{
  "source_id": "park-choi-2010",
  "locator": "Results, section Ground effect, and Fig. 12; model L3 (fins spread) and model N (body only); forces measured at 0 deg attack",
  "retrieved_at": "2026-10-31T18:18:51Z",
  "definition": "r = h / b, flight height over half wing span (b = S/2, S the full pectoral span)",
  "measured": [
    {"model": "L3", "surface": "solid", "r": 0.21, "drag_change": -0.09, "lift_change": "almost unaffected", "note": "drag nearly constant for r > 0.35"},
    {"model": "L3", "surface": "water", "r": 0.33, "drag_change": -0.14, "lift_change": "almost unaffected", "note": "drag starts to fall below r = 0.97; maximum reduction at r = 0.33"},
    {"model": "N (body only)", "surface": "solid", "r": 0.21, "drag_change": -0.04, "lift_change": null},
    {"model": "N (body only)", "surface": "water", "r": 0.33, "drag_change": -0.07, "lift_change": null}
  ],
  "lift_to_drag": "increased with decreasing r from r = 0.35 (solid) and r = 0.97 (water); values read from the text, not the figure",
  "qualifiers": [
    "The wind tunnel's water surface could not move with the air, so the authors expect a bigger drag reduction in the real situation (slip velocity of the sea surface).",
    "No measurement below r = 0.21 (solid) or 0.33 (water); the simulator holds the r = 0.33 reduction constant below it and says so.",
    "Lift-to-drag ratio values at each r are in Fig. 12B, which was not read; only the drag percentages stated in the text are used."
  ],
  "simulator_model": "drag multiplier f(r) = 1 for r >= 0.97; linear from 1 at r = 0.97 to 0.86 at r = 0.33 (water surface, model L3); 0.86 held for r < 0.33. Lift unchanged."
}