Suburbanization, Land Use Policy, and Heat Islands: A Case Study of the NY-NJ-CT Metropolitan Region

The objective of the paper is to evaluate the spatial structure of urban heat islands (UHI) in the New York-New Jersey-Connecticut Metropolitan (Tri-State) Region over the last century. On average the UHI effect denoted by the average minimum temperature difference was 3 degrees Celsius between Central Park, NY (urban stations) and the selected three rural stations (Mohonk Lake, NY; Falls Village CT; and Freeland, PA).

The Tri-State region has observed extensive and intensive growth that has impacted the regional climate regime (13,100 square miles). Suburban spatial development patterns have expanded UHI effect, reducing the amount of naturally vegetated areas and increasing the rates of impervious land consumption. Prior to 1965 the number of acres per person was approximately a fifteenth of an acre however today; those average rates are over a 1 acre per person. Professionals in the Tri-State Region realize that UHI effect and extreme heat circumstances contributed to over 150 fatalities in cities and old inner ring suburbs during the summer of 2006. This increased the amount of attention now given to UHI in a board range of fields including urban and regional planning. This is combined with recent recognition that UHI effect triggers and exacerbates other urban environmental hazards including increased air pollution, energy consumption, and increased heat related stress and mortality rates.

Multiple data sources and scales were investigated to develop a decision support system in a Geographic Information System (GIS). This paper combines historic station observation data from the United States Historical Climatology Network (USHCN) at 33 stations containing data from 1898 to 2005. To develop a more robust knowledge of the spatial dimensions of the UHI effect, additional air/surface temperature data was gathered data from USDA-NRCS Spatial Climate Mapping Project called Parameter elevation Regressions on Independent Slopes Model (PRISM) System products developed by Oregon State University. The PRISM dataset provided for a continuous spatially distributed monthly temperature from 1895-2005 for the entire region at 4km resolution. Statistical analysis provided information by monthly, seasonal, and annual for the period 1900 – 2005. Additionally it was decided to investigate UHI for micro-scale heat island effect. This was done using remote sensing data gathered from the MODIS Daily dataset which provides the opportunity to explore the diurnal pattern of UHI at the regional scale. Lastly, a Landsat TM dataset from 2002 provides the opportunity to examine the UHI effect at the neighborhood scale.

Since the UHI effect is closely linked to land-use type, urban form, presence of vegetation, the weather and human activities that release additional anthropogenic heat. This research seeks to established a statistical relationship between UHI effect and a set of explanatory variables following an urban-suburban-rural gradient that can be closely tied to land use and urban planning policies. The characterization of these differing UHI conditions allows for specification of a range of possible UHI mitigation scenarios, such as identifying areas for tree planting programs, Energy STAR programs, or building material changes.