01 — Parameter mapping

The Minimum Set of Key Data Parameters in Figure 2 is the shared data contract for both verticals. This page enumerates which parameters apply, which need vertical-specific extensions, and how each lands in the DuckLake knowledge graph as a column or as an edge / property on an observation node.

Observation node, in one place

Every report from a community user, every record pulled from an MCP server, and every agency case becomes a row whose canonical type is observation and whose properties are drawn from the eight Figure 2 parameter classes plus the per-vertical extensions below.

observation.<uuid>
  ├── kind = "report" | "mcp_pull" | "agency_case"
  ├── vertical = "vbd" | "heat"
  ├── source = "mobile" | "sms" | "voice" | "vectorsurv" | "nws_heatrisk" | …
  ├── (General properties — same for both verticals)
  ├── (Human properties — vertical-specific symptom set)
  ├── (Exposure properties — vertical-specific risk factors)
  ├── (Auxiliary properties — photo / digital biomarker / lab)
  ├── (Environmental properties — vertical-specific)
  └── edges to: pathogen.*, focus.*, county.*, tribe.*, outbreak.*,
                resource.*, population.*  (existing kg nodes)

Class-by-class mapping

General class — identical contract, both verticals

Figure 2 param VBD use Heat use DuckLake landing
Age Risk stratification (children, elderly) Vulnerability score (65+) kg.property(observation.X, "age", value_num)
Sex Demographic Demographic (81% of Maricopa heat deaths are male) …, "sex", value_text
Email Channel for results Optional (anonymous-friendly) …, "contact_email"
Unique ID Stable identifier for follow-up Same kg.node.node_id (uuid)
Occupation Outdoor work = WNV/tick risk Outdoor work = heat risk (OSHA) …, "occupation"
Date of report Drives Figure 3 milestones Same …, "reported_at", value_text (ISO)
Postal code County / tribe resolution Cooling-center proximity …, "postal_code"
Phone number SMS fallback 211 referral …, "contact_phone"
Household member ID Tick-shared-with-pet detection “Are others in your home at risk?” edge observation → relatesTo → observation
Geographical coordinates Map / cluster detection Cooling-center routing …, "lat", …, "lon" + edge to county.*, tribe.*

Human class — diverges by vertical

Figure 2 param VBD Heat
No symptoms ✓ (asymptomatic check-in) ✓ (“just checking in” for unsheltered)
Symptoms (general) ✓ ✓
Date of illness ✓ — clock starts here for Detect→Notify ✓
Cough / congestion ✓ (WNV neuro) —
Nausea / vomiting ✓ ✓ (heat exhaustion)
Difficulty breathing ✓ (hantavirus, severe WNV) ✓ (heat stroke)
Sore throat ✓ —
Rash ✓ (RMSF, dengue) —
Fever ✓ ✓ (heat stroke ≥ 104 °F)
Chills ✓ —
Diarrhea ✓ —
Bleeding from body openings (severity marker) ✓ —
Red eyes ✓ (dengue, leptospirosis) —
Muscle / body aches ✓ ✓
Discolored / bloody urine (severity marker) ✓ (leptospirosis, severe RMSF) ✓ (rhabdomyolysis)
Loss of smell or taste ✓ —
Yellow skin / yellow eyes (severity marker) ✓ (leptospirosis) —
Absent from work / school ✓ ✓
Sought health care ✓ (drives Notify / Verify) ✓
Heat-specific additions:    
Confusion / altered mental status — ✓ (critical heat-stroke indicator)
Stopped sweating / hot dry skin — ✓ (heat stroke)
Heavy sweating — ✓ (heat exhaustion)
Headache — ✓
Dizziness / fainting — ✓
Muscle cramps — ✓
Core body temp (if measured) — ✓ (numeric)

Exposure class — both extend Figure 2 differently

Figure 2 param VBD Heat
Attending a recent mass gathering ✓ ✓ (also: stadium / outdoor concert in heat)
Tick or insect bite ✓ (when, how many, where on body, attached duration) —
Animal bite ✓ (rabies) —
History of travel ✓ (dengue / Zika importation) ✓ (arrived from cooler climate?)
Contact with live animals ✓ —
Contact with dead or sick animals ✓ (plague, tularemia, HPAI) —
Contact with sick individual / confirmed case ✓ —
Heat-specific additions:    
Time spent outdoors in last 24 h — ✓
Access to working AC at home — ✓ (no AC = elevated risk)
Energy / utility security (any disconnect notices?) — ✓
Currently sheltered? — ✓ (unsheltered = top risk cohort)
Vehicle access for transport — ✓
Occupational heat exposure today — ✓
On medications affecting thermoregulation — ✓ (antipsychotics, diuretics, anticholinergics)

Auxiliary class — overlapping but the photo subject is different

Figure 2 param VBD Heat
Digital biomarker signal Pulse, body temp from wearable Critical: wearable skin-temp, heart-rate variability, sweat rate
Photo Tick / mosquito / sick animal / rash Optional (e.g. unsafe heat environment)
Diagnostic / lab confirmation Lab result reference (LOINC) ED diagnosis code (ICD-10 T67.*)

The digital-biomarker slot is the single biggest unlock for the Heat vertical — wearable data crosses the gap between “checking in” and “you need transport to a cooling center now.”

Environmental class — both heavily extend

Figure 2 param VBD Heat
Date of environmental incident ✓ ✓
Location of vector spotting ✓ —
Unusual presence of vectors ✓ —
Density or number of vectors ✓ —
Flooding ✓ (mosquito habitat) —
Water contamination ✓ (leptospirosis) —
Heat-specific additions:    
Ambient temperature (°F) — ✓ (NWS HeatRisk MCP)
Humidity — ✓
Heat index — ✓ (computed)
NWS HeatRisk level (Green/Yellow/Orange/Red/Magenta) — ✓
Urban-heat-island intensity at user’s block — ✓
Active extreme-heat watch/warning — ✓
VBD-specific additions:    
Standing water within X meters ✓ —
Recent rainfall (mm, last 7 days) ✓ —

Livestock + Wildlife classes — VBD only

The Livestock and Wildlife classes from Figure 2 are first-class for VBD (sick / dead animal reports drive zoonotic surveillance) and silent for Heat. They map to existing edges:

observation.X --reportsAbout--> reservoir.deer_mouse
observation.X --reportsAbout--> species.<from kg.deep/pathogens>
observation.X --reportsAbout--> outbreak.<if matches a known event>

Suppression rules

To avoid surveillance creep, the app does not collect certain Figure 2 parameters in some flows:

Suppression is enforced by the Intake Agent (see 03-agentic-architecture) and audited via a consent_profile property on every observation.

Knowledge-graph deltas required

Adding the application means seeding a small set of new node types into schema/deep/:

These additions are documented under 05-roadmap.md Phase 0.