About

New York State preliminary heat maps were developed using the i‑Tree Cool Air model, a physically based analytical spatial air‑temperature and humidity model developed to map urban heat islands and cooling potential of urban forests. ¹ These preliminary heat maps are available for download as .TIF and .SHP files, in addition to .JPG maps. Downloads are available on each urban area and county page in addition to statewide downloads available on the Home page.

The state maps depict the intensity of heat exposure, identified as the “hottest hour” for a geographic location between 2010 and 2020. The hottest hour is determined by spatially weighting the weather station inputs across the state, with each station's weighted contribution representing its peak‑heat event during 2010‑2020. The resulting continuous weather field is used as an input by the Cool Air model to estimate the upper‑boundary layer statewide. Ultimately, the model solves for air temperature at every 30 m pixel via a coupled water–energy balance. See About i-Tree Cool Air below for more details.

Preliminary maps in raster format (composed of pixels²) depict up to three measures of extreme heat:

Air Temperature³: simulated at a 6-foot heightand 30m resolution for the maximum temperature between 2010-2020;
Heat Index⁴: a measure of how hot it really feels when the relative humidity is combined with the actual air temperature;
Urban Heat Island Effect (UHI Effect)⁵: The occurence of higher temperatures in cities than surrounding landscapes. This is calculated with the temperature difference from the most forested pixel. This was only calculated for urban areas.

Examples of temperature‑related raster outputs — Example raster outputs: maximum air temperature, heat‑index, and UHI effect.

Draft static maps for each of the state’s urban areas (the metropolitan statistical area, or MSA⁶ ) and counties highlight are available in two delineations:

Polygon Census block‑group maps⁷ (spatially less detailed);
Raster maps at 30 m × 30 m (spatially more detailed).

The maps also depict the state disadvantaged communities to mobilize extreme heat analysis and mitigation efforts for the most underserved areas of New York.⁸

Block‑group versus raster comparison — Polygon block‑group (left) and 30 m raster (right) map formats

Inputs

i-Tree Cool Air requires spatial raster inputs of land-use, impervious cover, tree cover, and elevation. It also needs weather and radiation data from a nearby station (often an airport) and allows numerous location-specific model adjustments from soil properties to impervious-surface albedo. For the statewide preliminary extreme-heat maps, the following inputs were used:

2019 National Land Cover Database (NLCD)⁹ Rasters:
- Land-Use (water, forest, grasslands, agricultural lands, urban areas)
- Impervious Cover (0 – 100%)
- Tree Cover (0 – 100%)
NHDPlusV2¹⁰: Digital Elevation Model Raster (meters)
NCEI Integrated Surface Database¹¹: Hourly meteorology for 2010–2020 for 28 weather stations (23 in-state)
Model Parameter Adjustments (Input .XML):
- Statewide Soil Type: Silty clay loam
- Regional Values:
- Location Values:

Urban Tree Cover over Impervious Cover map

Interactive map of all 28 weather stations used in the preliminary map simulations. Click on a station to view the hottest day and hour simulated by i-Tree Cool Air.

More About i‑Tree Cool Air

The i‑Tree Cool Air model uses terrain elevation to laterally redistribute subsurface water and represent soil wetness based on the topographic index, adhering to TOPMODEL theory.¹³ It operates in tandem with the i‑Tree Hydro model, which represents hydrologic similarity across 30 topographic‑index bins to simulate temporally varying soil‑moisture deficit (comparable to groundwater‑table depth).

The Hydro model was calibrated with two full water years (October 1, 2011 - September 30, 2013). Calibration produced three groundwater parameters—initial groundwater‑drainage rate, depth to vadose zone, and a hydraulic‑conductivity decay parameter—anchored to an average soil‑moisture deficit for each region, then passed to Cool Air for fully distributed 30 m simulations.

Examples of calibrated hydrological parameters for the Mid-Hudson — Example calibration parameters for the Mid-Hudson, left to right: initial groundwater drainage rate, depth to vadose zone (m), hydraulic conductivity decay.

Model skill was confirmed by simulating July 9, 2020 statewide and validating maximum daily temperatures against 160 station observations from the SC-ACIS network.¹⁴

Areas of the highest error are explained by higher tree cover in the local area that is not always reflected at a 30-meter resolution.

Graph of maximum daily temperature (observed vs. simulated) on 9 July 2020.

Validation Map — Spatial validation of maximum daily temperature on 9 July 2020.

Next Steps in the Project

Future phases will simulate historic (2010-2019) and mid-century (2050-2059) heat-wave scenarios under climate-change projections to study exposure risks to heat over time.¹⁵ Products from this phase will be available for download in early 2026. Unlike the preliminary single-hour, 30 m products, the next phase will provide 20 years of hourly and daily data at 300 m resolution.

Statistics will summarize:

National Heat Advisory — expected maximum heat index 100‑104 °F for ≥ 2 days; nighttime ≥ 75 °F¹⁶
National Excessive Heat Warning — expected maximum heat index ≥ 105 °F for ≥ 2 days; nighttime ≥ 75 °F¹⁶
New York State Heat Advisory — maximum heat index 95‑104 °F for ≥ 2 consecutive hours¹⁷
New York State Excessive Heat Warning — maximum heat index ≥ 105 °F for ≥ 2 consecutive hours¹⁷
Minimum, maximum, and mean daily values for each of the three data types: air temperature, heat index, and wet bulb globe temperature (WBGT) ¹⁷