NYCHA Climate Risk Index

Technical Methodology — encoding, normalization, parameters, data sources

1. Risk Formula

Risk is calculated independently for each of 4 climate domains, then combined as a weighted average:

Per domain: adjusted = min_max(hazard × vulnerability) × (1 − effectiveness × mitigation)
Risk Index: = min_max( 0.50·Coastal + 0.40·Stormwater + 0.10·Heat )

All values normalized to [0, 1]. Higher = more risk. A development with no hazard exposure in a domain scores 0 for that domain, pulling the overall score down (weighted by that domain’s share). Per NYCHA direction, the Overall Risk score is weighted Coastal 0.50, Stormwater 0.40, Heat 0.10. Groundwater is weighted 0 — it does not contribute to the Overall Risk score; its per-domain risk is still computed and shown for reference.

2. Encoding Methods

MethodFormulaOutputUse Case
min_max(v − min) / (max − min)[0, 1]Higher raw value = higher score (solar radiation, building count below DFE)
min_max_inverse1 − (v − min) / (max − min)[0, 1]Lower raw value = higher score (groundwater depth: shallower = more hazard)
booleanTrue → 1.0, False → 0.0{0, 1}Binary presence/absence (stormwater exposure, cool roof)
categoricalLookup tableCustomOrdinal categories (Tidal onset: 2050s=1.0, 2080s=0.67, 2100=0.33, no onset=0; Equipment: Yes=1.0, Partial=0.5, none=0)

3. Normalization

StageMethodNotes
AttributesEncoded per method aboveAll output in [0, 1]
Sub-IndicesWeighted average of attributes, then min-max normalizedMissing attributes filled with 0.5 (neutral)
IndicesPer-site weighted average of sub-indices, then min-max normalizedConditional sub-indices (vulnerability/mitigation) that are N/A for a site are excluded from both the numerator and weight denominator for that site. Non-conditional missing sub-indices filled with 0.5 (neutral).
Domain RiskH × V product, min-max normalized, then mitigation appliedPer-domain before final averaging
Risk IndexWeighted sum of domain risks (Coastal 0.50, Stormwater 0.40, Heat 0.10; Groundwater 0.00 — excluded), then min-max normalizedFinal output

Edge Cases

ConditionBehavior
All values identicalAll set to 0.5
Single valid valueSet to 0.5
No hazard exposure in a domainDomain risk = 0 (pulls overall risk down)
Hazard not present → vulnerability / mitigationNaN (displayed as “Not applicable”); excluded from composite index aggregation for that site
Groundwater depth ≥ 16 ftCapped before normalization → hazard scores 0; groundwater V and M set to N/A
Flood-exposed site with no buildings below DFEVulnerability floor of 0.1 applied (coastal, stormwater, groundwater)

4. Parameters

ParameterValuePurpose
Coastal DFE Buffer (2050)40 inches (3.33 ft)Buffer added to FEMA BFE; combined freeboard and sea level rise
Coastal DFE Buffer (2080)52 inches (4.33 ft)Buffer added to FEMA BFE; combined freeboard and sea level rise
Stormwater DFE Buffer24 inches (2.0 ft)Buffer added to stormwater BFE
Groundwater Depth Cap16 ftUSGS HERA threshold; depths ≥ 16 ft = no hazard
Mitigation Effectiveness (default)0.6Max risk reduction for heat domain
Mitigation Effectiveness (flood)0.9Enhanced for coastal, stormwater, and groundwater (engineered flood-protection measures)
Basement Depth Below Grade10 ftUniversal assumption for basement floor elevation (z_grade − 10 ft)
Vulnerability Floor0.1Minimum vulnerability for hazard-exposed properties with zero building vulnerability (applies to coastal, stormwater, and groundwater domains)
High-Rise Threshold60 ft (≈ 6 stories)Campus type classification
API Cache TTL30 daysAuto-refresh for FEMA, coastal surge, building data

5. Design Flood Elevation (DFE)

The coastal and stormwater DFE follow different BFE derivations:

