Development of predictive indexes for evaluating UHI adaptation potential of green roof- and wall- based scenarios in the Mediterranean climate

Development of predictive indexes for evaluating UHI adaptation potential of green roof- and wall- based scenarios in the Mediterranean climate | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Development of predictive indexes for evaluating UHI adaptation potential of green roof- and wall- based scenarios in the Mediterranean climate Tiziana Susca, Jacopo Iaria, Fabio Zanghirella This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4259407/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 17 Sep, 2024 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Urban heat island (UHI) can jeopardize urban inhabitants, but the installation of green roofs (GRs) and walls (GWs) can contribute to mitigating it. The present study provides a novel index to easily predict the spatially median variation in air temperature at pedestrian height related to the application of GR- and GW- -based scenarios on the hottest hours of a typical summer day varying building height (BH), coverage percentage (COP), and leaf area index (LAI). The index is meant to be applied to built areas with 0.3–0.4 urban density in the Mediterranean climate and is derived from a linear regression model fed with the outputs of 269 simulations of three urban areas developed and run in ENVI-met software. The developed models are all highly significant. GR model shows that the mitigation is influenced by all three parameters, and it can estimate the mitigation with a mean standard error of 0.05°C. GW model shows that BH is not influential in decreasing air temperature compared to the other parameters. GF and living wall (LW) index can predict the mitigation with an error of 0.03°C and 0.04°C, respectively. However, for the LW model, further parameters should be considered to improve its reliability. Earth and environmental sciences/Climate sciences Earth and environmental sciences/Environmental sciences Physical sciences/Engineering Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Urban Heat Island (UHI), an increase in urban temperature compared to the rural counterparts 1 , is a serious issue globally 2 that can jeopardize urban inhabitants’ health 3 , increase building cooling energy demand 4,5 , and contribute to air quality deterioration 6 . It is therefore vital to adapt cities to UHI. Greenery offers an opportunity to reduce urban overheating 7,8 and although residual urban spaces are difficult to find in densely urbanized cities, building façades and rooftops can easily accommodate vegetation. Building Integrated Vegetation Technologies (BIVTs), namely green roofs and walls, can concomitantly reduce UHI and energy consumption for building summer cooling 9–11 . Green roofs are characterized by a horizontal growing medium layer. Depending on the thickness of the substrate green roofs are clustered into extensive green roofs (EGRs) when the thickness of the growing medium is 15 cm 13 EGRs because of their light weight can be extensively installed on existing buildings without structural intervention 14 . On the contrary, IGRs, because of their elevated load, can be installed just on buildings specifically designed to support their weight 15 . Green walls can be distinguished according to the position of the substrate: green façades (GFs) are characterized by a horizontal substrate while green walls (GWs) have a vertical substrate adherent to the wall 16 . Because of the beneficial effects of the deployment of BIVTs, many city councils —for example, London 17 , Sydney, and Brisbane 18 — incentivize their installation. However, due to the economic investment related to the application of BIVTs-based UHI adaptation plans 19 , their potential effectiveness should be assessed beforehand. Tools to forecast the UHI mitigation potential of BIVT plans do exist (e.g., 20 ), but they require specific knowledge, computational facilities, and time 21 , which are often missing. The use of simpler tools, such as indexes, to predict the effectiveness of the application of BIVTs might incentivize the development, application, and early evaluation of UHI adaptation plans. To the authors’ best knowledge, little research has been carried out on the development of such predictive indexes. For instance, using artificial neural networks, a study (i.e., 22 ) forecasts the surface UHI mitigation potential of green roofs limitedly to the rooftop surface rather than at the pedestrian level. Albeit of great interest, the model outputs can be hardly used by urban planners to verify beforehand the effectiveness of adaptation plans. Another study (i.e., 23 ) provides an estimation of air temperature variation due to the installation of green roofs, by linking evapotranspiration to mitigation. However, to include the effect easier to apply predictors such as urban geometry and percentage of coverage, high computational efforts are necessary hindering its wide use. The present study aims to fill this scientific gap by providing an equation capable to easily predict the variation in spatially median temperature in the hottest day hour of a typical summer day, at pedestrian level (i.e., 1.5 m above the ground) related to the application of adaptation plans based on the installation of a single BIVT. Specifically, we considered EGRs, GFs, and LWs, while IGRs were omitted because of their limited installation potential in built areas. The equations are meant to be applied to urban areas with urban density, defined as the ratio between buildings' basal area and the total surface of the considered zone, ranging from 0.3 to 0.4. Urban density has also been addressed as local climate zone (LCZ) expanding the concept from just surfaces to regions of uniform surface cover, material, human activity, and phenomenon 24 . Furthermore, since background climate is influential in green roofs and walls performances 10,11 , the index was calibrated on three study areas in three Italian cities in the Mediterranean climate zone since this latter is witnessing the highest intensification of heat stress 25 . Specifically, the selected cities belong to the Csa climate zone according to the Köppen Geiger classification 26 . The developed parametric equations calculate the variation of air temperature at pedestrian level (i.e., 1.5 m) varying the most influencing parameters for BIVTs performance in mitigating air temperature 27 : building height (BH), coverage percentage (COP) (i.e., the percentage of the surface occupied by BIVTs) and Leaf Area Index (LAI) of the installed plants. The equations were obtained through a linear regression model populated by the median air temperature variation at 1.5 m of 238 mitigation scenarios from 31 control ones. All the scenarios were developed and simulated in ENVI-met 20 . Results Air temperature variation The 31 control scenarios were developed considering three study areas in three Italian cities — Rome, Florence, and Bari — and varying, for each built area, the BH. Based on the control scenarios, the 238 adaptation scenarios were developed and simulated implementing a single BIVT each time and varying COP and LAI singularly or concomitantly. Figure 1 shows the median difference in air temperature between the mitigation scenario and the control scenario for each investigated BIVT. 5% of the EGR-based scenarios show a median mitigation effect greater than − 0.2°C, 39% of the scenarios mitigate by -0.2 – -0.1°C, 36% by -0.1– 0°C, and 20% have a positive variation in temperature. The maximum mitigation equals − 0.27°C and is achieved by a scenario in Bari with LAI 5, COP equal 100%, and BH of 5 m; the highest increase in temperature, + 0.08°C, is found in Rome in a scenario with 1.5 LAI, 25% COP and 20 m BH. Similarly, for GFs, only 4% of the scenarios achieve median mitigation equal to or greater than − 0.2°C, and 39% of the cases show a mitigation between − 0.2 and − 0.1°C and − 0.1 and 0°C, respectively; besides 22% of the temperature variation induced by the GF installation is positive. GFs’ maximum mitigation equals − 0.20°C and is achieved in Florence when a scenario characterized by LAI 5 covering 100% of the 40 m high buildings is applied. The highest increase in temperature equal to + 0.09°C is found in Bari in a scenario with GF of LAI 1.5 covering 25% of the 20 m high buildings. Eventually, 4% of the LW-based scenarios mitigate more than − 0.2°C, 50%, and 39% have a mitigation ranging from − 0.2 to -0.1°C, and from − 0.1 to 0°C, respectively. Only 6% of LW scenarios produce a positive temperature variation. The highest decrease and increase in temperature are − 0.21 and + 0.07°C, respectively. The former is achieved in Rome in a scenario with LAI 5, COP 100%, and BH 20 m and the latter is achieved in Bari with LAI 1.5, COP 25%, and BH 20 m. Indexes We used linear regression to calculate the relations between the median air temperature variation and BH, COP, and LAI to obtain three novel indexes. Such indexes are meant to predict the spatially median temperature variation at the pedestrian level when applying a specific BIVT varying, according to the user needs, the COP expressed as integer (i.e., for a 20% COP, 20 must be inserted), the mean BH in the built area expressed in meters, and LAI varying from a minimum value of 1.5 to a maximum one of 5. For the EGR, the equation is: $${I}_{EGR}=-0.0784 -0.0016COP +0.0722\text{ln}\left(BH\right) -0.0177LAI$$ 1 Table 1 shows that the EGR model is significant and explains 0.61 of the data variances. Also, all the predictors are significant in the model and their relationship with the temperature variation is depicted in Fig. 2 . All the models have been tested for normality of residuals using a Shapiro-Wilkins test and for homoscedasticity using a Breusch-Pagan test as well as checking for the influence of single observations on the regression