Transpiration effect is expected to mitigate urban thermal environment but efforts to predict urban trees’ transpiration are complicated by our limited knowledge of how much and different whole tree’s water use is. The difficulty in understanding urban trees’ transpiration is mainly caused by the limitation of measurement method, so we newly developed hanging-type gravimeter method using container grown trees and beam-type load cells, which is convenient to weigh a number of samples in a short period.

The purpose of this study is to (1) Classify morphological and physiological characteristics which help us to predict transpiration rate and (2) Investigate transpiration responses to vapor pressure deficit (VPD) as represented by the change of canopy conductance (Gc). We weighed 11 trees of 11 popular urban tree species in Japan at farm field, Aichi prefecture (35°8′N, 137°6′E) in summer 2010, 2012 and 2013. Studied trees are under isolated and soil-restricted condition which assuming urban environment and ranged 3~6.7 m in height, 1.5~9.2 m2 in crown projection area, deciduous/evergreen, and diffuse-porous/ring-porous/tracheid in xylem porosity.

First, we verified our new gravimeter accuracy using high time resolution recording and AR model and showed the mean value of relative error (95% C.I.) in half-hourly transpiration rate of all studied trees was 9.2%, and the error of Quercus myrsinifolia in 2012, which use least water and have largest error, was 19.8% when wind velocity is under 2 m/s.

Second, we compared transpiration rate of summer irrigation days and found (1) Daily water use vary 10~30 kg in individuals, equivalent to 0.62±0.36 times of water evaporation from the same size of crown projection area, (2) Maximum latent heat was 2.3 kW/tree of Zelkova serrata in 2010 and Quercus serrata, (3) Zelkova serrata and Quercus myrsinifolia changed Gc threefold due to physiological change, (4) Ring-porous species tend to use much water (p=0.045), which suggest the significance of hydraulic capacity under volume-restricted soil condition.

Finally, we tested the hypothetical relationship between Gc at VPD = 1 kPa (Gcref) and the Gc sensitivity to VPD (-dGc/dlnVPD) suggested by Oren et al.(1999) [Plant, Cell and Environment 22, 1515–1526]. The results showed good agreement between theory and measurement across almost all studied trees, but few were rather insensitive.

One of the ecosystem services provided by urban trees is the cooling effect caused by their transpiration. However, while the transpiration of forest trees has been widely studied, little research has been conducted on the daytime and night-time transpirational cooling effect of mature urban trees. Knowledge about the transpiration of street and park trees and its response to different environmental factors can prove useful in estimating the thermal influence of urban greenery as well as in urban planning and management. The aim of this study is to i) quantify the magnitude and diurnal variation of transpiration of common urban tree species in a high latitude city (Gothenburg, Sweden), ii) analyse the influence of weather conditions and fraction of permeable surfaces within the vertically projected crown area on tree transpiration, and iii) find out whether transpiration of urban trees remains active during the night and therefore contributes to nocturnal cooling.

Measurements were conducted on mature street and park trees of seven tree species common in Gothenburg: Tilia europaea (Common lime), Quercus robur (English oak), Betula pendula (Silver birch), Acer platanoides (Norway maple), Aesculus hippocastanum (Horse chestnut), Fagus sylvatica (European beech) and Prunus serrulata (Japanese cherry). Stomatal conductance and leaf transpiration were measured using a LI-6400XT Portable Photosynthesis System (LI-COR Biosciences) at daytime and night-time on warm summer days of 2012-2013 in Gothenburg. Leaf area index (LAI) of the studied trees was measured with a LAI-2200 Plant Canopy Analyser (LI-COR Biosciences) in order to estimate the latent heat flux due to tree transpiration.

Leaf transpiration was found to increase with vapour pressure deficit and photosynthetically active radiation, with on average 22% of the midday incoming solar radiation being converted into latent heat flux. Midday rates of sunlit leaves varied between species, ranging from less than 1 mmol m-2 s-1 (B. pendula) to over 3 mmol m-2 s-1 (Q. robur). Daytime stomatal conductance was positively related to the fraction of permeable surfaces within the vertically projected tree crown area. A simple estimate of available rainwater, comprising of precipitation sum and a fractional surface permeability within the tree crown area, was found to explain 68% of variation in midday stomatal conductance. The results indicate that a high fractional surface permeability can minimize the frequency of water stress experienced by urban trees and enhance their transpirational cooling.

Night-time transpiration was observed in all studied species and was positively related to daytime tree water use. Nocturnal transpiration amounted to 7% and 20% of midday transpiration of sunlit and shaded leaves, respectively. With an estimated latent heat flux of 27 W m-2, evening tree transpiration enhanced the cooling rates around and 1-2 hours after sunset, but not later in the night.

