In 1995 a temporary array of 25 stations using 1 Hz 3-component sensors was set across the Parkway neighborhood, in the Wainuiomata intermontaine lacustrine basin across the Harbor from Wellington City.

• A 100-m-deep basin with very low velocities, Parkway has been studied by Taber and Yu, IGNS, and engineering firms.
• It is an urban and noisy environment.
• Of many earthquake records available, we examine one from 15 km north, and one 80 km south.

The 1 Hz array was somewhat elongated N-S, with basin stations in the middle and bedrock stations on the margin.

• N-S elongation was reason to pick events to N and S.
• Analysis used only basin stations.
• 110 s event records, 55 s noise records.
• 3 Aug. 1995 and noise record included 12 and 14 stations.
• 29 Aug. 1995 record included only 7 stations.

Velocity-spectral analyses run separately on each component of the basin array record of the 3 Aug. 1995 event, 15 km north.

• The vertical component record produces a discontinuous but pickable dispersion trend.
• The north component shows the same trend, plus a lower-velocity pick at 2.5 Hz (0.4 s period)
• The east component is a mix of both.

• The Kyoto Univ. group only uses verticals, to eliminate Love-wave interference.

Analyses of two events and two noise records on the Parkway array.

• Upper left: Aug. 3 event as above, with picked dispersion.
• Lower left: Aug. 29 event, with only 7 stations. Below white-dotted spatial-aliasing boundary, the horizontal banding demonstrates velocity wraparound.
• Upper right: noise record 1, with good dispersion to 7 Hz, but high apparent velocities at 2-4 Hz.
• Lower right: noise record 2, with lack of coherence. Likely contained E-W-propagating energy, crossing the N-S analysis direction of the noise records.

3 refraction records from Eric Duggan's Parkway surveys along Black Stream Reserve, and one from Mike Stevens's March 1999 survey in Taieri basin near Dunedin airport (lower right).

• Records were hapahazardly chosen and taken directly from the RT-24 seismograph.
• I did not get any of the existing Parkway records longer than a half second.
• The very short-offset Parkway records at left are with vertical geophones above, and horizontals below.

• An unusually high Poisson's ratio at Taieri basin is clear from the record.

Velocity-spectral analyses of Parkway and Taieri refraction records.

• The short half-second Parkway records severely limit frequency resolution.
• The short-offset records are not of ``normal modes;'' propagation is only in the very shallowest layer.

• The Taieri analysis is clear.

Log-log plot of dispersion picks from Parkway 3 Aug. 1995 event and noise records, and refraction explosion record from Taieri basin near Dunedin Airport; against synthetic dispersions for each locale.

• Low-velocity Parkway picks from horizontal and short refraction records were ignored.

Shallow S-wave velocity models for Parkway and Taieri basins.

• Shallow velocities at <50 m tightly constrained.
• Velocity gradients would fit both data sets better than steps.

Very shallow velocity model comparisons from an IGNS report, including a consultant's SASW dropped-blocks, and IGNS cone penetrometer tests.

• Excellent agreement with 1 Hz array and some refraction record results with the 210 m/s velocity to 15 m depth.

Velocity-spectral image shows aliasing and velocity wraparound, but peaks at low values may explain why Seatoun showed largest acceleration in Wellington.