using Cook’s distance. For EGRs, it resulted that the residuals are normal (p-val = 0.06) and homoscedastic (p-val = 0.22) (Table 1 ). One observation was removed as it showed Cook’s distance a magnitude greater than all the others (i.e., EGR with LAI 1.5 COP 25 BH 10, in Florence). As a measure of errors on the indexes, we computed the root mean square error (RMSE) and mean absolute error (MAE) between predicted and observed results and obtained 0.049°C and 0.041°C for the EGR index. Table 1 Linear regression EGR model outputs Extensive green roof Median T reduction = a + b*COP + c*ln(BH) + d*LAI Model coefficients N. obs Adj. R 2 F-value p-value 82 0.61 43.45 10 − 16 Variables coefficients Estimate St. Err t-value p-value a -0.0784 0.0353 -2.22 0.03 b -0.0016 0.0002 -8.11 10 − 12 c 0.0722 0.0095 7.58 10 − 11 d -0.0177 0.0038 -4.63 10 − 6 For the GFs, we derived the following index: $${I}_{GF}= 0.1757 -0.0014COP -0.1164\text{ln}\left(LAI\right)$$ 2 Table 2 reports the GF model coefficients’ estimates, standard error t, and p-value. The model neglects BH since it has been found non-significant. GF’s model explains 70% of the variance and has a p-value equal to 10 − 16 . The residuals are normal (p-val = 0.11) and homoscedastic (p-val = 0.29). One observation (i.e., GF LAI 1.5 COP 25 BH 10 in Florence) was removed since it was an outlier. Figure 3 depicts the relation between the single predictor and the median air temperature variation in the scenarios. It shows that predictions occupy a smaller interval than the observed ones and when LAI equals 3 predicted values coincide. The estimated RMSE and MAE for the index equal 0.041°C and 0.033°C, respectively. Table 2 Linear regression GF model outputs Green façade Median T reduction = a + b*COP + c*ln(LAI) Model coefficients N. obs Adj. R 2 F-value p-value 76 0.70 90.18 10 − 16 Variables coefficients Estimate St. Err t-value p-value a 0.1757 0.0204 8.82 10 − 13 b -0.0014 0.0001 -8.17 10 − 12 c -0.1164 0.0095 -12.18 10 − 16 For LWs we obtained the following index: $${I}_{LW }= 0.096 -0.00157COP -0.0201LAI$$ 3 As for GFs, also for LWs BH was non-significant as a predictor and was removed to improve the model’s residuals. The model is significant and has an adjusted R squared of 0.54, the residuals are homoscedastic (p-val = 0.3) but non-normal (p-val = 0.01) (Table 3 ). One observation (i.e., LAI 1.5 COP 25 BH 10 in Florence) presented a Cook’s distance of magnitude greater than the other ones and was removed. The predicted values for LWs occupy a smaller interval of values than the observed ones and tend to cluster (Fig. 4 ); RMSE and MAE for the index are equal to 0.043°C and 0.036°C, respectively. Table 3 Linear regression LW model outputs Living wall Median T reduction = a + b*𝐶𝑂𝑃 + c*L𝐴𝐼 Model coefficients N. obs Adj. R 2 F-value p-value 77 0.54 46.05 10 − 13 Variables coefficients Estimate St. Err t-value p-value a 0.096 0.0215 4.47 10 − 5 b -0.00157 0.0002 -8.76 10 − 13 c -0.0201 0.0033 -5.93 10 − 8 Discussion This study provides a tool to ease the beforehand assessment of UHI adaptation plans based on the application of single BIVT scenarios. Specifically, the authors developed three equations capable of providing median UHI mitigation values in an urban area by applying EGRs, GFs, or LWs varying COP, BH, and LAI. Compared to previous research the main aim of the present study is to provide an easy-to-use tool that can be widely used by city council technicians and urban planners as it requires a limited amount of input data and computational time. To the authors’ best knowledge, just a few studies focused on the interaction between BIVTs and pedestrian-level air temperature do exist and are based on complex physical models. To provide some examples, Alexandri et al. 28 developed an ad-hoc dynamic micro-scale model that analyzes the effect of green roofs and green walls on the air temperature in urban canyons and at the rooftop level, Djeddjig et al. 29 developed a TRNSYS hygrothermal model of green walls and a model of mass flows in street canyons. Both these studies provide an estimation of air temperature variation within an urban canyon rather than in a selected urban area. Therefore, they cannot be easily and widely used for UHI mitigation forecasting since they provide punctual air temperature values and need many parameters and input variables. Furthermore, Yang and Wang 30 analyzed the application of EGRs on the urban canopy by applying Monte Carlo simulations, Mazzeo et al. 22 investigated the effect of EGRs on the temperature at the rooftop level using artificial neural networks, and Suter et al. 23 developed an equation for calculating the air temperature variation of the surface layer (from roofs height to 280 m) depending on green roofs. Although such studies provide useful information as roof surface temperature 22,30 and domain averaged evaporation rate 23 , such data can hardly be used by urban planners to evaluate the UHI mitigation potential at the pedestrian level of adaptation plans. The model developed in this study is based on microclimate simulations conducted with ENVI-met, a tool based on the laws of thermodynamics and fluid dynamics, which can simulate the interactions between air, surfaces, and vegetation in an urban context 20 . Therefore, the output reliability depends on the accuracy of both the ENVI-met software and the regression model. Tsoka et al. 31 found that ENVI-met can accurately predict air temperature values under different meteorological conditions. Besides, the regression model has been found to be significant and explains 0.61 and 0.70 of the data variances for EGRs and GFs, respectively, and for both the residuals are normal and homoscedastic. On the other hand, the LW model has been found significant, and its residuals are homoscedastic, but its adjusted R squared equals 0.54 and residuals are non-normal. Nevertheless, by including the effect of the different urban arrangements as intercepts, the accuracy of the LW model greatly improves (i.e., adjusted R squared = 0.66) and its residuals are normal (i.e., p-val = 0.12). Therefore, as far as the LWs are concerned, the reliability of the model greatly depends on the urban arrangement. However, adding the urban arrangement effect might hinder the genericity of the current model. The developed EGR regression model shows a direct correlation between both LAI and COP values and median air temperature mitigation. Our results are in accordance with Suter et al. 23 regarding coverage. As regards LAI, our findings are in agreement with Jamel et al. 32 , even though they observed that after a certain LAI threshold, the temperature reduction reaches a plateau. Furthermore, we found that increasing COP is more effective in mitigation than increasing LAI. Such a finding is also confirmed by Iaria and Susca 27 . Contrarily, Fig. 2 , also shows an indirect correlation between BH and air temperature mitigation, meaning that the effectiveness of EGRs in mitigating air temperature at the pedestrian level decreases when the height of the rooftops where EGRs are installed increases. Such a result is in accordance with previous literature (e.g. 33 ) and is explained with a greater distance of the mitigation source from the target. Furthermore, as in 27 , we found that BH equal to approximately 40 m can be considered as a threshold above which the application of EGRs has a negligible effect on air temperature reduction at the pedestrian level 34,35 . As far as the GF and the LW models are concerned, BH was found to be a negligible input parameter, in comparison with LAI and COP and was therefore removed from the models. This finding might indicate that proximity to the target is as important –or even more– as the greened area increases in mitigating air at the desired height. Indeed, the mitigation effect of GFs and LWs at the pedestrian level is greater when the source of evapotranspiration is closer to the target (i.e., street level). This result should be further investigated in future studies. In Fig. 3 and Fig. 4 both the GF and the LW models, respectively, predict a direct relationship between both LAI and COP and the mitigation in air temperature. When applying GFs, low values of LAI (e.g., LAI = 1.5) lead to a negligible median reduction or even an increase in temperature, while high values (e.g., LAI = 5) lead to a median reduction of about 0.1°C. Moreover, when GFs are deployed, an increase in LAI is more effective than an increase in COP in mitigating air temperature. Contrarily, in the application of LWs, the effect of an increase in COP on temperature mitigation (ranging from about − 0.04°C with COP = 25% to about − 0.12°C with COP = 100%) is significantly greater than an increase in LAI (ranging from about − 0.07°C with LAI = 1.5 to about − 0.09°C with COP = 100%). As far as limitations and shortcomings are concerned, one limit of the proposed model resides in its applicability to the Mediterranean climate only, specifically, the Csa climate zone. Nonetheless, the same methodology might be applied to other climate zones. Furthermore, climatic variables such as wind direction and speed, that have an impact on mitigation 35 , were kept constant in the study areas by using typical summer days as forcing climatic conditions. Future research might improve the index with changing climatic variables to get insights about the mitigation performance in changing conditions. Also, the model assumes that green roofs and walls are fully irrigated at any time. BIVTs’ performance depends on irrigation and in particular water availability for plants 11,36 , therefore, in real conditions water availability is a crucial factor. Lastly, the vegetation layer of the BIVTs is composed of standard, well-investigated plants such as Sedum sediforme, Nephrolepis exalata , and Hedera helix for EGR, GF, and LW, respectively. We neglected the use of different local species as it was