The Place-Evaluation Framework

The attributes that most affect health are the ones conventional search ignores.

A place is a trajectory, not a snapshot. The attributes that most affect health — acoustic quiet, dark skies, clean air, nature access, and social connection — are precisely the ones conventional real-estate search (price, beds, baths, school ratings) leaves out. Still is built as a weighted, multi-scale (state → building) multi-criteria engine that layers authoritative public data over commercial property feeds, and scores a place not as a single number but as a value, a direction of travel, and a watchlist of what is entitled nearby.

The evidence base is strong enough to act on. Combining seven longitudinal studies, the WHO review found road-traffic noise raises ischemic-heart-disease incidence with a relative risk of 1.08 (95% CI 1.01–1.15) per 10 dB L_den. White et al. (2019, n = 19,806) found that 120–179 minutes per week in nature was associated with markedly better self-reported health (OR 1.59, 95% CI 1.31–1.92) and wellbeing (OR 1.23, 95% CI 1.08–1.40). Green, quiet, and dark attributes are independently linked to mental, physical, and contemplative outcomes — and no mainstream real-estate tool measures any of them.

This paper sets out the rationale, the geographic-scale taxonomy, the health and environmental dimensions, the temporal and governance layers, the seek-side amenities, the decision method (MCDA / AHP / weighted overlay), the data sources, and the caveats. It is the why and the what behind the product, published openly because the research should be accessible, not proprietary.

Place is a health variable — and a measurable one.

"Where you live shapes everything — your sleep, your breath, your sense of belonging."

The restorative case is well-evidenced and under-served.

Attention Restoration Theory (Kaplan & Kaplan, 1989; Kaplan, 1995) holds that natural environments restore depleted directed attention through four properties — being away, extent, fascination (especially "soft fascination"), and compatibility. Stress Reduction Theory (Ulrich et al., 1991) shows physiological stress recovery — heart rate, blood pressure, skin conductance — from viewing nature within roughly four minutes. These are complementary pathways: ART is cognitive and attentional, SRT is affective and physiological. Both predict that low-stimulation, natural, quiet settings confer measurable health benefits — yet no mainstream real-estate tool measures acoustic, light, or restorative quality.

Environmental stressors cause measurable disease.

The WHO Environmental Noise Guidelines for the European Region (2018) recommend keeping road traffic below 53 dB L_den and night noise below 45 dB L_night; the supporting systematic review (van Kempen et al., 2018) found a relative risk of 1.08 (95% CI 1.01–1.15) per 10 dB L_den for ischemic heart disease — rated high-quality evidence, with risk rising continuously from about 50 dB. A more recent umbrella review (Münzel et al., 2024) gives a more conservative combined cardiovascular risk of 3.2% (95% CI 1.1–5.2%) per 10 dB. The FAA uses DNL 65 dB as its residential land-use compatibility threshold for aircraft noise, with a 10 dB nighttime penalty, though most standards bodies treat 55 dB DNL as the threshold of impact in residential areas. Fine particulate matter (PM2.5) is linked to higher depressive-symptom scores (≈0.07 SD per 10 µg/m³ in one cohort) as well as cardiopulmonary mortality.

The sensitive-person rationale.

This is the project's ethical and practical core. Aron & Aron (1997), using their 27-item Highly Sensitive Person Scale, identified Sensory Processing Sensitivity in the top 15–20% of the population — a heritable trait characterized by deeper processing, greater emotional reactivity, and heightened sensitivity to subtle stimuli; fMRI work (Acevedo, Aron et al., 2014) shows stronger activation in attention and empathy regions. For highly sensitive people, environmental quality is not a preference but a determinant of functioning: noise, crowding, harsh light, and pollution impose genuine physiological and mental-health costs. Because conventional search tools are blind to these attributes, sensitive users are systematically mismatched to places — the precise gap Still fills.

Existing location-scoring products optimize the wrong things.

Walk Score, AreaVibes, and Niche are useful but converge on amenity density, cost, crime, and schools. None measures acoustic refuge, dark skies, soundscape naturalness, circadian light exposure, hyperlocal air quality, or contemplative quality — exactly the dimensions sensitive people and restoration-seekers need. The data to measure them exists, mostly free and authoritative, but is fragmented across agencies and geographic scales. The core challenge is integration and normalization, not data scarcity.

Solitude ↔ Belonging is one tunable spectrum, not two products.