The results of transpiration measurements will be combined with vegetation data derived from LIDAR and LAI measurements to estimate neighbourhood- to city-scale cooling effect provided by urban trees.

Increasing the vegetation cover in cities is one of the key approaches to mitigate urban heat. Trees are the most important vegetation type since from their height they intercept sunlight, cast shade and cool the air by transpiration. However, still very little is known about the magnitude of the cooling related to the transpiration and how this varies on a diurnal scale and in response to heat and drought stress.

Here we present results from a study linking diurnal tree transpiration to urban cooling by comparing daily urban air temperature and humidity pattern from June 2013 to October 2014 in areas of contrasting tree cover (suburban park, an urban garden, a courtyard, a street and a square) in the city of Mainz, Germany. Transpiration is indirectly estimated by high resolution measurement of sap flow and stem radius change , which are indicators of tree stem water balance, in three common tree species in European cities; Platanus × acerifolia, Acer platanoides and Tilia platyphyllus. The measurements were analyzed in view of general weather situation and water availability in order to evaluate the effect of tree heat and drought related stress on the diurnal cooling.

As expected, we observed a cooling effect of the examined trees on the urban air temperatures both at day and night, although nocturnal cooling were more pronounced. Preliminary results of stem water balance indicate that transpiration of the examined trees can be linked to the magnitude of the cooling at a diurnal scale, since stronger cooling observed in the vegetated areas significantly correlate with increasing sap flow and daily stem radius changes. The cooling effect was especially pronounced in the examined urban courtyard. In this location, five large trees in an otherwise bare and dense urban environment caused a stronger cooling compared to the sparsely vegetated street directly outside, and humidity patterns more similar to those found in the suburban park than to the street.

The examined trees did not suffer from heat and drought stress, except during a prolonged hot and dry period during the 2013 growing season. Towards the end of this period we could observe that the stressed trees changed their evapotranspiration pattern. This changes corresponds to a weaker daytime and a stronger evening cooling of the air, possibly caused by a drought adaptive behavior where mid-day evapotranspiration is reduced for water saving purposes.

In this presentation, we present a first indication on how the cooling effect of urban trees vary on a diurnal scale, as well as in relation to heat and drought stress. This could provide very useful information in the approaches to mitigate heat stress for the urban inhabitants in a future warmer climate.

Planting trees is one of the countermeasures to urban heat island because of the cooling effect by the transpiration or the sunlight cover effect. It is difficult to construct the large scale forest in urban area, so isolated trees such as roadside tree are generally planted. In the case, it is desirable that trees are efficiently located on understanding the performance of heat exchange between the tree crown and the atmosphere. The heat budget related to the performance of a tree cannot be directly measured in detail. In this study, heat budget of isolated plant unit is numerically evaluated with thermal plant model including radiation transfer model for net radiation and transpiration model for latent heat transfer, and the numerical results are validated by comparison with that of outdoor measurement for a potted plant.

Heat budget is evaluated as the balance of net radiation, sensible and latent heat transfer between the leaves and the atmosphere and conduction heat transfer to the foliage. Conduction heat transfer is omitted in this study. Latent heat transfer related to transpiration on leaf surface can be evaluated by Jarvis model. Net radiation of foliage is evaluated by solving the Ross’s radiation transfer equations and sensible heat transfer is by difference between the net radiation and latent heat transfer.

In the prediction of net radiation of foliage, Computer graphics (CG) model with software AMAP, which can draw tree characteristics such as shapes of a tree crown and a leaf and leafing arrangement on a branch, is applied to decision of structure parameters of leaf surface density and distribution function of leaf surface direction vectors which are included in Ross’s radiation transfer equations in plant vegetation. It is necessary for evaluation of the net radiation to consider direct and diffuse insolation from the sky and scattering light in tree crown for the component of short wavelength and thermal radiations from surroundings including atmospheric radiation and from leaf surfaces for the component of long wavelength. Leaf temperature which is necessary in evaluating the thermal radiation from leaf surfaces and transpiration rate is evaluated as satisfying the heat budget on leaf surface. Parameters of Jarvis model in transpiration of leaves is decided by measurement of transpiration rate of actual leaves of several kinds of trees. By applying the numerical simulation, shielding efficiency of insolation and ratio of latent and sensible heat transfers to net radiation of isolated and vegetated tree are evaluated for several kinds of trees.