Coastal DFE = FEMA PFIRM STATIC_BFE (per development) + freeboard buffer
Stormwater DFE = max(LiDAR DEM elevation within stormwater flood zone intersection) + 24″ freeboard

Coastal BFE — FEMA PFIRM source

Per NYCHA methodology, the coastal Base Flood Elevation is sourced from FEMA’s Preliminary Flood Insurance Rate Map (PFIRM) STATIC_BFE field, available on FEMA AE zones (1%-annual-chance flood zones with a regulatory flood elevation). For each development:

  1. Overlapping AE zones: Take the maximum STATIC_BFE from any FEMA AE zone whose geometry intersects the development polygon.
  2. Nearest-neighbor fallback: If no AE zone overlaps the development, use the STATIC_BFE of the spatially nearest AE zone (per NYCHA’s “closest to development” wording). This is needed for sites outside the current FEMA floodplain but exposed under projected 2050/2080 coastal surge.
  3. No FEMA reference available: If no AE zone exists anywhere, the coastal DFE is N/A — no regulatory flood elevation exists to serve as the BFE.

A single BFE value per development is then used for both horizons; the 2050 and 2080 DFEs differ only by the freeboard buffer (40″ vs 52″), giving an exact 12″ differential per the NYC Climate Resiliency Design Guidelines v4.1, Table 5.

Stormwater BFE — DEM-derived

There is no regulatory BFE for stormwater flooding, so the stormwater BFE is derived topographically: the highest ground elevation (NAVD88) on the campus exposed to stormwater flooding — computed as the maximum LiDAR DEM value within the stormwater flood zone’s intersection with the development polygon. A 24″ freeboard buffer is then applied to produce Stormwater_DFE_2050 and Stormwater_DFE_2080.

Per-scenario gating

Each scenario’s DFE is only assigned when the development is exposed in that horizon:

Sites exposed only in 2080 therefore have Coastal_DFE_2050 = None and only the 2080 DFE is displayed. The sidebar and PDF hide DFE rows for scenarios where no value is populated.

ScenarioBFE SourceFreeboardOutput Field
Coastal current (building counts)FEMA STATIC_BFE (with DEM fallback inside FEMA zones)40 inBuildings_with_Inhabitation / Basements_Below_Coastal_DFE
Coastal 2050FEMA STATIC_BFE (overlap or nearest AE zone)40 inCoastal_DFE_2050
Coastal 2080FEMA STATIC_BFE (overlap or nearest AE zone)52 inCoastal_DFE_2080
Stormwater 2050Max LiDAR DEM within Stormwater_2050_Moderate.tif intersection24 inStormwater_DFE_2050

Vulnerable-building counts

  1. Using the Building Elevation Subgrade dataset (z_floor = first inhabited floor elevation, NAVD88), count buildings on the development where z_floor < DFE.
  2. Estimate basement elevation as z_grade − 10 ft (universal assumption; configurable via BASEMENT_DEPTH_BELOW_GRADE_FT). Count buildings where basement_elevation < DFE. This captures vulnerability to building systems and infrastructure even when inhabited floors are above the flood line.

If a development has no flood zone overlap, DFE is not calculated and the building count defaults to 0.

6. Conditional Relationships

Vulnerability and mitigation sub-indices are only calculated when the corresponding hazard is present; otherwise set to NaN (“Not applicable”). When the composite hazard/vulnerability/mitigation indices are aggregated, N/A conditional sub-indices are excluded from both the numerator and the weight denominator for that site — so sites are not penalized by forced zeros for domains that do not apply to them.