Still treats Solitude/Restoration and Community/Belonging not as separate tools but as opposite poles of a tunable weight vector over one shared criteria taxonomy. A solitude profile up-weights acoustic quiet, dark skies, low density, nature proximity, and clean air, and may down-weight density-dependent amenities. A belonging profile up-weights social associations, civic engagement, walkable third places, contemplative-community presence, and residential stability. A balanced profile blends them.

Crucially, some criteria — clean air, green space, low pollution, healthcare access — are health-positive under any profile, and carry a floor weight that a user cannot zero out. The community case is as evidence-based as the solitude case but distinct: social capital, measured by Chetty et al.'s Social Capital Atlas (Nature, 2022, built from 21 billion Facebook friendships among 72.2 million U.S. users) predicts upward mobility — a one-SD increase in county economic connectedness is associated with an 8.2-percentile-rank increase in the adult income of children from low-income families — and correlates with health. A user weighting toward belonging wants high social infrastructure; a user weighting toward solitude may actively want low density. The engine must let these pull in opposite directions.

Eight geographic scales, from state to building.

Data resolution and the meaningful "unit of decision" change with scale. Health (mental / physical / contemplative), environment (human + animal), and time (current / historic / future) thread through each.

Mental, physical, and contemplative health.

Mental health

Chronic transportation noise drives sleep disruption, annoyance, stress hormones, and links to depression and anxiety; artificial light at night disrupts circadian rhythm; crowding and density add load. Against these, green-space access protects — meta-analyses link higher NDVI within 300 m of home to lower depressive symptoms — alongside soft fascination, social connection (versus isolation), and access to mental-health care.

Physical health

Air quality — PM2.5, ultrafine particulates, ozone, NO₂ — measured by the EPA regulatory network and hyperlocal sensors; water quality; walkability and active living; food access (USDA low-income / low-access tracts); proximity to pollution sources (highways, industry, agriculture); heat and climate exposure; healthcare access; and community morbidity patterns from CDC PLACES.

Contemplative health

This is where Still differentiates. Access to nature and awe — in a field experiment (Piff et al., 2015, 5 studies, N = 2,078), participants who gazed up at a grove of towering eucalyptus for one minute subsequently picked up more dropped pens for a researcher than those who stared at a tall building — plus dark skies, quiet and contemplative space, contemplative-community presence, and a sense of place and rootedness (residential tenure, length of community history). Forest-bathing research provides the bridge to physical health: it raises natural-killer-cell activity, lowers cortisol, and shifts toward parasympathetic dominance.

Sound, light, air, water, life, and hazard.

Acoustic environment

Still's signature dimension, and the hardest to source. Road traffic is modeled from state-DOT Annual Average Daily Traffic counts using standard noise-propagation logic, or taken from the US DOT National Transportation Noise Map (aviation + road + rail, modeled dB(A) L_den at ~30 m). Aircraft: FAA DNL contours; DNL 65 dB is the residential-incompatibility threshold. Natural quiet: the NPS Natural Sounds program measures ambient levels and the "noise-free interval"; soundscape ecology decomposes the environment into biophony, geophony, and anthrophony — high biophony and low anthrophony is both an ecological-health and a human-restoration indicator. A single jet at 38,000 ft can degrade roughly 1,000 square miles of natural quiet.

Light environment

Measured by the Bortle scale (Class 1 pristine → Class 9 inner city) and SQM sky brightness (mag/arcsec²), derived from VIIRS satellite night-lights (~500 m annual composites). Artificial light at night suppresses melatonin and disrupts sleep; it also disorients migrating birds, kills nocturnal insects, and deters sea-turtle nesting. Proximity to a certified International Dark Sky Place is a strong dark-sky proxy.

Air quality

EPA AQI and the regulatory monitoring network (attainment / nonattainment); AirNow for real-time AQI; the PurpleAir low-cost PM2.5 network for hyperlocal coverage (dense but requiring calibration — it over-reads PM2.5 by default and needs the EPA correction); traffic-related pollution near highways; and rising wildfire-smoke exposure, especially in the West.

Water

Drinking water from EPA SDWIS (public-water-system violations) and UCMR 5 PFAS occurrence — the 2024 PFAS rule set enforceable limits of 4.0 ppt for PFOA and PFOS; surface and groundwater from USGS NWIS and the Water Quality Portal; contamination proximity from EPA Superfund and Brownfields. Private wells, common in rural areas, are not in the federal database — Still flags that and recommends local testing.

