Microclimate Variation in the Suburban Environment

C.K. Smith, H. Akbari

One focus of research in the Heat Island Project is the use of shade trees and high-albedo surfaces to reduce urban temperatures. In addition to cooling surfaces and structures in their immediate vicinity, such features can contribute to the reduction in air temperature over an entire neighborhood, resulting in a neighborhood-wide or indirect cooling effect.

Experiments to assess the potential magnitude of these indirect effects often compare air temperatures measured near the ground in neighborhoods having different albedos or densities of trees. Necessity dictates the measurements be made at one or a few discrete locations. Yet the urban and suburban environments are filled with man-made and natural features that alter the ambient air temperature in their vicinity, i.e., they give rise to localized or microclimate effects. If discrete air temperature measurements are to reflect an entire neighborhood, the influence of microclimate effects on these measurements must be minimized.

We performed an experiment to monitor ambient air temperature at 18 locations around a suburban residence, from which we identified a multitude of microclimate effects. Temperature data, along with horizontal insolation, relative humidity, and wind speed and direction, were logged every 2.7 seconds for the period September 11 to November 7, 1994.

In the absence of any objective measure of the unperturbed neighborhood climate of the site, we used the spatial average temperature, derived from 13 locations monitored continuously over the course of the experiment. For each location, the difference from the spatial average temperature was taken to be a measure of the microclimate effects.

The microclimate effects at each location followed a diurnal pattern characteristic of the local environment. Fourteen days of temperature-difference data were co-averaged to show a diurnal microclimate profile for each location (see Figure).

Figure. Diurnal patterns of microclimate variation (expressed as the temperature difference from the average): (a) near house; (b) under large walnut tree; (c) next to redwood wind break; (d) in open yard.

Locations close to the home showed a daytime peak in their difference profiles, while those near the location of shade trees and away from the home showed a daytime trough. The amplitude of the features reached 2 to 3°C on a warm day (25°C). The vertical offset of the diurnal profile was also related to features in the local environment. Nighttime behavior is governed mainly by the view factor (fraction of full sky seen from each location) and the distance from high thermal mass features (such as the home).

Some locations had diurnal difference profiles that peak at about 0°C (i.e., the temperature was close to the spatial average). The same locations show a depressed nighttime temperature relative to the spatial average. They have high view factors, are away from surfaces of low reflectivity, and are out of the wind shadows of structures. These are optimal locations for neighborhood temperature measurement.

The sensitivity of the microclimate variations to changes in the overall temperature was explored by regressing the daytime extremes of the temperature difference against the maximum daily average temperature. The daily peaks generally became higher and daily troughs became lower by a few tenths of a degree per 10-degree increase in average temperature. Put simply, the microclimate effects that dominate during the day become stronger on a warmer day.

The results of the analysis allowed us to formulate a protocol for minimizing the effects of microclimate variation on neighborhood temperature measurements in the urban and suburban environment. The protocol will be used in designing experiments to assess the impact of indirect cooling effects on urban air temperatures.

Reference

Smith CK, Akbari H, Bretz S. Microclimate Effects Near the Ground in the Suburban Environment. Lawrence Berkeley National Laboratory Report No. LBL-37876, 1996.

| Return to Contents | Return to Previous Paper | Next Paper |