there is no confidence-limit conversion, because the table\nprints none, and no split between the real and quadrature parts, because the\ncaption forbids one. These are real published uncertainties, not a constant\nfill, and they are the reason the variance blocks are present here where\nthey are absent from other minted files in this collection.\n\nERROR FLOOR. Twenty-one of this table's error cells print 0.00. That is a\ntwo-decimal rounding of an error below 0.005, not a claim of perfect\ndetermination, and fourteen of them are at CNB, the reference site itself.\nA printed 0.00 is written here as a variance of 0.005 squared, that is\n2.5E-05, which is the smallest error the print could have rounded to zero\nfrom. Writing a zero variance would publish an infinite weight on a rounded\nnumber. The substitution is applied to those cells only; every other\nvariance in this file is the printed error squared, untouched.\n\nIMPEDANCE AND CHANNELS. None. CICADA97 deployed three-component\nmagnetometers and recorded no electric field, so this file declares HX, HY\nand HZ only and carries no impedance block, no apparent resistivity, no\nphase and no electric channel of any kind. Nothing here may be read as\nmagnetotelluric data.\n\nCAMPAIGN: CICADA97, a ten-site vertical-field induction array run from the\nResearch School of Earth Sciences of The Australian National University down\nthe New South Wales south coast and out onto the Tasman abyssal plain. Land\nrecording opened 30 September 1997 with a second deployment from 18 October\n1997, and BBB was re-occupied 13 to 18 February 1998. Land instruments were\nFlinders three-component ringcore fluxgates after Chamalaun and Walker\n(1982) sampling at 5 s; CNB is a Narod fluxgate at 1 s. No per-station\nacquisition date is printed, so ACQDATE is deliberately absent rather than\nfilled with a boilerplate value.\n\nSTATION: Solo, thesis site code Solo, Continental slope, Tasman Sea,\noffshore New South Wales. Table 6.2 position -35.82167, 150.58167, elevation\n-523 m. The 20 rows are this station's block of Table B.2, 6826.7 s down to\n10.0 s.\n\nCOORDINATES: GPS, Table 6.2.\n\nELEV is the sea-floor depth of 523 m below sea level, sounded by RV Franklin\nand printed in Table 6.2. It is a depth, not a land elevation.\n\nCAVEAT ON USE, THIS STATION ONLY. This record carries a documented\ninstrument defect that the printed standard errors do not express. The 11\nbands 320.0 s and shorter are published so that the table is not silently\nshortened, but they are NOT sound and they must not be used at the weight\ntheir variances state. Weight them down or exclude them. The affected\nperiods are exactly these, in seconds, and no other row in this file is\naffected:\n320.0 213.3 160.0 106.7 80.0 53.3 40.0 26.7 20.0 13.3 10.0\nThe 9 bands 6826.7 s down to 426.7 s stand, and carry the coast effect this\nsite was deployed to measure.\n\nWHAT IS WRONG WITH THEM. Reading only the bands whose printed error does\nnot already exceed the component it covers, the real arrow drifts smoothly\nfrom 6826.7 s down to 213.3 s, a total swing of 23.2 degrees, and then\nreverses through about 180 degrees: at 160.0 s it turns 161.9 degrees from\nthe band above it, and below 160.0 s the band-to-band change reaches 155.6,\n130.0 and 158.7 degrees. Across the 7 affected bands whose real arrow still\nreaches 0.10 and so still carries a direction, the printed standard errors\nstay between 0.04 and 0.08 and flag none of those rows. At 106.7 s and 53.3\ns the arrow points 328.6 and 162.5 degrees, 166.1 degrees apart, while their\nreal north components differ by 0.95 against a combined standard error of\n0.06, about 15 standard errors. One site cannot hold both as a response.\n\nAND IT BITES BEFORE IT SHOWS. At 320.0 s, 213.3 s and 160.0 s the\nquadrature arrow is LONGER than the real one, 1.21 against 0.63, 1.09\nagainst 0.15 and 0.90 against 0.44, where Hitchman states at printed page 50\nthat the quadrature response at these sites is weak relative to the real.\nThose bands are already corrupt while their azimuths still sit on the drift\nabove, so a consumer who keeps rows on the look of the arrow alone will keep\nthe worst of them.\n\nTHE DOCUMENTED CAUSE, printed page 49. The instrument was inadvertently\nprogrammed to sample faster than it could sustain, so every write of a block\nto memory lost the following 12 samples, that is 60 s at the 5 s sampling\ninterval. Blocks were written every 78 records, that is every 390 s, so\nabout 15 per cent of the record was lost on a strict periodic beat. The\nmissing points were then filled by inference from the corresponding\ncomponent at CNB, and CNB is the same site that supplies the reference\nhorizontal field for every transfer function in this file, so the fill\nenters both sides of the estimate. This site also ran for 5 days against 15\nat the land sites.