Trees help reduce rising heat
Evaluating the impacts of trees on residential thermal conditions in Los Angeles using community science
A StoryMap about Objective 4 of the project “Planting Resilience: Assessing Vulnerabilities and Benefits of L.A.’s Urban Forest” funded by Accelerate Resilience L.A., a sponsored project of Rockefeller Philanthropy Advisors.
Background
As Los Angeles heats up, public health risks rise and solutions are needed
As the planet warms, heat-vulnerable communities in Los Angeles and in cities around the world face increased risks of heat-related problems including lost productivity, reduced learning outcomes, illness, and even death. Of all of the changes induced or exacerbated by climate change, extreme heat has the potential to impact the largest number of people is L.A. because many of the region’s residents lack the resources necessary to cope.
Continued warming is projected to increase average temperatures 4-5°F by mid-century, and by 5-8°F by the end of the century, with temperature extremes expected to be expressed both in the rising number of extremely hot days, and heat extremes growing by up to 10°F compared to today’s hottest days. Heat causes a host of health problems — ranging from those directly caused by heat exposure, such as heat exhaustion and heat stroke, to underlying conditions that become exacerbated by high temperatures, such as diabetes or cardiovascular diseases — and instances of these problems are expected to rise in California.
Like many shifts brought on or exacerbated by climate change, heat raises equity concerns, as the burden of extreme heat disproportionately affects low-income urban populations and people of color. These communities often live in neighborhoods of denser development that have older, lower-quality building stock, less urban forest cover, and fewer buildings with air conditioning — living conditions which contribute to a pronounced urban heat-island, and which create a feedback loop of heating effects.
Heat-health risk can be mitigated by trees, but there are critical gaps in knowledge
Despite the growing threat of heat, effective approaches to alleviate urban heat do exist, ranging from risk mitigation strategies designed to facilitate response during extreme heat events, to built environment strategies that focus on reducing urban temperatures. Air conditioning access is a remarkably effective approach to protecting health, but it is not a sustainable strategy in its current form because it generates climate-changing emissions and is often prohibitively costly for low-income households to operate. Tree planting is a heat mitigation strategy that has received investment in a growing number of cities around the world, but there are significant gaps in knowledge that stand in the way of optimizing the cooling potential of trees in the urban context.
One area that needs more research is understanding the effects that trees can have on indoor thermal conditions, and specifically on a room-by-room basis and at different times of day. This matters because people spend more than 85% of their time indoors, and when and where indoor activities happen in the home affects heat exposure and risk. Understanding thermal conditions in the bedroom, for example, is critically important during nighttime hours, when people are most likely to be sleeping and when the body attempts to rest and recover.
How trees affect microclimates, and what some of the tradeoffs are
Urban forests provide an array of benefits to urban communities ranging from carbon sequestration and improved air quality, to stormwater capture and provision of wildlife habitat.
At the neighborhood scale, trees change local climate conditions through shading and evapotranspiration, with one study finding that trees decrease park air temperatures by up to 11°F. In Los Angeles, city blocks that have more than 30% tree cover are about 5°F cooler than blocks without trees. The percentage of shaded tree cover over those streets accounts for more than 60% of land surface temperature variations, compared with only 30% of variation explained by factors such as topography and distance to the coast.
At the parcel scale, trees reduce temperature by providing shade, intercepting solar radiation, modifying wind patterns, and increasing humidity through transpiration. Cooling at this micro scale also impacts energy demand. Tree shade reduces building heat gain and shaded air conditioners work more efficiently, and simultaneously provide carbon sequestration and prevent the combustion of carbon by reducing energy demand.
However, the energy, ecosystem, and health protection services that trees provide are not free from tradeoffs, and it is important to think strategically about how and where to plant to maximize benefit and reduce risk. Trees can provide cooling through shading of buildings during hot weather but can increase the need for wintertime heating and have a wind shielding effect that reduces mixing and dilution of pollutants — potentially contributing to poor air quality. Lower wind speed influenced by trees can produce more conductive heat gain on surfaces in the built environment — a phenomenon that can be beneficial in cool weather but detrimental during hot weather. While shading and reduction of solar radiation by building-adjacent trees and vegetation reduce temperature, trees can raise indoor humidity — which can promote improved thermal comfort in dry climates or during dry heat waves can, but has the opposite effect in humid climates or during humid heat waves.
Understanding how to address these tradeoffs requires a holistic exploration of the topic, and additional research into how to best utilize trees as new climate records are set.
Methods
We evaluated the impacts of trees on indoor and outdoor thermal conditions at residential sites in Los Angeles using community science
We undertook a study to provide empirical evidence of the impacts of trees on indoor and outdoor thermal conditions in residential sites. We engaged residents in southeast Los Angeles County to host sensors in their homes. The original plan called for interested community members to host thermal sensors in their homes and allow study personnel to visit their home to install the sensors and download the data several times during the project period. As the COVID-19 pandemic unfolded, the scope was modified, and we pivoted to an approach that deputized participants as "community scientists."