outside the scope of the study, but as transpiration is dependent on plant traits, differences might occur when other species are implemented in BIVTs. By applying a single BIVT most of the mitigation values range from 0 to − 0.2°C. Such values seem to be low, nevertheless, mitigation is not evenly distributed within the areas: The effect of BIVTs tends to be distributed near buildings or in the upwind direction. Specifically, in 25% of the urban areas the temperature mitigation is significantly higher than the median one (Fig. 5 ), and punctual mitigation values can be up to − 0.83°C. Nevertheless, the proposed model might also indicate a potential ineffectiveness of BIVT-based adaptation plans. Such results are equally important since might entail the implementation of more synergic solution sets 19 . Conclusions The present study reports on the development of an index to predict the mitigation potential of single BIVT-based scenarios — specifically, EGRs, GFs, and LWs — to eradicate UHI in urban areas in the Csa climate zone characterized by an urban density ranging from 0.3 to 0.4. The developed indexes can be applied to built areas by varying the parameters that mostly influence the UHI mitigation potential of BIVT-based scenarios: BH, COP, and LAI. The developed indexes are useful since they can ease the work of planners and city council technicians who want to assess beforehand the effectiveness of BIVT-based UHI adaptation plans. The developed indexes can forecast and compare the effect of adaptation BIVT-based plans using easily available data overcoming the limits of other methods (i.e., 20 ), and their outputs remain within an acceptable range of error. Furthermore, the indexes are novel, since no studies have been found in published literature providing such a tool and, differently from other research that focuses on providing punctual temperature variation values, the current study provides spatially median variation in urban temperature which can more reliably support urban planners in designing UHI adaptation plans. The novelty of the proposed indexes also resides in the use of a methodology based both on the development of 269 ENVI-met models and simulations and on the linear regression models. The proposed methodology can be used for developing indexes for other climate zones. The precision of the model greatly relies on that of the ENVI-met tool, therefore, in the future, when the ENVI-met model will be further improved, the developed indexes, fed with other simulations, might decrease their error. Yet, limitations exist. For instance, the model might be further enriched with other parameters, such as plant watering data or the models can be developed on single urban arrangements. However, although the precision of the model would probably benefit from such implementations, the implementation of more parameters would likely increase the complexity of the indexes and the focus on single urban arrangements would make the indexes less applicable, altogether, such improvements might hinder urban planners from their use. Altogether the developed indexes can greatly contribute to ease the work of planners to eradicate urban overheating and the proposed methodology can pave the way for the development of other indexes capable of assessing, with different levels of error, the effectiveness of UHI adaptation plans. Methods For the development of the indexes, three urban areas prone to the UHI formation in three Italian cities belonging to the Csa climate zone were selected. The study areas of Via Lanciani in Rome, Viale Kennedy in Bari, and the neighborhood of Gavinana in Florence are characterized by similar extension, LCZ, and urban density, but different layouts were chosen (Fig. 6, 1. Identification of the study areas). Indeed, the three urban areas have been chosen to represent the main typical regional layouts 33,35 to generalize the results as much as possible. Specifically, via Lanciani urban area presents a scattered layout with similar-sized squared buildings and roads arranged in N-S and E-W directions; Bari shows an array layout with buildings displayed along parallel streets, while Florence buildings are characterized by an enclosing layout with buildings forming closed polygons with internal courtyards. Another requisite for the selection of the areas was the access to monitored historical meteorological data recorded by a meteorological station positioned within the selected built area and another one positioned in the rural environs. These latter data were used to force the ENVI-met models (Fig. 6). The used methodology has been thoroughly described in a previous study 27 , however, the present study has been enriched with 152 new scenarios Tab.S1. All the models were run in ENVI-met for the typical summer day for each city. To identify the typical summer days, we compared the average hourly temperature profile of the hottest month (data from 37 ) with the real-day temperature profile recorded in the urban stations, after removing the rainy days and selecting the one with the smallest RMSE and MAE. 1. UHI-mitigation criteria and values selection Three parameters are predicted to be influential in UHI mitigation by BIVTs: BH 10,11,35,38 , COP 39 , and LAI 38 . For the development of mitigation and control scenarios, we assumed the following BH: 5, 10, 20, 30, and 40 m. The incremental height has been chosen as multiple of the grid used for the development of the scenarios (see section 3. Mitigation scenarios). The maximum has been set at 40 m since previous research has shown that above this threshold green roofs have null mitigation potential 32 . As far as COP is concerned, we considered 25, 50, 75, and 100%, besides, LAI values were set to 1.5, 3, and 5 standing for low, medium, and high foliage density, respectively. All the mitigation scenarios were created by applying a single BIVT to the study area with combinations of the possible values of the three criteria. 2. Study areas modeling The selected areas were modeled in ENVI-met. The software requires the 3D arrangement of all urban elements (e.g., buildings, trees, shrubs, paved and natural surfaces, BIVTs) as well as their inner structure and physical properties to simulate the interaction between them and produce microclimatic outputs such as the potential air temperature. Additional information necessary to improve the model accuracy is meteorological data to input as forcing conditions. Data from the urban station were used to test the ENVI-met accuracy while data from the rural stations were used as forcing to the baseline and mitigation scenarios (See methodology in 27 ). 3. Mitigation scenarios In addition to the 117 original scenarios developed by 27 , created by varying the value of one parameter per time while keeping the other constant at the maximum value, we added 139 additional intermediate scenarios –scenarios with the parameters changing together at low values– to account for BIVTs efficacy in sub-optimal conditions and improve the results. All scenarios are shown in tab.S.1 supplementary materials. The intermediate scenarios are representative of the sub-optimal condition of LAI (i.e., LAI 1.5) and coverage percentage (i.e., COP 25%, 50%, and 75%) for the different classes of building height in the three study areas. 4. ENVI-met simulations Every scenario was simulated using the same grid parameters. Specifically, we choose a grid of 5 m (x) x 5 m (y) x 5 m (z) to get accurate outputs in a relatively short computational time ranging from six to eight hours. For each simulation two days were simulated, the first one for stabilizing the model and the second one to get the microclimate outputs. All the simulations were run with an AMD Ryzen 53,600 6-Core processor and 32.0 Gb RAM. The original scenarios were simulated using ENVI-met 4.4.6 while for the additional scenarios, ENVI-met 5.1 was used. The difference in the ENVI-met version required the baseline scenarios to be simulated in both 4.4.6 and 5.1 to nullify any difference between the different releases. 5. Data extraction and analysis Using ENVI-met LEONARDO, we extracted in CSV format the potential air temperature data at pedestrian level related to the hottest hour in the three study areas, namely 1.00 pm for Rome and 2.00 pm for Bari and Florence of all the simulated scenarios. By means of 40 we calculated for each cell of the modeled areas the difference between the mitigation scenario and the control one (e.g., mitigation scenarios in Rome with BH equal to five were subtracted from baseline scenario of Rome with BH equal to five). Subsequently, we calculated the mean, median, 1st and 3rd quartile air temperature difference of each mitigation scenario. Then, we used linear regression to calculate the mitigation index of each BIVT (function lm(), baseline R). We tested heteroskedasticity with a Breusch-Pagan test, package lmtest 41 function bptest(), and normality with function shapiro.test() in baseline R; Cook’s distance was considered by plotting the model’s residuals. Lastly, we used functions rmse() and mae() of 42 to assess the RMSE and MAE of the equations in predicting mitigation. For graphs packages, ggplot2 and gghalves 43 were used. Declarations Acknowledgments This work was supported by MASE (Italian Ministry for the Environment and Energy Security) [Fondo per il finanziamento delle attività di ricerca e di sviluppo di interesse generale per il sistema elettrico nazionale]. Author contributions Conceptualization, TS and JI; methodology, TS and JI; formal analysis JI; investigation TS, JI, and FZ; writing-original draft preparation TS, JI, and FZ; writing-review and editing TS, JI and FZ; visualization JI; supervision TS. All authors have read and agreed to the published version of the manuscript. Competing interests The authors declare no competing interests. Data availability The datasets necessary for preparing ENVI-met simulations