Heat stress modification capacity of urban trees is widely acknowledged and makes these natural landscape elements very important in the field of climate conscious urban planning. Many studies (measurement and/or modeling based investigations) have proven already that shading, i.e. the reduction of direct solar radiation is the most effective way to moderate summer heat stress among Central European climatic conditions. Shading potential of trees is described usually by their transmissivity values. Although the transmissivity depends on the leaf density, leaf orientation and other tree crown-related characteristics, most microclimate simulation software set this attribute as default in the case of all trees. The first goal of our investigation is to determine the transmissivity for four tree species which occur frequently in Hungarian cities: Sophora japonica, Tilia cordata, Celtis occidentalis and Aesculus hippocastanum. To accomplish, one-year long systematic measurement campaign is carried out in the South-Hungarian city of Szeged. Beside the measurement of short-wave radiation under the crown of different species, other microclimatic parameters and the actual condition of the tree crown (seasonal status of the canopy, health conditions) are also recorded. Although the results indicate clearly different average transmissivity values for the investigated species, transmissivity varies in a wide range in every case depending on many influencing factors (e.g. intensity of unobstructed global radiation, seasonal condition of tree crown). As a next step, simulations are carried out to investigate the effect of tree species selection (meaning transmissivity differences) on the resulted reduction of radiation load. Solweig software is applied to model the spatial distribution of mean radiant temperature (Tmrt) on a square planted with different types of trees incorporating the measurement-based transmissivity values. These types of assessments can help finding the appropriate tree species for urban landscape planning. Besides the micro-scale results, they can contribute to the methodological development of local scale heat stress mapping, moreover to the indicator development for mapping climate regulation ecosystem services of urban green spaces.

Water availability is widely recognized as being an essential factor for tree survival, growth and for maximizing their ability in mitigating urban heat islands (UHI) through evapotranspiration. In urban areas, where the ground surface is highly impervious and the trees are not regularly irrigated, the reduction of precipitation infiltration into the soils may put the trees face an increased water stress1,2. It is also generally predicted that trees in urban sites have higher water losses than trees in natural forests due to increased evapotranspiration demand3. There is currently insufficient data to generalize the physiological responses of trees to the complex urban environment, where both climatic and management factors are entangled. Especially, little information is available on the effect of water stress on tree health, and their consequences on ecosystem services such as UHI mitigation. Furthermore, on-going climatic changes make it all the more necessary to anticipate the potential trajectories of urban trees, in terms of (i) risk assessment of tree survival, and (ii) their potential ability to provide cooling services. In this context, a retrospective approach of the long-term relations of trees to urban climate can provide a way to enhance our understanding of current urban tree hydric state and to gain insights on future levels of water stress levels under future climatic conditions.

It is well known that there is a close relationship between tree growth and climate. Indeed, the size and the state of tree-rings are affected by the yearly sequence of favourable and unfavourable climates4. In turn, climate phenomena can be identified and reconstructed through ring-width sequences5. Thus, dendrochronology and study of wood structure can be used as informative tools in order to understand the long-term influence of past climate on urban trees growth. Consequently, understanding the past trajectory of trees under the past climate provides insights on their response to future climate projections. Since tree cooling potential is tightly linked to water availability, negative feedback of water stress to tree cooling potential can be expected.

To further investigate these issues, we are implementing a 3 years project, financed by CDC-Biodiversité and we are testing this approach by studying 95 urban Tilia tomentosa Moench individuals in Paris (France), selected along a 10-120-year age gradient. Tree age was estimated using wood core samples, and the health status of each tree was visually assessed according to the "Visual Tree Assessment". Number of tree-rings and ring-width were measured using a Lintab measurement table and compared to meteorological data at local and regional scales (Paris, Île-de-France region). The xylem vessel diameter and density were used as a proxy of hydraulic conductance efficiency, embolism vulnerability and health status of trees6,7. The sampling design will enable a comparative approach of trees of different ages, which is used to assess whether chronic water stress is influenced by both tree age and climate history.

The communication describes the current results, and we use them to discuss the ways climate change could further impact tree-growth and thus potential tree cooling effects in Paris City.

Key references

1 Whitlow and Bassuk (1987) A 3-year study of water relations of urban street trees. Journal of applied ecology 29: 436-450

2 Clark and Kjelgren (1990) Water as a limiting factor in the development of urban trees. Journal of Arboriculture 16: 203–208

3 McCarthy and Pataki (2010) Drivers of variability in water use of native and non-native urban trees in the greater Los Angeles area. Urban ecosystems 13(4): 393-414

4 Fritts (1976) Tree rings and climate. The Blackburn press p567

5 Hughes (2002) Dendrochronology in climatology - the state of the art. Dendrochronologia 20(1-2):95-116

6 Gricar et al. (2013) Anatomical characteristics and hydrologic signals in tree-rings oaks (Quercus robur L.). Trees 27: 1669-1680

7 Tulik (2010) Diminished vessel diameter as a possible factor in the decline of European ash (Fraxinus excelsior L.). Annals of Forest Science 67: 103