\n\nTHE AUTHOR SAYS SO, printed page 50. The response at periods less than 200\ns 'is difficult to determine, however, due to the inconsistent length and\norientation of the arrows'. Those bands were nonetheless plotted unmarked\nin Figure B.1 and in Figure 6.4, and the warning was put in prose only.\nFigure B.1 carries no error bars, no confidence marks and no rejection marks\nof any kind: it replots Table B.2 point for point, and it cannot be cited as\nevidence that these rows are good.\n\nTHE PRINTED ERRORS DO NOT CAPTURE THIS, which is why this note exists. The\nvariance blocks in this file are the printed standard errors squared and\nnothing else, so across the affected bands they state a precision of a few\nper cent on values that cannot all be a response. A variance-weighted\nconsumer reading this file without reading this note will treat those rows\nas strong and well determined. They are not.\n\nTHE SOURCE-FIELD EXPLANATION IS REFUTED, so do not reach for it. Hitchman's\nown criterion is that beta below 2.0 indicates amphidromic conditions. Beta\nhere is 11.5, 10.1, 10.9, 9.3, 7.2, 2.4 and 4.6 over the 7 affected bands\nabove, and never falls below 2.0. DRS, the station this thesis demonstrates\nto be amphidromic, has beta 2.8, 2.6, 2.4, 1.9, 1.6, 1.6 and 1.5 at the same\nperiods, below 2.0 at 4 of them, and its real arrow holds between 175.9 and\n188.0 degrees across all 7 with no band-to-band change larger than 4.2\ndegrees. Amphidromic conditions in this campaign do not produce reversals,\nand the station that does produce them is not amphidromic.\n\nTHE 9 RETAINED BANDS ARE NOT UNAFFECTED EITHER. A phase-related amplitude\nuncertainty that the printed standard errors do not carry grows as the\nperiod falls, and is already of order 20 to 30 per cent by 426.7 s. Treat\nthe printed errors on those bands as a lower bound. The real arrow\nexceeding 1.0 at long period is NOT part of this defect: that is the\nCNB-referenced coast effect set out above, and it is expected here.\n\nRESOLUTION: at 10.0 s, 13.3 s the printed standard error reaches or exceeds\nthe component it covers, so those rows constrain little. They are published\nrather than gated out, because the variance blocks in this file already\ncarry that message and a silently shortened table would hide the band over\nwhich this station resolved nothing.\n\nNEW SITE: this position is not a re-occupation. Measured when this file is\nwritten, the nearest station already published from the Tasman Project,\nwhich is this collection's only other coverage of this transect on land and\non the sea floor alike, is au.tasman-land-1983-84.Durras, 31.97 km away.\nFour of this campaign's ten sites are new, this among them; the other six\nre-occupy Tasman Project land stations.\n\nThis file is published as station au.cicada97-1997-98.Solo. Dates in this\nfile are ISO 8601." }, "distribution": { "edi_available": true, "license": "CC-BY-4.0", "edi_path": "edi/cicada97-1997-98/Solo.edi" }, "provenance": { "pipeline": "ausmt/extract.build_portal", "pipeline_version": "0.2.1", "extractor": "mt_metadata (community canonical)", "software": { "python": "3.12.14", "mt_metadata": "1.0.10", "mth5": "0.6.8" }, "git_commit": "25c90921c42c72bbc5455b4bbdab9137cf8b8e48", "parameters": { "dimensionality": { "beta_per_period_deg": 3.0, "skew_3d_deg": 5.0, "pct_periods_3d_threshold": 40, "ellip_2d_deg": 0.1, "min_rez_row_sine": 0.01, "beta_physical_cap_deg": 15.0, "min_usable_period_frac": 0.5, "skew_aggregation": "median" }, "diagnostic": "completeness/smoothness (median rel error + coverage + smoothness)" }, "generated": "2026-09-09T06:22:47.791693+00:00", "input_file": "Solo.edi", "input_sha256": "1a7b685502649927b434d9ea3683748c74bfa6f9e4fcbbfac7db847ce046a4ef" }, "coordinate_qc": null, "canonical_conditioning": null, "frame": { "evidence": { "branch": "mt", "zrot": null, "trot": null, "rotspec": null, "hmeas_azimuths": { "hx": 0.0, "hy": 90.0 }, "emeas_azimuths": { "ex": null, "ey": null } }, "impedance_rotation_deg_source": null, "tipper_rotation_deg_source": null, "derotated": false, "frame_served": "declared-zero", "declared_azimuth_deg": 0.0, "convention_check": { "verdict": "insufficient", "phs_xy_median_deg": null, "phs_yx_median_deg": null, "n_periods_used": 0, "detail": "only 0 usable period(s) (< 5) — no convention verdict" } }, "resources": [ { "id": "edi", "kind": "transfer_function", "format": "edi", "provenance_role": "source", "representation_role": "original", "path": "edi/cicada97-1997-98/Solo.edi" }, { "id": "mth5", "kind": "transfer_function", "format": "mth5", "provenance_role": "derived", "representation_role": "alternate", "path": "h5/cicada97-1997-98/Solo.h5" }, { "id": "edi-zip", "kind": "archive", "format": "zip", "path": "bundles/cicada97-1997-98-edi.zip" }, { "id": "survey-mth5", "kind": "archive", "format": "mth5", "path": "bundles/cicada97-1997-98-tf.h5" } ] }