are available in 27 supplementary materials. All ENVI-met-related files, extracted temperature layers from simulation results, R Scripts and data generated for computing models are available on request to Jacopo Iaria, email: [email protected] . References Landsberg, H. E. The Urban Climate . (Academic Press, 1981). Zhang, P., Imhoff, M. L., Wolfe, R. E. & Bounoua, L. Characterizing urban heat islands of global settlements using MODIS and nighttime lights products. Canadian Journal of Remote Sensing 36, 185–196 (2010). Huang, W. T. K. et al. Economic valuation of temperature-related mortality attributed to urban heat islands in European cities. Nat Commun 14, 7438 (2023). Santamouris, M., Cartalis, C., Synnefa, A. & Kolokotsa, D. On the impact of urban heat island and global warming on the power demand and electricity consumption of buildings—A review. Energy and Buildings 98, 119–124 (2015). Santamouris, M. On the energy impact of urban heat island and global warming on buildings. Energy and Buildings 82, 100–113 (2014). Li, H. et al. Interaction between urban heat island and urban pollution island during summer in Berlin. Science of The Total Environment 636, 818–828 (2018). Wong, N. H., Tan, C. L., Kolokotsa, D. D. & Takebayashi, H. Greenery as a mitigation and adaptation strategy to urban heat. Nat Rev Earth Environ 2, 166–181 (2021). Edmondson, J. L., Stott, I., Davies, Z. G., Gaston, K. J. & Leake, J. R. Soil surface temperatures reveal moderation of the urban heat island effect by trees and shrubs. Sci Rep 6, 33708 (2016). Susca, T., Zanghirella, F. & Del Fatto, V. Building integrated vegetation effect on micro-climate conditions for urban heat island adaptation. Lesson learned from Turin and Rome case studies. Energy and Buildings 295, 113233 (2023). Susca, T., Zanghirella, F., Colasuonno, L. & Del Fatto, V. Effect of green wall installation on urban heat island and building energy use: A climate-informed systematic literature review. Renewable and Sustainable Energy Reviews 159, 112100 (2022). Susca, T. Green roofs to reduce building energy use? A review on key structural factors of green roofs and their effects on urban climate. Building and Environment 162, 106273 (2019). Dunnett, N. & Kingsbury, N. Planting Green Roofs and Living Walls . vol. VII (Timber Press, Portland, 2004). Environmental Protection Agency. Reducing Urban Heat Islands: Compendium of Strategies. Green Roofs. (2008). Castleton, H. F., Stovin, V., Beck, S. B. M. & Davison, J. B. Green roofs; building energy savings and the potential for retrofit. Energy and Buildings 42, 1582–1591 (2010). Environmental Protection Agency. Reducing Urban Heat Islands: Compendium of Strategies. Green Roofs. (2008). Manso, M. & Castro-Gomes, J. Green wall systems: A review of their characteristics. Renewable and Sustainable Energy Reviews 41, 863–871 (2015). London City Council. London Plan 2011. London City Hall https://www.london.gov.uk//what-we-do/planning/london-plan/past-versions-and-alterations-london-plan/london-plan-2011 (2011). Irga, P. J. et al. The distribution of green walls and green roofs throughout Australia: Do policy instruments influence the frequency of projects? Urban Forestry & Urban Greening 24, 164–174 (2017). Zhao, Y. et al. Beating urban heat: Multimeasure-centric solution sets and a complementary framework for decision-making. Renewable and Sustainable Energy Reviews 186, 113668 (2023). High-Resolution 3D Modeling of Urban Microclimate with ENVI-met Software. ENVI-met https://www.envi-met.com/ . Balany, F., Ng, A. W., Muttil, N., Muthukumaran, S. & Wong, M. S. Green Infrastructure as an Urban Heat Island Mitigation Strategy—A Review. Water 12, 3577 (2020). Mazzeo, D., Matera, N., Peri, G. & Scaccianoce, G. Forecasting green roofs’ potential in improving building thermal performance and mitigating urban heat island in the Mediterranean area: An artificial intelligence-based approach. Applied Thermal Engineering 222, 119879 (2023). Suter, I., Maksimović, Č. & van Reeuwijk, M. A neighbourhood-scale estimate for the cooling potential of green roofs. Urban Climate 20, 33–45 (2017). Stewart, I. D. & Oke, T. R. Local Climate Zones for Urban Temperature Studies. Bulletin of the American Meteorological Society 93, 1879–1900 (2012). Diffenbaugh, N. S., Pal, J. S., Giorgi, F. & Gao, X. Heat stress intensification in the Mediterranean climate change hotspot. Geophysical Research Letters 34, (2007). Kottek, M., Grieser, J., Beck, C., Rudolf, B. & Rubel, F. World Map of the Köppen-Geiger climate classification updated. metz 15, 259–263 (2006). Iaria, J. & Susca, T. Analytic Hierarchy Processes (AHP) evaluation of green roof- and green wall- based UHI mitigation strategies via ENVI-met simulations. Urban Climate 46, 101293 (2022). Alexandri, E. & Jones, P. Temperature decreases in an urban canyon due to green walls and green roofs in diverse climates. Building and Environment 43, 480–493 (2008). Djedjig, R., Bozonnet, E. & Belarbi, R. Modeling green wall interactions with street canyons for building energy simulation in urban context. Urban Climate 16, 75–85 (2016). Yang, Jiachuan & Wang, Zhi-Hua. Physical parameterization and sensitivity of urban hydrological models: Application to green roof systems. Building and Environment 75, 250–263 (2014). Tsoka, S., Tsikaloudaki, A. & Theodosiou, T. Analyzing the ENVI-met microclimate model’s performance and assessing cool materials and urban vegetation applications–A review. Sustainable Cities and Society 43, 55–76 (2018). Jamei, E. et al. Investigating the cooling effect of a green roof in Melbourne. Building and Environment 246, 110965 (2023). Ng, E., Chen, L., Wang, Y. & Yuan, C. A study on the cooling effects of greening in a high-density city: An experience from Hong Kong. Building and Environment 47, 256–271 (2012). Chen, H., Ooka, R., Huang, H. & Tsuchiya, T. Study on mitigation measures for outdoor thermal environment on present urban blocks in Tokyo using coupled simulation. Building and Environment 44, 2290–2299 (2009). Jin, C., Bai, X., Luo, T. & Zou, M. Effects of green roofs’ variations on the regional thermal environment using measurements and simulations in Chongqing, China. Urban Forestry & Urban Greening 29, 223–237 (2018). Vaezizadeh, F., Rashidisharifabad, S. & Afhami, R. Investigating the Cooling Effect of Living Walls in the Sunken Courtyards of Traditional Houses in Yazd. European Journal of Sustainable Development 5, 27–27 (2016). CTI Comitato Termotecnico Italiano -. https://www.cti2000.it/ . Morakinyo, T. E., Lai, A., Lau, K. K.-L. & Ng, E. Thermal benefits of vertical greening in a high-density city: Case study of Hong Kong. Urban Forestry & Urban Greening 37, 42–55 (2019). Morakinyo, T. E., Dahanayake, K. W. D. Kalani. C., Ng, E. & Chow, C. L. Temperature and cooling demand reduction by green-roof types in different climates and urban densities: A co-simulation parametric study. Energy and Buildings 145, 226–237 (2017). R: The R Project for Statistical Computing. https://www.r-project.org/ . Zeileis, A. & Hothorn, T. Diagnostic Checking in Regression Relationships. R News vol. 2 (2002). mfrasco. /Metrics . (2024). Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial.docx Cite Share Download PDF Status: Published Journal Publication published 17 Sep, 2024 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 10 Jun, 2024 Reviews received at journal 07 Jun, 2024 Reviewers agreed at journal 21 May, 2024 Reviews received at journal 10 May, 2024 Reviewers agreed at journal 07 May, 2024 Reviewers invited by journal 07 May, 2024 Editor assigned by journal 03 May, 2024 Editor invited by journal 03 May, 2024 Submission checks completed at journal 03 May, 2024 First submitted to journal 12 Apr, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4259407","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":300054543,"identity":"bd95aec9-b180-41a5-a130-cdab468d9d56","order_by":0,"name":"Tiziana Susca","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5UlEQVRIiWNgGAWjYDACZgaGAwheBYQ6gE0lDPCgajkD1YJPDw8Kj7GNgbA19uzsDw/83GHHID8j/eGHn/Ns8vjZDzAe/oDfYQkHe88kMxjcyDGW7N2WVizZk0DAYcxAed42ZgYDiRw2Bt5thxM3HCCohbHh4N+2epDDnjH+nfM/cf/5B4S0MDMc5m07zMBwI8GMmbfhQOIGCUK2HGZjOCzbdpzH4MwbY2mZY8mJM248bDhwBo8W9v7jjz++bauWk29Pf/jxTY1dYn9/8uEPFXi0wG1DYjM2EKFhFIyCUTAKRgE+AADlyFE4hRCJvQAAAABJRU5ErkJggg==","orcid":"","institution":"National Agency for New Technologies, Energy and Sustainable Economic Development","correspondingAuthor":true,"prefix":"","firstName":"Tiziana","middleName":"","lastName":"Susca","suffix":""},{"id":300054544,"identity":"4c8ec01f-535f-43ad-b701-f96e9b7862ae","order_by":1,"name":"Jacopo Iaria","email":"","orcid":"","institution":"University of Bologna","correspondingAuthor":false,"prefix":"","firstName":"Jacopo","middleName":"","lastName":"Iaria","suffix":""},{"id":300054545,"identity":"401fc8de-8c71-4579-a5e5-3453db78b733","order_by":2,"name":"Fabio Zanghirella","email":"","orcid":"","institution":"National Agency for New Technologies, Energy and Sustainable Economic Development","correspondingAuthor":false,"prefix":"","firstName":"Fabio","middleName":"","lastName":"Zanghirella","suffix":""}],"badges":[],"createdAt":"2024-04-12 18:59:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4259407/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4259407/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-024-67567-9","type":"published","date":"2024-09-17T15:57:20+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":56142834,"identity":"069f38bc-1ab4-485d-ba06-e2622b1cf69c","added_by":"auto","created_at":"2024-05-09 05:02:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":91777,"visible":true,"origin":"","legend":"\u003cp\u003eMedian variation in air temperature of the mitigation scenarios grouped by BIVT.