Studies show that during the extreme weather mortality increases in urban areas. The mean of heat island in New York City during summer was estimated at 3.9 C and 3.0 C degrees during winter and spring. This study looks at fine scale structure in the urban heat island of Manhattan at street level in order to relate it to surface characteristics such as building geometry, vegetation, proximity to water, and albedo. If these relationships can be found and related to satellite derived classifications, a pathway exists to predict fine scale temperature variations at the street level based on satellite imagery. Eight mobile units bearing temperature and relative humidity sensors were used for this project. A preliminary campaign during the summer of 2011 measured the average Temperature and RH at street level in 8 different neighborhoods on 6 days. During the summer and early fall of 2012 an extensive field campaign deployed mobile sensors across parallel routes along both the short and long axes of Manhattan on 14 days. The results show some persistent patterns that could be related to surface characteristics. A surface classification scheme using LandSat 30 m resolution data shows promise in this direction. This work was funded as part of the health component of the Consortium for Climate Risk in the Urban Northeast (CCRUN), a NOAA Regional Integrated Science Assessment (RISA) that includes universities from New York, Philadelphia, and Boston.