Ecology & biodiversity

Habitat connectivity and wildlife corridors; biodiversity as an environmental-health indicator; tree canopy (NLCD, NDVI); and proximity to protected lands (USGS PAD-US). The nature-dose literature (White et al., 2019) found benefits peaking around 200–300 minutes per week, with no further gain beyond.

Climate & natural hazards

FEMA National Risk Index (18 hazards — wildfire, flood, drought, heat, coastal flood, county + tract); the National Flood Hazard Layer; NOAA climate data and projections; and forward-looking multi-horizon flood / fire / heat models. Projections are modeled, not observed, and are labeled as such.

A place is a trajectory, not a snapshot.

Every incumbent sells a snapshot. Still's defining claim is that two streets equally quiet today are not equal if one has grown louder every year and has an entitled subdivision behind the tree line. The temporal dimension maps onto three lenses.

Past — read the public record as a curve

Traffic-count history (a road on a rising curve predicts the next decade better than today's number), Census change over decades (densifying / aging / gentrifying / hollowing), building-permit history (the rate of nearby construction), and air-monitor and night-lights time series. Direction of travel is itself a criterion: a score is a value plus a trend arrow.

Present — the snapshot

What incumbents already sell. Necessary, but the least differentiated layer.

Future — two distinct kinds that must not be conflated

Planned change (governance data) — comprehensive plans, zoning and pending rezonings, capital-improvement plans, and the approved-but-unbuilt development pipeline. This is the gold nobody mines: the entitled townhomes behind the quiet field, the funded road-widening, the commercial rezoning across the street. It is public, local, and messy — which is exactly why no consumer product touches it, and why it is defensible once built. Projected change (modeled futures) — 30-year climate projections, plus regional-transportation and demographic projections where published.

What will owning here cost me — and what does the law let my neighbors do?

No consumer tool surfaces this, and that framing is the differentiator. The data is mostly public but text-heavy — state statute, county assessor methodology, municipal code — rather than clean API data, and the assembly cost is real.

State

Income-tax structure; property-tax methodology; homestead exemptions; capital-gains treatment on home sale; estate and inheritance tax; landlord–tenant balance; HOA-law strength; short-term-rental legality.

City / town

Local tax add-ons; zoning-code flexibility; and — important for sensitive buyers — local noise and lighting ordinances and whether they are enforced. A town with an enforced noise ordinance is structurally different from one without: the existence of an ordinance is a weak signal, enforcement is the strong one. Still captures both — ordinance present? enforced? complaint process?

Solitude is not just the absence of bad — it is the presence of good.

The seek bundle carries equal weight to the avoid bundle.

These carry trend arrows too: a 15-year-old farmers market signals a different community than a one-season-old one; a just-passed dark-sky ordinance is a trajectory, not a static fact.

Combining weighted criteria — MCDA, transparently.

Multi-Criteria Decision Analysis is the governing framework. The core pipeline:

1. Criteria taxonomy. Organize criteria hierarchically — goal → categories (acoustic, light, air, water, nature, social, hazard, health, cost) → sub-criteria → measurable indicators with units.
2. Weight elicitation. The Analytic Hierarchy Process (Saaty) uses pairwise comparisons on a 1–9 scale, builds a comparison matrix, derives weights from its principal eigenvector, and checks a consistency ratio below 0.10 — letting a user express, e.g., "acoustic quiet is strongly more important than walkability."
3. Normalization. Convert heterogeneous indicators (dB, µg/m³, mag/arcsec², percentiles) to a common 0–1 suitability scale via value functions. Threshold functions encode evidence-based breakpoints — a steep penalty above 53 dB L_den, a bonus below 45.
4. Aggregation. Weighted Linear Combination: suitability = Σ(weightᵢ × normalized scoreᵢ). Ordered Weighted Averaging allows risk-averse aggregation — appropriate for sensitive users, where one dealbreaker (a flight path) should be able to veto.
5. Spatial overlay. The McHarg "layer-cake" overlay tradition, implemented as raster algebra or vector zonal statistics.
6. Accessibility. Score proximity by travel time, not Euclidean distance, using isochrones for drive / walk / bike access to work, nature, healthcare, and third places.
7. Multi-scale conflict resolution. When a tract scores well but the specific parcel abuts a highway, the finer scale overrides for exposure criteria (worst-case local value) while coarser scales govern context criteria (social capital, climate). This precedence is documented explicitly.

The core data sources.

Most are free and authoritative; the engineering work is integration and normalization across scales and update cadences. Every value is labeled measured or modeled, with its source and retrieval date.

Caveats and ethics.

References.