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-4259407/v1/0b812286822d9739721be035.png"},{"id":56142874,"identity":"b9bd4406-1096-4ecd-82dc-f609c9fb59cc","added_by":"auto","created_at":"2024-05-09 05:02:16","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":187037,"visible":true,"origin":"","legend":"\u003cp\u003eRight panel: comparison between observed and predicted values for the EGR model. Left panel: relation median temperature variation and the predictors with trend lines for the EGR model.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-4259407/v1/e8fb6926d209a4243f97efd8.png"},{"id":56142794,"identity":"d5efa27d-49d8-4658-bd8f-25a2ae65f280","added_by":"auto","created_at":"2024-05-09 05:01:42","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":119581,"visible":true,"origin":"","legend":"\u003cp\u003eRight panel: comparison between observed and predicted values for the GF model. Left panel: relation median temperature variation and the predictors with trend lines for the GF model.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-4259407/v1/7e9142e4944a61f7ce820e4d.png"},{"id":56142674,"identity":"10feee97-42d5-40b1-b00d-79df7ff45763","added_by":"auto","created_at":"2024-05-09 05:01:01","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":148037,"visible":true,"origin":"","legend":"\u003cp\u003eRight panel: comparison between observed and predicted values for the LW model. Left panel: relation median temperature variation and the predictors with trend lines for the LW model.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-4259407/v1/d703ab22739a1ad0cb06993c.png"},{"id":56142625,"identity":"91255b3c-043a-4f2e-b17c-77ace78247a8","added_by":"auto","created_at":"2024-05-09 05:00:50","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":49948,"visible":true,"origin":"","legend":"\u003cp\u003eBoxplot of the 25\u003csup\u003eth\u003c/sup\u003e, 50\u003csup\u003eth\u003c/sup\u003e, and 75\u003csup\u003eth\u003c/sup\u003e percentile of the spatial observed temperature mitigation values, and the 50\u003csup\u003eth\u003c/sup\u003e percentile of the spatial predicted temperature mitigation values\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-4259407/v1/e3f331e91c31e865ccb4566f.png"},{"id":56142640,"identity":"529cd0d5-3960-400c-9989-80a364a76447","added_by":"auto","created_at":"2024-05-09 05:00:54","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":230057,"visible":true,"origin":"","legend":"\u003cp\u003eMethodology workflow.\u003c/p\u003e","description":"","filename":"image6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4259407/v1/3aad13b12ea5567029ec964f.jpg"},{"id":65103950,"identity":"068ca598-612a-4991-8e45-3d9dbc8f642f","added_by":"auto","created_at":"2024-09-23 16:10:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1306421,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4259407/v1/ad3c640e-3030-4f10-bb63-1625d8dbc921.pdf"},{"id":56143954,"identity":"c5a92197-6556-4911-91a4-521ec79e5f48","added_by":"auto","created_at":"2024-05-09 05:10:07","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":47154,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-4259407/v1/5a8f9112dcb35c4b51afd4a2.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Development of predictive indexes for evaluating UHI adaptation potential of green roof- and wall- based scenarios in the Mediterranean climate","fulltext":[{"header":"Introduction","content":"\u003cp\u003eUrban Heat Island (UHI), an increase in urban temperature compared to the rural counterparts\u003csup\u003e1\u003c/sup\u003e, is a serious issue globally\u003csup\u003e2\u003c/sup\u003e that can jeopardize urban inhabitants\u0026rsquo; health\u003csup\u003e3\u003c/sup\u003e, increase building cooling energy demand\u003csup\u003e4,5\u003c/sup\u003e, and contribute to air quality deterioration\u003csup\u003e6\u003c/sup\u003e. It is therefore vital to adapt cities to UHI. Greenery offers an opportunity to reduce urban overheating\u003csup\u003e7,8\u003c/sup\u003e and although residual urban spaces are difficult to find in densely urbanized cities, building fa\u0026ccedil;ades and rooftops can easily accommodate vegetation. Building Integrated Vegetation Technologies (BIVTs), namely green roofs and walls, can concomitantly reduce UHI and energy consumption for building summer cooling\u003csup\u003e9\u0026ndash;11\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eGreen roofs are characterized by a horizontal growing medium layer. Depending on the thickness of the substrate green roofs are clustered into extensive green roofs (EGRs) when the thickness of the growing medium is \u0026lt;\u0026thinsp;12 cm\u003csup\u003e12\u003c/sup\u003e, and intensive green roofs (IGRs) when it is \u0026gt;\u0026thinsp;15 cm\u003csup\u003e13\u003c/sup\u003e EGRs because of their light weight can be extensively installed on existing buildings without structural intervention\u003csup\u003e14\u003c/sup\u003e. On the contrary, IGRs, because of their elevated load, can be installed just on buildings specifically designed to support their weight\u003csup\u003e15\u003c/sup\u003e. Green walls can be distinguished according to the position of the substrate: green fa\u0026ccedil;ades (GFs) are characterized by a horizontal substrate while green walls (GWs) have a vertical substrate adherent to the wall\u003csup\u003e16\u003c/sup\u003e. Because of the beneficial effects of the deployment of BIVTs, many city councils \u0026mdash;for example, London\u003csup\u003e17\u003c/sup\u003e, Sydney, and Brisbane\u003csup\u003e18\u003c/sup\u003e \u0026mdash; incentivize their installation. However, due to the economic investment related to the application of BIVTs-based UHI adaptation plans\u003csup\u003e19\u003c/sup\u003e, their potential effectiveness should be assessed beforehand. Tools to forecast the UHI mitigation potential of BIVT plans do exist (e.g.,\u003csup\u003e20\u003c/sup\u003e), but they require specific knowledge, computational facilities, and time\u003csup\u003e21\u003c/sup\u003e, which are often missing.\u003c/p\u003e \u003cp\u003eThe use of simpler tools, such as indexes, to predict the effectiveness of the application of BIVTs might incentivize the development, application, and early evaluation of UHI adaptation plans. To the authors\u0026rsquo; best knowledge, little research has been carried out on the development of such predictive indexes. For instance, using artificial neural networks, a study (i.e.,\u003csup\u003e22\u003c/sup\u003e) forecasts the surface UHI mitigation potential of green roofs limitedly to the rooftop surface rather than at the pedestrian level. Albeit of great interest, the model outputs can be hardly used by urban planners to verify beforehand the effectiveness of adaptation plans. Another study (i.e.,\u003csup\u003e23\u003c/sup\u003e) provides an estimation of air temperature variation due to the installation of green roofs, by linking evapotranspiration to mitigation. However, to include the effect easier to apply predictors such as urban geometry and percentage of coverage, high computational efforts are necessary hindering its wide use.\u003c/p\u003e \u003cp\u003eThe present study aims to fill this scientific gap by providing an equation capable to easily predict the variation in spatially median temperature in the hottest day hour of a typical summer day, at pedestrian level (i.e., 1.5 m above the ground) related to the application of adaptation plans based on the installation of a single BIVT. Specifically, we considered EGRs, GFs, and LWs, while IGRs were omitted because of their limited installation potential in built areas. The equations are meant to be applied to urban areas with urban density, defined as the ratio between buildings' basal area and the total surface of the considered zone, ranging from 0.3 to 0.4. Urban density has also been addressed as local climate zone (LCZ) expanding the concept from just surfaces to regions of uniform surface cover, material, human activity, and phenomenon\u003csup\u003e24\u003c/sup\u003e. Furthermore, since background climate is influential in green roofs and walls performances\u003csup\u003e10,11\u003c/sup\u003e, the index was calibrated on three study areas in three Italian cities in the Mediterranean climate zone since this latter is witnessing the highest intensification of heat stress\u003csup\u003e25\u003c/sup\u003e. Specifically, the selected cities belong to the Csa climate zone according to the K\u0026ouml;ppen Geiger classification\u003csup\u003e26\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe developed parametric equations calculate the variation of air temperature at pedestrian level (i.e., 1.5 m) varying the most influencing parameters for BIVTs performance in mitigating air temperature\u003csup\u003e27\u003c/sup\u003e: building height (BH), coverage percentage (COP) (i.e., the percentage of the surface occupied by BIVTs) and Leaf Area Index (LAI) of the installed plants. The equations were obtained through a linear regression model populated by the median air temperature variation at 1.5 m of 238 mitigation scenarios from 31 control ones. All the scenarios were developed and simulated in ENVI-met\u003csup\u003e20\u003c/sup\u003e.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eAir temperature variation\u003c/h2\u003e \u003cp\u003eThe 31 control scenarios were developed considering three study areas in three Italian cities \u0026mdash; Rome, Florence, and Bari \u0026mdash; and varying, for each built area, the BH. Based on the control scenarios, the 238 adaptation scenarios were developed and simulated implementing a single BIVT each time and varying COP and LAI singularly or concomitantly.