Participants in seven homes installed and maintained a thermal sensor network, contributing continuous half-hourly readings for indoor (bedroom and living room) and outdoor (eave) temperatures for a period of 11 weeks.
Approximate locations of study sites, with orange icons showing sites that have little to no tree cover and green icons having moderate or high tree cover. Icons have been moved slightly and randomly in order to protect participant anonymity.
We engaged residents living in "non-treehouses" — residences where parcel tree canopy cover falls below L.A. County’s average canopy cover of 18%.
We also engaged residents living in "treehouses" — residences where parcel tree canopy cover exceeds L.A. County’s average canopy cover.
Each participant installed three thermal sensors that recorded half-hourly readings — for a total of 48 readings per day — for a period of 11 weeks between September 1 and November 15, 2020. One sensor was installed in the bedroom and another in the living room.
A third sensor was installed outdoors, under the eave. Outdoor sensors were placed in a small paper cup to help shield them from "noisy" readings that might be caused by direct sunlight or strong wind.
We then compared thermal conditions at treehouses vs. non-treehouses on "hot days" — days with a maximum temperature at or above 90°F as recorded at the nearby National Weather Service Los Angeles Downtown/USC weather station — versus "non-hot days" below 90°F.
Results & Discussion
On average, we found that indoor temperatures in treehouses warm 1.1°F less on hot days compared to non-treehouses. If homes in heat-vulnerable parts of Los Angeles were 1°F cooler we could reduce heat-related deaths by 10-20%, and with additional tree canopy and solar reflectance increases the number of lives saved could grow to 30% or more.
Bedroom Average Temperatures
This graphical version of the difference-in-differences estimate for bedroom temperatures shows that over the study period, bedrooms in treehouses actually experienced 2.1°F higher average temperatures on non-hot days. There are a host of reasons why this could be the case, including building materials and sun exposure as a function of the orientation of the bedroom relative to the rest of the house. This fact alone does not diminish the potential of urban cooling by trees, and it underscores the aptness of a difference-in-differences research design.
The data show that on average, bedrooms in treehouses are 5.0°F warmer on hot days than on non-hot days, and that bedrooms in non-treehouses are 6.1°F warmer on hot days than on non-hot days. The difference between the two groups of homes being 2.1°F on non-hot days and shrinking down to 1.0°F on hot days suggests that trees have a 1.1°F dampening effect during extreme heat conditions. Without trees, we would expect that treehouses would be hotter and expose residents to higher temperatures.
Living Room Average Temperatures
Living rooms in treehouses are 1.2°F warmer on non-hot days and 0.2°F warmer on hot days relative to non-treehouses, with a difference-in-differences of approximately 1.0°F. The estimated effect for the living room is similar to our estimate for the bedroom, indicating the benefits of trees are not confined to one area of the house.
Eave Average Temperatures
Averaged over the study period, treehouse temperatures are actually warmer outdoors than at non-treehouses. In contrast with indoor temperatures, we see that eave temperatures in treehouses actually rise by a greater amount than eaves in non-treehouses during hot weather. On average, eaves at treehouses are 10.5°F warmer on hot days than on non-hot days, whereas non-treehouse eaves are 9°F warmer on hot days than on non-hot days. The difference between the two kinds of sites is 1.1°F on non-hot days and grows to 2.6°F on hot days, suggesting that tree houses are actually warming 1.5°F more outside on hot days.
There are a variety of site-specific reasons that could account for this unexpected finding, and while we cannot conclusively ascribe this differential to any specific factors given the data at hand, we expect that average eave temperatures in treehouses would grow even more significantly if trees were absent. This fact suggests that our findings above, which already support a cooling benefit of trees, might even be understated.
The regression table replicates the prior figures in table format and presents estimates that are identical to those previously shown.
Bedrooms in non-treehouses are 6.1°F warmer on hot days than non-hot days (Hot day >90F). Bedrooms in treehouses are an average 2.1°F warmer than non-treehouses on non-hot days (Moderate / High tree cover), but temperatures in treehouse bedrooms increase by 1.1°F less than they do in non-treehouses (Tree x Hot day), once again pointing to indoor temperature modulation impacts of trees. Given the standard error of 0.28, the estimates are statistically significant (p = 0.0000). The number of observations varies due to variations in thermal sensor performance over the 76-day study period. The community scientist nature of the project led to data downloads occurring sporadically, at times causing a delay in identifying and troubleshooting sensor issues.
While a difference-in-differences of 1.1°F is small, we note that this study was intentionally conducted in neighborhoods that have low tree cover in order to yield data about the parcel-level function of trees while excluding potential neighborhood-level tree cover influence. Even where the parcel had high tree cover, we expect no additional tree cover benefit to come from neighborhood-level tree cover, because all neighborhoods have less than the L.A. County average of 18% tree cover.
