{
  "version": "8.0",
  "declared": "2026-10-19",
  "target": "Relative biophysical potential for repeated production by cultivated edible olives under a stated cultivar and water scenario; not a probability of annual yield or economic profitability.",
  "freezeCalibration": "QC-passed station annual minima, >=95% annual and winter completeness. Compare matched years with original ERA5-EX. Stations receive equal aggregate weight. Fit residual models and test by withholding complete geographic regions. Retain the simplest correction with improved out-of-region error; report every fold and the uncorrected baseline.",
  "freezeSurface": "Correct the distribution of annual minima using transferable climatic features, never distance-to-literature or a score boost at orchard coordinates. Preserve source quantiles and identify corrected estimates separately. Expose uncertainty and coverage independently of suitability.",
  "flowering": "Arbequina warm response: optimum7.3C, breakpoint21.7C, rate-0.50,339CU; Leccino/Frantoio original20.7C/-0.56 and612/671CU. Both normal weather and daily reconstructions use matched cultivar responses. Seven winter-month temperature residuals are assimilated using fixed250km Gaussian,max500km,source winter-min3C and winter-mean4.5C climate similarity; monthly support shrinks toward zero and short records receive min(1,accepted month-years/15) weight. Separate native/overview logistic fits recalibrated after weather correction. Entire geographic regions are removed from weather nodes and chilling-fit training before evaluation. Huasco2014/15/24/25 and Lalamilo2-4month-years are explicitly short development records; no30yearnormal is fabricated.",
  "orchards": "Freeze primary orchard outcome labels before testing scores. Do not treat alternate bearing as a negative climate label. Consecutive yield series, multi-year means, short trials, flowering failures and freeze damage are separate outcomes. No point/radius score overrides. Literature already seen by developers is an external compatibility audit, not newly independent predictive validation.",
  "evaluation": "Report station geographic holdout accuracy separately from orchard compatibility and numerical/software checks. Compare irrigated evidence with irrigated scores; rainfed with rainfed. Declare contradictions, exclusions, distinct study clusters, and geographic gaps. Do not claim a globally calibrated reliable-yield probability.",
  "wetness": "Add CHELSA BIO17 driest consecutive3month total. Its numeric interpretation exactly matches independently fetched monthly-pr sums at32nativecontrols. Management response [mm,score]:[0,100],[75,100],[150,85],[225,75],[300,65],[450,50],[600,30],[900,5]; combine with prior warm-quarter wetness by minimum. If wetness<90, cap potential at79. These thresholds are explicit agronomic planning choices, not fitted infection or crop-loss probabilities.",
  "newDevelopmentCases": [
    "Huasco: newINIAstationweather and2003-2007irrigatedcultivarharvests",
    "LaYaradaPeru: officialstation/cropmonitoring, Sevillanarequirementunknown",
    "LalamiloHawaii: incompletehistoricalstationrecord andyounglowchillcultivarfruitingtrial",
    "LaPlata/Uruguay/CoronelDorrego: actualproductionwithdisease/springfrost/alternatebearingconstraints"
  ],
  "palette": "Green reserved for80-100; amber40-79; red0-39. Interpolation within each tier, no green below80.",
  "heat": "Globally cap final relative climate potential by both existing heat and summer response components, without allowing a1.35thermal allowance or compensating favorable conditions. Response curves are explicit planning choices; monthly climate does not provide flowering-stage heatwave records or calibrated crop-loss probabilities.",
  "resourceDisplay": "For the irrigation-assumed potential view, annual CHELSA BIO12<200mm and climate score>=40 triggers muted amber/neutral diagonal striping globally. Low scores remain low. The literal rainfall flag is separate from numeric scores and does not infer water availability, salinity, drainage or sustainable supply. Identical rule flags Huasco,Negev and Sahara, with no orchard-location exceptions.",
  "newDesertChecks": "Five-year ElOuedharvests for local/table cultivars, measured Negev Arbequina production, two-year WadiAlNatrunArbequina production, Siwa salinity/yield decline and long-term Revivim salinity. Development compatibility checks only; do not treat unknown cultivar parameters as validated."
}