\u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows the median difference in air temperature between the mitigation scenario and the control scenario for each investigated BIVT. 5% of the EGR-based scenarios show a median mitigation effect greater than \u0026minus;\u0026thinsp;0.2\u0026deg;C, 39% of the scenarios mitigate by -0.2 \u0026ndash; -0.1\u0026deg;C, 36% by -0.1\u0026ndash; 0\u0026deg;C, and 20% have a positive variation in temperature. The maximum mitigation equals \u0026minus;\u0026thinsp;0.27\u0026deg;C and is achieved by a scenario in Bari with LAI 5, COP equal 100%, and BH of 5 m; the highest increase in temperature, +\u0026thinsp;0.08\u0026deg;C, is found in Rome in a scenario with 1.5 LAI, 25% COP and 20 m BH. Similarly, for GFs, only 4% of the scenarios achieve median mitigation equal to or greater than \u0026minus;\u0026thinsp;0.2\u0026deg;C, and 39% of the cases show a mitigation between \u0026minus;\u0026thinsp;0.2 and \u0026minus;\u0026thinsp;0.1\u0026deg;C and \u0026minus;\u0026thinsp;0.1 and 0\u0026deg;C, respectively; besides 22% of the temperature variation induced by the GF installation is positive. GFs\u0026rsquo; maximum mitigation equals \u0026minus;\u0026thinsp;0.20\u0026deg;C and is achieved in Florence when a scenario characterized by LAI 5 covering 100% of the 40 m high buildings is applied. The highest increase in temperature equal to +\u0026thinsp;0.09\u0026deg;C is found in Bari in a scenario with GF of LAI 1.5 covering 25% of the 20 m high buildings. Eventually, 4% of the LW-based scenarios mitigate more than \u0026minus;\u0026thinsp;0.2\u0026deg;C, 50%, and 39% have a mitigation ranging from \u0026minus;\u0026thinsp;0.2 to -0.1\u0026deg;C, and from \u0026minus;\u0026thinsp;0.1 to 0\u0026deg;C, respectively. Only 6% of LW scenarios produce a positive temperature variation. The highest decrease and increase in temperature are \u0026minus;\u0026thinsp;0.21 and +\u0026thinsp;0.07\u0026deg;C, respectively. The former is achieved in Rome in a scenario with LAI 5, COP 100%, and BH 20 m and the latter is achieved in Bari with LAI 1.5, COP 25%, and BH 20 m.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eIndexes\u003c/h2\u003e \u003cp\u003eWe used linear regression to calculate the relations between the median air temperature variation and BH, COP, and LAI to obtain three novel indexes. Such indexes are meant to predict the spatially median temperature variation at the pedestrian level when applying a specific BIVT varying, according to the user needs, the COP expressed as integer (i.e., for a 20% COP, 20 must be inserted), the mean BH in the built area expressed in meters, and LAI varying from a minimum value of 1.5 to a maximum one of 5.\u003c/p\u003e \u003cp\u003eFor the EGR, the equation is:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$${I}_{EGR}=-0.0784 -0.0016COP +0.0722\\text{ln}\\left(BH\\right) -0.0177LAI$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows that the EGR model is significant and explains 0.61 of the data variances. Also, all the predictors are significant in the model and their relationship with the temperature variation is depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. All the models have been tested for normality of residuals using a Shapiro-Wilkins test and for homoscedasticity using a Breusch-Pagan test as well as checking for the influence of single observations on the regression using Cook\u0026rsquo;s distance. For EGRs, it resulted that the residuals are normal (p-val\u0026thinsp;=\u0026thinsp;0.06) and homoscedastic (p-val\u0026thinsp;=\u0026thinsp;0.22) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). One observation was removed as it showed Cook\u0026rsquo;s distance a magnitude greater than all the others (i.e., EGR with LAI 1.5 COP 25 BH 10, in Florence). As a measure of errors on the indexes, we computed the root mean square error (RMSE) and mean absolute error (MAE) between predicted and observed results and obtained 0.049\u0026deg;C and 0.041\u0026deg;C for the EGR index.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eLinear regression EGR model outputs\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eExtensive green roof\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eMedian T reduction\u0026thinsp;=\u0026thinsp;a\u0026thinsp;+\u0026thinsp;b*COP\u0026thinsp;+\u0026thinsp;c*ln(BH)\u0026thinsp;+\u0026thinsp;d*LAI\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel coefficients\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eN. obs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdj. R\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eF-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e43.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;16\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eVariables coefficients\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eEstimate\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eSt. Err\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003et-value\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003ep-value\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.0784\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0353\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-2.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.0016\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-8.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;12\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ec\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.0722\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0095\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;11\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.0177\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0038\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-4.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFor the GFs, we derived the following index:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$${I}_{GF}= 0.1757 -0.0014COP -0.1164\\text{ln}\\left(LAI\\right)$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e reports the GF model coefficients\u0026rsquo; estimates, standard error t, and p-value. The model neglects BH since it has been found non-significant. GF\u0026rsquo;s model explains 70% of the variance and has a p-value equal to 10\u003csup\u003e\u0026minus;\u0026thinsp;16\u003c/sup\u003e. The residuals are normal (p-val\u0026thinsp;=\u0026thinsp;0.11) and homoscedastic (p-val\u0026thinsp;=\u0026thinsp;0.29). One observation (i.e., GF LAI 1.5 COP 25 BH 10 in Florence) was removed since it was an outlier. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e depicts the relation between the single predictor and the median air temperature variation in the scenarios. It shows that predictions occupy a smaller interval than the observed ones and when LAI equals 3 predicted values coincide. The estimated RMSE and MAE for the index equal 0.041\u0026deg;C and 0.033\u0026deg;C, respectively.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eLinear regression GF model outputs\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eGreen fa\u0026ccedil;ade\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eMedian T reduction\u0026thinsp;=\u0026thinsp;a\u0026thinsp;+\u0026thinsp;b*COP\u0026thinsp;+\u0026thinsp;c*ln(LAI)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel coefficients\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eN. obs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdj. R\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eF-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e90.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;16\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eVariables coefficients\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eEstimate\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eSt. Err\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003et-value\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003ep-value\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.1757\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0204\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;13\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.0014\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-8.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;12\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ec\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.1164\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0095\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-12.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;16\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFor LWs we obtained the following index:\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$${I}_{LW }= 0.096 -0.00157COP -0.0201LAI$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eAs for GFs, also for LWs BH was non-significant as a predictor and was removed to improve the model\u0026rsquo;s residuals. The model is significant and has an adjusted R squared of 0.54, the residuals are homoscedastic (p-val\u0026thinsp;=\u0026thinsp;0.3) but non-normal (p-val\u0026thinsp;=\u0026thinsp;0.01) (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). One observation (i.e., LAI 1.5 COP 25 BH 10 in Florence) presented a Cook\u0026rsquo;s distance of magnitude greater than the other ones and was removed. The predicted values for LWs occupy a smaller interval of values than the observed ones and tend to cluster (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e); RMSE and MAE for the index are equal to 0.043\u0026deg;C and 0.036\u0026deg;C, respectively.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eLinear regression LW model outputs\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eLiving wall\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eMedian T reduction\u0026thinsp;=\u0026thinsp;a\u0026thinsp;+\u0026thinsp;b*\u0026#119862;\u0026#119874;\u0026#119875; + c*L\u0026#119860;\u0026#119868;\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel coefficients\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eN. obs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdj. R\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eF-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e46.