ConditionTrigger (ANY)Applies To
COASTAL_TIDAL_HAZARDPFIRM=True, Surge 2050=True, Surge 2080=True, or Tidal ≠ emptyCoastal mitigation
STORMWATER_HAZARDStormwater 2020, 2050, or 2080 = TrueStormwater mitigation
GROUNDWATER_HAZARDDepth_to_Groundwater_ft < 16 ft (the no-hazard cap)Groundwater vulnerability and mitigation
HEAT_HAZARDSolarRadGroundPct > 0, bldg_height_mean > 0, Total_Roof_Area > 0, or Outdoor_Temperature_Rank presentHeat vulnerability and mitigation

7. Data Sources

Auto-Refreshing (API, 30-day cache)

DataAPIFields Used
FEMA PFIRM Flood Zoneshazards.fema.gov NFHL MapServer/28FLD_ZONE, ZONE_SUBTY, SFHA_TF, STATIC_BFE
Coastal Surge 2050/2080 (NPCC projections)NYC MOCEJ ArcGIS FeatureServer, serving New York City Panel on Climate Change (NPCC) sea-level-rise and coastal flood projections (reference link shows the 2050s 100-year floodplain; 2080s and tidal-onset layers are separate NPCC datasets published by NYC Open Data)Geometry (boolean intersection)
Building Elevation SubgradeNYC Open Data SODA (bsin-59hv)the_geom, z_floor
NYC Building FootprintsNYC Open Data SODA (5zhs-2jue)the_geom, height_roof; Shape_Area computed

Manual Refresh (local rasters)

FileSourceUpdate Trigger
Stormwater_2020_Light.tifNYC Department of Environmental Protection (DEP) — NYC Stormwater Flood MapsNew DEP release or NPCC projection update
Stormwater_2050_Moderate.tifNYC Department of Environmental Protection (DEP) — NYC Stormwater Flood MapsNew DEP release or NPCC projection update
Stormwater_2080_Extreme.tifNYC Department of Environmental Protection (DEP) — NYC Stormwater Flood MapsNew DEP release or NPCC projection update
NYC_DEM_LiDAR.tifUSGS 3DEP / NYC DoITTNew LiDAR acquisition
nyc_groundwater_depth_interpolated.tifUSGS / NYC DEPNew interpolation
NYC_SolarRadiation_Merged.tifNYC / CUNY Solar MapNew analysis

NYCHA-Maintained

FileContents
NYCHA Development Data Book.csvAdministrative data, Sandy/IDA flags, completion date, building type
2025 Resident Data Book.csvDemographics (Minors %, Seniors %)
Flood Protection Components.xlsxFlood hardening inventory, critical equipment status
335 Developments Hazard Exposure desktop.xlsxTidal flooding timeline, Cloudburst designations
Developments with Replaced Roofs.xlsxCool roof installations
DFTA NYCHA Cooling Centers.xlsxDFTA senior cooling center locations (union-flagged with DYCD into OnsiteCoolingCenter)
DYCD NYCHA Cooling Centers.xlsx (optional)DYCD community/youth cooling center locations; OR-combined with DFTA into the OnsiteCoolingCenter attribute. Loads gracefully if absent.
nycha_climate_adaptation_plan_outdoor_temp.csvPer-development Outdoor Temperature Rank (Low/Medium/High) extracted from the NYCHA Climate Adaptation Plan (October 2021), Appendix: Climate Variables. Source methodology: 2018 summer Landsat surface temperature averaged within each development polygon, tercile-ranked.

8. Spatial Operations

SourceMethodAggregation
Building Elevation PointsSpatial join (points within polygon)min/max/mean z_floor; count below DFE
NYC Building FootprintsCentroid-in-polygon joinmean height, sum area, count
Rasters (groundwater, stormwater, solar)Polygon mask extractionMean (groundwater depth), boolean any>0 (stormwater exposure), sum of ground-area pixels > 0 (solar radiation, continuous input to min_max)
FEMA NFHL / Coastal Surge vectorsDissolved union, prepared geometry intersectionBoolean intersects + DFE via DEM sampling
Tabular (Excel, CSV)Exact name match on DEVELOPMENTFirst match or aggregated per domain rules

All spatial overlays use EPSG:2263 (NY State Plane). Vectors reprojected to match raster CRS before extraction. Invalid geometries repaired via make_valid before dissolve.

