The Effect of Trees on Summertime Below Canopy Urban Climates: A Case Study Bloomington, Indiana

The study area

The measurements were conducted in Bloomington, Indiana (39°10’N, 86°31’W), a typical mid-size city in the Midwest, USA, with an active tree planting program. Summertime throughout the Midwest is hot and humid, consequently air conditioning demand is high. The data for this study were collected during a 15 day period August 7 - 21,1991 (inclusive). The measurement period was very typical for August, with maximum daily temperatures between 28 and 30°C, minimum temperatures 18 - 20°C, and average dew point temperature 18.2°C (12). Most days in the period were sunny without clouds. A few days had afternoon cumulus, and one day was uniformly overcast with heavy rain in the morning. Measurements were not conducted on this day.

Adult, well established, healthy trees were chosen for the study. Where possible ten similar trees (replicates) were identified and studied in each of the following categories:

• 1) Acer saccharum (sugar maple) individuals over grass

• 2) Acer saccharum individuals along streets (over concrete)

• 3) Acer saccharum clumps of trees over grass (3 or 4 trees define a clump)

• 4) Juglans cinera (black walnut) individuals over grass

• 5) Quercus palustris (pin oak) individuals over grass

Sugar maple, pin oak and walnut were chosen because they are relatively common and are actively being planted in urban areas of the Midwest. The three environments studied: individuals over grass, clumps over grass, and individuals along streets over concrete, are common niches for trees within urban areas. Because of problems of finding well matched trees only 7 pin oak and 7 sugar maple in clumps were studied. The replicates in each category, the 44 trees overall, represent a trade-off between logistical constraints of manually taking measurements (see below) and the desire to maximize the number of replicates of each species in each environment. All individuals over grass and the clumps of sugar maple were located on the campus of Indiana University, located in the center of Bloomington (Figure 1). The sugar maple individuals along the roadside were located along a downtown street, in close proximity to the University, with mixed commercial and residential use (Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1:

Location of each of the trees studied in relation to Indiana University, Bloomington, Indiana

Field measurements

All trees were physically described by measuring height (using a hand-held clinometer), diameter at breast height (dbh), height of the base of the canopy, and the radius of the canopy to the dripline (Table 1). The leaf area index (LAI) of each tree was determined using a Licor LAI 2000 Plant Canopy Analyzer (PCA). The PCA estimates of LAI (Table 1) are lowerthan expected. As yet this technique has not been independently tested for isolated trees in urban areas, so the absolute values must be treated with caution. However, the LAI data obtained provide a basis for relative comparison between the trees studied.

Measurements were taken below the tree canopy, at a height of approximately 1.5 m, representative of the environment in which humans live. Wet and dry bulb temperatures were taken simultaneously at the edge of the canopy and underthe middle of the canopy cover (half distance from trunk to dripline). The circuit to measure the trees took approximately 2 hours to complete. To remove the effects of day to day variability in synoptic conditions and measurement time, reference data from the IU climate station were collected. All below tree-canopy readings were compared to these reference data and differences between the open reference site and under the trees calculated. The IU climate station is located at the edge of the campus of Indiana University (Figure 1), at an open site over grass. Intercomparison of the instruments at the climate station and those used to study the trees showed combined systematic and unsystematic measurement errors on the order of the resolution of the instruments (temperature ±0.12 °C). Consequently the data were not adjusted.

Under the trees, temperature and relative humidity were determined using sling psychrometers, which measure wet (T_w) and dry (T_d) bulb air temperatures (°C). From these two temperatures actual (e_a) and saturation (e_s) vapor pressure (mb), and relative humidity (%) can be calculated:

where e_a is actual vapor pressure (mb), e_s is saturation vapor pressure (mb) calculated (5):

C^p is specific heat of air at constant pressure (1010 J/kg/K), E is 0.622, the ratio of the molecular weight of water vaporto that of dry air, L_vis latent heat of vaporization (J/kg):

with T_d dry bulb temperature (K), and p atmospheric pressure (mb). Relative humidity (%) can then be calculated:

Relative humidity is strongly inversely correlated with temperature because of the temperature dependence of the saturation vapor pressure (e_s). Forthis reason, in the subsequent analysis, actual vapor pressure (e_a) is also considered to investigate the effect of the trees on the vapor content of the atmosphere.

Results are grouped into three times of the day: early morning (0700 - 0900 LAT); midday/early afternoon (1200 -1400 LAT); and late afternoon-early evening (1700 - 1900 LAT). Four or five sets of readings for each time of day were taken over the 15 day period. Thus there are 24 - 50 measurements for each category of tree at each time of day, 614 observations in total.

The effects of trees

Morning temperatures under all of the other trees were close to the reference site (0.7°C warmer to 0.1 °C cooler) (Table 2). By midday all trees are significantly cooler than the open site by the order of 0.7-1.3°C (Table 2). The notable exception is the street trees, which in the middle of the day are only slightly cooler on average (0.1 °C). In the evening it is slightly warmer under all the trees, significantly so under the sugar maple street trees and walnut (0.4 - 0.6°C). The warming in the evening is likely due to the reduced sky view factor under the trees: longwave radiation emitted by the ground to the tree is reflected, absorbed and re-emitted back to the ground.