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;13\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eVariables coefficients\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eEstimate\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eSt. Err\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003et-value\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003ep-value\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.096\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0215\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.00157\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-8.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;13\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ec\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.0201\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-5.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study provides a tool to ease the beforehand assessment of UHI adaptation plans based on the application of single BIVT scenarios. Specifically, the authors developed three equations capable of providing median UHI mitigation values in an urban area by applying EGRs, GFs, or LWs varying COP, BH, and LAI. Compared to previous research the main aim of the present study is to provide an easy-to-use tool that can be widely used by city council technicians and urban planners as it requires a limited amount of input data and computational time.\u003c/p\u003e \u003cp\u003eTo the authors\u0026rsquo; best knowledge, just a few studies focused on the interaction between BIVTs and pedestrian-level air temperature do exist and are based on complex physical models. To provide some examples, Alexandri et al.\u003csup\u003e28\u003c/sup\u003e developed an ad-hoc dynamic micro-scale model that analyzes the effect of green roofs and green walls on the air temperature in urban canyons and at the rooftop level, Djeddjig et al.\u003csup\u003e29\u003c/sup\u003e developed a TRNSYS hygrothermal model of green walls and a model of mass flows in street canyons. Both these studies provide an estimation of air temperature variation within an urban canyon rather than in a selected urban area. Therefore, they cannot be easily and widely used for UHI mitigation forecasting since they provide punctual air temperature values and need many parameters and input variables. Furthermore, Yang and Wang\u003csup\u003e30\u003c/sup\u003e analyzed the application of EGRs on the urban canopy by applying Monte Carlo simulations, Mazzeo et al.\u003csup\u003e22\u003c/sup\u003e investigated the effect of EGRs on the temperature at the rooftop level using artificial neural networks, and Suter et al.\u003csup\u003e23\u003c/sup\u003e developed an equation for calculating the air temperature variation of the surface layer (from roofs height to 280 m) depending on green roofs. Although such studies provide useful information as roof surface temperature\u003csup\u003e22,30\u003c/sup\u003e and domain averaged evaporation rate\u003csup\u003e23\u003c/sup\u003e, such data can hardly be used by urban planners to evaluate the UHI mitigation potential at the pedestrian level of adaptation plans. The model developed in this study is based on microclimate simulations conducted with ENVI-met, a tool based on the laws of thermodynamics and fluid dynamics, which can simulate the interactions between air, surfaces, and vegetation in an urban context\u003csup\u003e20\u003c/sup\u003e. Therefore, the output reliability depends on the accuracy of both the ENVI-met software and the regression model. Tsoka et al.\u003csup\u003e31\u003c/sup\u003e found that ENVI-met can accurately predict air temperature values under different meteorological conditions. Besides, the regression model has been found to be significant and explains 0.61 and 0.70 of the data variances for EGRs and GFs, respectively, and for both the residuals are normal and homoscedastic. On the other hand, the LW model has been found significant, and its residuals are homoscedastic, but its adjusted R squared equals 0.54 and residuals are non-normal. Nevertheless, by including the effect of the different urban arrangements as intercepts, the accuracy of the LW model greatly improves (i.e., adjusted R squared\u0026thinsp;=\u0026thinsp;0.66) and its residuals are normal (i.e., p-val\u0026thinsp;=\u0026thinsp;0.12). Therefore, as far as the LWs are concerned, the reliability of the model greatly depends on the urban arrangement. However, adding the urban arrangement effect might hinder the genericity of the current model.\u003c/p\u003e \u003cp\u003eThe developed EGR regression model shows a direct correlation between both LAI and COP values and median air temperature mitigation. Our results are in accordance with Suter et al.\u003csup\u003e23\u003c/sup\u003e regarding coverage. As regards LAI, our findings are in agreement with Jamel et al.\u003csup\u003e32\u003c/sup\u003e, even though they observed that after a certain LAI threshold, the temperature reduction reaches a plateau. Furthermore, we found that increasing COP is more effective in mitigation than increasing LAI. Such a finding is also confirmed by Iaria and Susca\u003csup\u003e27\u003c/sup\u003e. Contrarily, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, also shows an indirect correlation between BH and air temperature mitigation, meaning that the effectiveness of EGRs in mitigating air temperature at the pedestrian level decreases when the height of the rooftops where EGRs are installed increases. Such a result is in accordance with previous literature (e.g.\u003csup\u003e33\u003c/sup\u003e) and is explained with a greater distance of the mitigation source from the target. Furthermore, as in \u003csup\u003e27\u003c/sup\u003e, we found that BH equal to approximately 40 m can be considered as a threshold above which the application of EGRs has a negligible effect on air temperature reduction at the pedestrian level\u003csup\u003e34,35\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAs far as the GF and the LW models are concerned, BH was found to be a negligible input parameter, in comparison with LAI and COP and was therefore removed from the models. This finding might indicate that proximity to the target is as important \u0026ndash;or even more\u0026ndash; as the greened area increases in mitigating air at the desired height. Indeed, the mitigation effect of GFs and LWs at the pedestrian level is greater when the source of evapotranspiration is closer to the target (i.e., street level). This result should be further investigated in future studies. In Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e both the GF and the LW models, respectively, predict a direct relationship between both LAI and COP and the mitigation in air temperature. When applying GFs, low values of LAI (e.g., LAI\u0026thinsp;=\u0026thinsp;1.5) lead to a negligible median reduction or even an increase in temperature, while high values (e.g., LAI\u0026thinsp;=\u0026thinsp;5) lead to a median reduction of about 0.1\u0026deg;C. Moreover, when GFs are deployed, an increase in LAI is more effective than an increase in COP in mitigating air temperature. Contrarily, in the application of LWs, the effect of an increase in COP on temperature mitigation (ranging from about \u0026minus;\u0026thinsp;0.04\u0026deg;C with COP\u0026thinsp;=\u0026thinsp;25% to about \u0026minus;\u0026thinsp;0.12\u0026deg;C with COP\u0026thinsp;=\u0026thinsp;100%) is significantly greater than an increase in LAI (ranging from about \u0026minus;\u0026thinsp;0.07\u0026deg;C with LAI\u0026thinsp;=\u0026thinsp;1.5 to about \u0026minus;\u0026thinsp;0.09\u0026deg;C with COP\u0026thinsp;=\u0026thinsp;100%).\u003c/p\u003e \u003cp\u003eAs far as limitations and shortcomings are concerned, one limit of the proposed model resides in its applicability to the Mediterranean climate only, specifically, the Csa climate zone. Nonetheless, the same methodology might be applied to other climate zones. Furthermore, climatic variables such as wind direction and speed, that have an impact on mitigation\u003csup\u003e35\u003c/sup\u003e, were kept constant in the study areas by using typical summer days as forcing climatic conditions. Future research might improve the index with changing climatic variables to get insights about the mitigation performance in changing conditions. Also, the model assumes that green roofs and walls are fully irrigated at any time. BIVTs\u0026rsquo; performance depends on irrigation and in particular water availability for plants\u003csup\u003e11,36\u003c/sup\u003e, therefore, in real conditions water availability is a crucial factor. Lastly, the vegetation layer of the BIVTs is composed of standard, well-investigated plants such as \u003cem\u003eSedum sediforme, Nephrolepis exalata\u003c/em\u003e, and \u003cem\u003eHedera helix\u003c/em\u003e for EGR, GF, and LW, respectively. We neglected the use of different local species as it was outside the scope of the study, but as transpiration is dependent on plant traits, differences might occur when other species are implemented in BIVTs. By applying a single BIVT most of the mitigation values range from 0 to \u0026minus;\u0026thinsp;0.2\u0026deg;C. Such values seem to be low, nevertheless, mitigation is not evenly distributed within the areas: The effect of BIVTs tends to be distributed near buildings or in the upwind direction. Specifically, in 25% of the urban areas the temperature mitigation is significantly higher than the median one (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), and punctual mitigation values can be up to \u0026minus;\u0026thinsp;0.83\u0026deg;C.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eNevertheless, the proposed model might also indicate a potential ineffectiveness of BIVT-based adaptation plans. Such results are equally important since might entail the implementation of more synergic solution sets\u003csup\u003e19\u003c/sup\u003e.