9. Ranking & Classification

All indices use descending rank (Rank 1 = highest/worst). For flood-related sub-indices, only exposed properties are ranked; zero-exposure properties show “N/A”.

third = ceil(total / 3)
Rank ≤ third → HIGH  |  Rank ≤ 2 × third → MEDIUM  |  Rank > 2 × third → LOW

10. Mitigation Action Matching

Mitigation action recommendations are tailored to each development based on its hazard exposure level. The system determines which recommendations to show using the tertile ranking of each hazard sub-index.

Decision Logic

Hazard DomainHIGH (top third)MEDIUM (middle third)LOW (bottom third)
Coastal/TidalFull mitigation packageFull mitigation packageNo section shown
StormwaterComprehensive package (H&H analysis + critical systems + site strategies + entrances + buffering)Reduced package (H&H analysis + buffering only)No section shown
Extreme HeatShown for all exposed developments (any level except N/A)
GroundwaterInvestigation stepsInvestigation stepsNo section shown

Coastal Flood Mitigation Actions

BASELINE — Protect Critical Systems (boilers, electrical switchgear, generators, elevators, fire alarm, BMS, telecom):

Passive strategies preferred. Deployable strategies may be used where passive strategies are not feasible.

Protect entrances and exits: protect or relocate ground-floor spaces below DFE; install passive barriers at entrances.

BETTERMENT — Buffering strategies to reduce flood impacts: elevated curbs or floodwalls, regraded site to direct water away from buildings, vegetated berms along waterfront edges.

Stormwater Flood Mitigation Actions — HIGH Risk

BASELINE — All coastal baseline actions plus:

BETTERMENT — Buffering per NYCHA Stormwater Design Manual: elevated curbs, adapted topography, vegetated berms, subsurface tanks and nature-based solutions.

Stormwater Flood Mitigation Actions — MEDIUM Risk

BASELINE — Build H&H Model to refine DFE per Stormwater Design Manual.

BETTERMENT — Same buffering strategies as HIGH risk.

Extreme Heat Mitigation Actions

BASELINE:

BETTERMENT:

High Groundwater Mitigation Actions

Mitigation credit is given to properties with the following measures (weights for the Groundwater Mitigation sub-index):

Recommended further investigation steps for high-groundwater sites:

  1. Review existing geotechnical surveys and soil borings
  2. Consult geotechnical engineer on groundwater impacts
  3. Install groundwater monitoring wells
  4. Evaluate active subsurface drainage strategies

11. Key Assumptions

  1. Relative ranking: Scores rank NYCHA properties against each other, not against external safety standards.
  2. Weighted domain aggregation: The Overall Risk Index weights Coastal/Tidal 0.50, Stormwater 0.40, Heat 0.10, and Groundwater 0.00. Per NYCHA direction, groundwater does not contribute to the Overall Risk score; its per-domain risk is still computed and shown for reference. This weighting reflects NYCHA’s prioritization of coastal and stormwater flood risk for capital planning.
  3. Hazard-present gating: Mitigation only assessed where hazard exists; otherwise “Not applicable.”
  4. Residual risk: Full mitigation reduces risk by 90% (flood) or 60% (heat), never to zero.
  5. Near-term weighting: Current conditions (2020) weighted more heavily than distant projections (2080).
  6. Freeboard buffers: 40 inches coastal (current, 2050 scenarios), 52 inches coastal (2080 scenario), and 24 inches stormwater — conservative margins above the computed base flood elevation (FEMA BFE for coastal, derived stormwater BFE for stormwater).
  7. Groundwater vulnerability proxy: Estimated from building age (weight 1.0) since direct below-grade condition data is unavailable. Age is min-max normalized across the portfolio (oldest completion date → 1.0, newest → 0.0) — the same continuous-range approach used for groundwater depth, just in reverse. Applied only to sites where groundwater is a concern (depth < 16 ft).
  8. Basement depth assumption: Every building is assumed to have a basement 10 ft below grade (z_grade). This universal assumption may be replaced with site-specific data when available from NYCHA engineering teams or verified subgrade observations.
  9. Vulnerability floor: Flood-exposed properties with zero building vulnerability (no inhabited floors or basements below DFE) receive a minimum vulnerability score of 0.1, reflecting site-level risks such as access disruption, grounds flooding, and infrastructure damage.