On Sept. 6, 2020, L.A. County recorded its highest-ever temperature —121°F in Woodland Hills. On that day, the high for our study’s reference weather station at Downtown/USC was 111°F. The hottest of the study sites — a residence in Huntington Park with no trees or air conditioning — topped out at 107.4°F in the living room and 99.7°F in the bedroom. Such extreme temperatures are dangerous even for healthy people, and sustained exposure can prove deadly. As the planet warms and Los Angeles becomes more prone to hotter and longer heat waves, heat-protection strategies are needed to prevent more Angelenos from being in harm’s way.
Bedroom Hourly Averages
Looking at the same data in hourly average format, we see that on hot days the difference in temperatures between bedrooms in treehouses and non-treehouses is smaller at all times of day than it is on non-hot days, suggesting a temperature attenuation effect by trees.
The fact that the benefits extend to nighttime hours is particularly beneficial to public health, because while occupants are sleeping the body seeks to recuperate after the day’s heat exposure. Indoor peak temperatures occur around 5:00pm, later than outdoor peak temperatures, as heat continues to be retained and conveyed even after outdoor temperatures begin to cool off.
Living Room Hourly Averages
Living rooms in non-treehouses are generally cooler. However, we see that temperatures in treehouses increase by a lesser amount on hot days and that non-treehouse temperatures actually exceed those in treehouses as daily temperatures increase between about 11:00am and 6:00pm. This shows that trees have an even larger cooling effect in living rooms during hours when daily temperature is on the rise. This switch is not observed in bedrooms and could potentially be attributed to factors such as insulation quality and azimuth/cardinal direction of the living room relative to the rest of the house.
Eave Hourly Averages
The data show that outdoor temperatures in treehouses are on average higher than those observed at non-treehouses during the cooler parts of the day. Importantly, we see that the relationship flips during peak temperature hours (between about 12:00pm and 5:00pm), when temperatures at treehouses are cooler. This occurs both during non-hot and hot days, though the differential at the coolest part of the day is larger on hot days.
These observations suggest different possibilities: a) trees provide some, albeit relatively less, cooling at night than during the day, or b) trees trap heat and have a warming effect at night. Nighttime warming is attributable to wind shielding and longwave radiation emitted from the ground being reflected by the tree back down to the ground due to limited sky view factor.
Disentangling these two competing hypotheses is difficult with the limited data at hand. But the first hypothesis seems more likely since the theoretical benefits of trees are largest during the hottest part of hot days, suggesting that we would expect there to be even less difference between day and night for treehouses on hot days, which is what we observe.
Conclusions
This study contributes new empirically-derived support for the heat-protective function of trees in an urban environment. We find that on average, bedroom temperatures in treehouses warm 1.1°F less on hot days compared to non-treehouses and 1.0°F in living rooms. These temperature benefits extend to all times of the day, which is critical from a public health perspective given nighttime vulnerability to heat exposure. These temperature reductions can help reduce heat-related public health costs among heat-vulnerable communities — a fact of critical importance as the study also finds that exposure to extreme heat can and does reach dangerously high levels in older residences without trees or air conditioning.
Future directions for this research could include a larger-scale study involving 100 to 200 homes segmented by neighborhood and site characteristics. This would enable a deeper exploration of tree and housing type characteristics. Incorporating household-level energy data for the study period would enable quantification of the impacts of trees on energy demand. Such an analysis could be linked both to in situ sensors, such as the ones used in this study, and remote-sensed land surface temperature data.
Further investigation of the daytime vs. nighttime effects of trees on thermal conditions is another critical area that should be explored, especially in the context of how exposure to heat at different times of day and in different rooms of the house impacts public health outcomes.
Acknowledgments
The Institute of the Environment and Sustainability at UCLA acknowledges our presence and the presence of the homes involved in this study on the traditional, ancestral and unceded territory of the Gabrielino/Tongva, Chumash, and Kizh peoples.
The study team wishes to thank:
- The participants of this study and their families for their curiosity and commitment to growing our collective knowledge about the role of trees in making our neighborhoods more livable.
- Andy Lipkis, Jen Bravo, Deborah Bloome, and Zenya Prowell at Accelerate Resilience Los Angeles for the partnership and support provided during this study.
- Luis Rodriguez at TreePeople for his role co-managing the project and liaising with study participants, Dr. Yujuan Chen for her role co-managing the project, and Alejandro Fabian and Eileen Garcia for their help in recruiting participants.
- Charlotte Schulman at Vassar College for compiling information on thermal sensor selection and placement.
- Dr. V. Kelly Turner and Dr. Aradhna Tripati at UCLA for advising on research design.
Photos credits: TreePeople, Pexels
