Nisqually (Washington) CPTu soundings — curated Parquet for the conic library

Curated CPTu subset of the 2001 Nisqually, Washington, earthquake liquefaction case-history dataset (Rasanen, Geyin & Maurer, 2023), prepared as an analysis-ready Parquet file for the conic library. The source compiles 24 CPT-based liquefaction case histories from the Mw6.8 Nisqually earthquake in the Puget Sound region. The 15 soundings carrying piezocone pore-pressure measurements (u2) were selected — the 9 plain CPTs without u2 were excluded — using the source's Cone Tip Resistance Type field. The source tip for these 15 is the pore-pressure-corrected qt; the uncorrected qc was recovered as qc = qt − u2·(1 − a) and stored under the qc name for cross-dataset consistency (downstream the corrected qt is recomputed). Source units (q, fs, u2 all in MPa) were converted to the canonical schema (qc in MPa, fs and u2 in kPa). Cleaning steps: the synthetic pre-drill infill interval (depth below the pre-drill depth) was dropped; rows at exactly equal depth within a sounding collapsed by averaging the measurements; negative sleeve friction truncated to zero; rows with no usable cone reading dropped. Because the source supplies neither density nor in-situ stresses, vertical stresses were derived: the soil unit weight was estimated per depth from the CPT via Robertson & Cabal (2010) and integrated down each profile to give the total stress . The pore pressure u0 is hydrostatic from the per-site groundwater table depth. A constant cone area ratio (0.80) was assumed, as the source records none. A dense per-file key (ID) was assigned in source case-history order while preserving the site name (Site name). The file is sorted by sounding and depth and stored as zstd-compressed Parquet. This curated derivative is released under CC BY 4.0. Changes are limited to subsetting, cleaning, qc recovery, stress derivation, re-keying, unit conversion, and format conversion. Caveats: the unit weight is correlated from the CPT, not measured, so σv_eff and any quantity normalized by it inherit that assumption; the cone area ratio is assumed; 17 of the original 24 sites map to artificial fill, which may affect the applicability of correlations developed for natural soils. Original dataset: Rasanen, R.A., Geyin, M., Maurer, B.W. (2023). Select Liquefaction Case Histories from the 2001 Nisqually, Washington, Earthquake: A Digital Data Set and Assessment of Model Performance. Earthquake Spectra, 39(3), 1534–1557. DesignSafe-CI, PRJ-3758.