The midday reductions in temperature under the trees in this study, while statistically significant, are lower than those documented in previous studies. However, this reduced effect is to be expected. Bloomington is located further north than the previous study sites (Sacramento, Phoenix, Miami, etc). Consequently solar radiation is lower, and the trees would be expected to have less of a shading effect. More importantly, the humidity in the Midwest is much greater than the southwest, resulting in reduced vapor gradients and evaporative cooling.

Relative humidities were statistically higher under all the trees than at the open site (Table 3), except for pin oak in the morning. Relative humidity was highest in the morning, due to the combination of lower temperatures and the evaporation of dew. Through midday and evening, relative humidity under the trees was 27-33% greater than at the open site.

The absolute level of vapor pressure at all sites decreased through the day (Table 4). Most evaporation (as opposed to transpiration) occurs in the morning when solar radiation is coming from a low sun angle and there is dew on the surface which can be readily evaporated. By midday evaporation of soil and grass moisture decreases as less than 20% of the incoming solar radiation may penetrate the canopy (13). Vapor pressure continues to decrease in the late afternoon-early evening as evaporative demand induces the trees’ stomata to close. Vapor pressure is greater at the open site in the morning. The difference by the middle of the day becomes significantly greater below the trees and remains so through the evening.

Conditions at the edge of the tree canopy

Higher temperatures are observed at the edge of the canopy than under the middle of the tree (Table 2). Where significant cooling occurs under the tree in the middle of the day, no such effect is observed at the edge of the canopy (Table 2). At all times of the day the conditions at the edge of the street trees are warmer than the reference site. However, this is to be expected given that the reference site is located over grass and the street trees grow over concrete.

Very similar results are observed both under the tree and in the middle of the tree for relative humidity and vapor pressure (Tables 3 and 4), i.e. moisture is significantly enhanced around, not just under the tree. In general, there is greater variability at the edge of the canopy (greater standard errors, see Tables 3 and 4). These results suggest that there is a marked gradient away from the trees, and the edge is less affected by shade and more by small scale advection of heat.

Effects of tree species

In terms of deviations from the refernce site, in the morning, under the middle of the canopy, pin oaks are significantly warmer than both sugar maple and walnut (Table 5), and the reference site. A distinguishing feature of pin oak is the drooping of their lower branches toward the ground, which may reduce overnight long-wave radiative losses from underthe canopy. In contrast, Sugar maple has a longer trunk, with branches that extend straight out, thus there should be a greater movement of air through the canopy, reducing the temperature. By midday/early afternoon there are no significant differences between the species (Table 5a). By evening walnut and sugar maple are significantly different, with walnut warmer on average, although neither of the two species is significantly different from pin oak.

In terms of relative humidity and vapor pressure, all the trees are significantly different from one another in the morning (Table 5a). Walnut and sugar maple have different relative humidities in the evening (as with temperature) but no difference in vapor pressure, indicating the difference in relative humidity is likely an artifact of a temperature control.

Trees of the same species in different environments

Differences between environments are more significant than between species in the same environment (compare Tables 5a and 5b). In terms of mid-canopy temperature, trees growing along streets are warmer than trees growing alone or in clumps. By evening, the street trees are approximately 0.5°C warmer (Table 2). At this time the surrounding concrete may be reflecting and re-emitting radiation back onto the trees resulting in higher temperatures. There are no significant differences between the clumps and individuals.

The vapor pressures below the canopy in all three environments are remarkably similar (Tables 4 and 5). The street trees are consistently lower, although only significantly so in the morning, probably due to the absence of evaporation of dew from grass. In the morning and evening, vapor pressure is less variable under the street trees than in the other environments, a pattern indicative of more uniform stress for all the street trees. The pattern of relative humidity largely reflects (inversely) the pattern of temperature. It is highest under street trees in the morning then decreases as temperature rises, even though the moisture content of the air (vapor pressure) is increasing. By evening it decreases, despite decreasing temperatures, as a consequence of changes in vapor pressure.

Correlation of effects with physical characteristics of the trees

Linear correlations were determined between the temperature and vapor pressure reductions for each time period and physical characteristics of the trees (dbh, LAI, and radius of the tree to dripline). Firstly for each of the five classes of tree separately, and secondly for all the trees together (Table 6). Caution should be exercised in interpreting the first set of results because of the small sample size.

No consistent significant correlations were determined between temperature reduction under the trees and their physical characteristics. The only significant correlations were with LAI for sugar maple clumps in the morning, and sugar maple street trees in the middle of the day. The only significant correlations for vapor pressure were with LAI for walnut in the morning. These results suggest that factors such as the exact location of the tree are more important than physical characteristics of the tree, at least in the urban situation, in determining the microclimate under a tree canopy.