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThe present study reports on the development of an index to predict the mitigation potential of single BIVT-based scenarios \u0026mdash; specifically, EGRs, GFs, and LWs \u0026mdash; to eradicate UHI in urban areas in the Csa climate zone characterized by an urban density ranging from 0.3 to 0.4. The developed indexes can be applied to built areas by varying the parameters that mostly influence the UHI mitigation potential of BIVT-based scenarios: BH, COP, and LAI.\u003c/p\u003e \u003cp\u003eThe developed indexes are useful since they can ease the work of planners and city council technicians who want to assess beforehand the effectiveness of BIVT-based UHI adaptation plans. The developed indexes can forecast and compare the effect of adaptation BIVT-based plans using easily available data overcoming the limits of other methods (i.e., \u003csup\u003e20\u003c/sup\u003e), and their outputs remain within an acceptable range of error.\u003c/p\u003e \u003cp\u003eFurthermore, the indexes are novel, since no studies have been found in published literature providing such a tool and, differently from other research that focuses on providing punctual temperature variation values, the current study provides spatially median variation in urban temperature which can more reliably support urban planners in designing UHI adaptation plans. The novelty of the proposed indexes also resides in the use of a methodology based both on the development of 269 ENVI-met models and simulations and on the linear regression models. The proposed methodology can be used for developing indexes for other climate zones. The precision of the model greatly relies on that of the ENVI-met tool, therefore, in the future, when the ENVI-met model will be further improved, the developed indexes, fed with other simulations, might decrease their error. Yet, limitations exist. For instance, the model might be further enriched with other parameters, such as plant watering data or the models can be developed on single urban arrangements. However, although the precision of the model would probably benefit from such implementations, the implementation of more parameters would likely increase the complexity of the indexes and the focus on single urban arrangements would make the indexes less applicable, altogether, such improvements might hinder urban planners from their use.\u003c/p\u003e \u003cp\u003eAltogether the developed indexes can greatly contribute to ease the work of planners to eradicate urban overheating and the proposed methodology can pave the way for the development of other indexes capable of assessing, with different levels of error, the effectiveness of UHI adaptation plans.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eFor the development of the indexes, three urban areas prone to the UHI formation in three Italian cities belonging to the Csa climate zone were selected. The study areas of Via Lanciani in Rome, Viale Kennedy in Bari, and the neighborhood of Gavinana in Florence are characterized by similar extension, LCZ, and urban density, but different layouts were chosen (Fig.\u0026nbsp;6, 1. Identification of the study areas). Indeed, the three urban areas have been chosen to represent the main typical regional layouts\u003csup\u003e33,35\u003c/sup\u003e to generalize the results as much as possible. Specifically, via Lanciani urban area presents a scattered layout with similar-sized squared buildings and roads arranged in N-S and E-W directions; Bari shows an array layout with buildings displayed along parallel streets, while Florence buildings are characterized by an enclosing layout with buildings forming closed polygons with internal courtyards. Another requisite for the selection of the areas was the access to monitored historical meteorological data recorded by a meteorological station positioned within the selected built area and another one positioned in the rural environs. These latter data were used to force the ENVI-met models (Fig.\u0026nbsp;6). The used methodology has been thoroughly described in a previous study\u003csup\u003e27\u003c/sup\u003e, however, the present study has been enriched with 152 new scenarios Tab.S1.\u003c/p\u003e\n\u003cp\u003eAll the models were run in ENVI-met for the typical summer day for each city. To identify the typical summer days, we compared the average hourly temperature profile of the hottest month (data from \u003csup\u003e37\u003c/sup\u003e) with the real-day temperature profile recorded in the urban stations, after removing the rainy days and selecting the one with the smallest RMSE and MAE.\u003c/p\u003e\n\u003cp\u003e1. UHI-mitigation criteria and values selection\u003c/p\u003e\n\u003cp\u003eThree parameters are predicted to be influential in UHI mitigation by BIVTs: BH\u003csup\u003e10,11,35,38\u003c/sup\u003e, COP\u003csup\u003e39\u003c/sup\u003e, and LAI\u003csup\u003e38\u003c/sup\u003e. For the development of mitigation and control scenarios, we assumed the following BH: 5, 10, 20, 30, and 40 m. The incremental height has been chosen as multiple of the grid used for the development of the scenarios (see section 3. Mitigation scenarios). The maximum has been set at 40 m since previous research has shown that above this threshold green roofs have null mitigation potential\u003csup\u003e32\u003c/sup\u003e. As far as COP is concerned, we considered 25, 50, 75, and 100%, besides, LAI values were set to 1.5, 3, and 5 standing for low, medium, and high foliage density, respectively. All the mitigation scenarios were created by applying a single BIVT to the study area with combinations of the possible values of the three criteria.\u003c/p\u003e\n\u003cp\u003e2. Study areas modeling\u003c/p\u003e\n\u003cp\u003eThe selected areas were modeled in ENVI-met. The software requires the 3D arrangement of all urban elements (e.g., buildings, trees, shrubs, paved and natural surfaces, BIVTs) as well as their inner structure and physical properties to simulate the interaction between them and produce microclimatic outputs such as the potential air temperature. Additional information necessary to improve the model accuracy is meteorological data to input as forcing conditions. Data from the urban station were used to test the ENVI-met accuracy while data from the rural stations were used as forcing to the baseline and mitigation scenarios (See methodology in\u003csup\u003e27\u003c/sup\u003e).\u003c/p\u003e\n\u003cp\u003e3. Mitigation scenarios\u003c/p\u003e\n\u003cp\u003eIn addition to the 117 original scenarios developed by\u003csup\u003e27\u003c/sup\u003e, created by varying the value of one parameter per time while keeping the other constant at the maximum value, we added 139 additional intermediate scenarios \u0026ndash;scenarios with the parameters changing together at low values\u0026ndash; to account for BIVTs efficacy in sub-optimal conditions and improve the results. All scenarios are shown in tab.S.1 supplementary materials. The intermediate scenarios are representative of the sub-optimal condition of LAI (i.e., LAI 1.5) and coverage percentage (i.e., COP 25%, 50%, and 75%) for the different classes of building height in the three study areas.\u003c/p\u003e\n\u003cp\u003e4. ENVI-met simulations\u003c/p\u003e\n\u003cp\u003eEvery scenario was simulated using the same grid parameters. Specifically, we choose a grid of 5 m (x) x 5 m (y) x 5 m (z) to get accurate outputs in a relatively short computational time ranging from six to eight hours. For each simulation two days were simulated, the first one for stabilizing the model and the second one to get the microclimate outputs. All the simulations were run with an AMD Ryzen 53,600 6-Core processor and 32.0 Gb RAM. The original scenarios were simulated using ENVI-met 4.4.6 while for the additional scenarios, ENVI-met 5.1 was used. The difference in the ENVI-met version required the baseline scenarios to be simulated in both 4.4.6 and 5.1 to nullify any difference between the different releases.\u003c/p\u003e\n\u003cp\u003e5. Data extraction and analysis\u003c/p\u003e\n\u003cp\u003eUsing ENVI-met LEONARDO, we extracted in CSV format the potential air temperature data at pedestrian level related to the hottest hour in the three study areas, namely 1.00 pm for Rome and 2.00 pm for Bari and Florence of all the simulated scenarios. By means of\u003csup\u003e40\u003c/sup\u003e we calculated for each cell of the modeled areas the difference between the mitigation scenario and the control one (e.g., mitigation scenarios in Rome with BH equal to five were subtracted from baseline scenario of Rome with BH equal to five). Subsequently, we calculated the mean, median, 1st and 3rd quartile air temperature difference of each mitigation scenario. Then, we used linear regression to calculate the mitigation index of each BIVT (function lm(), baseline R). We tested heteroskedasticity with a Breusch-Pagan test, package lmtest\u003csup\u003e41\u003c/sup\u003e function bptest(), and normality with function shapiro.test() in baseline R; Cook\u0026rsquo;s distance was considered by plotting the model\u0026rsquo;s residuals. Lastly, we used functions rmse() and mae() of\u003csup\u003e42\u003c/sup\u003e to assess the RMSE and MAE of the equations in predicting mitigation. For graphs packages, ggplot2 and gghalves\u003csup\u003e43\u003c/sup\u003e were used.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by MASE (Italian Ministry for the Environment and Energy Security) [Fondo per il finanziamento delle attivit\u0026agrave; di ricerca e di sviluppo di interesse generale per il sistema elettrico nazionale].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization, TS and JI; methodology, TS and JI; formal analysis JI; investigation TS, JI, and FZ; writing-original draft preparation TS, JI, and FZ; writing-review and editing TS, JI and FZ; visualization JI; supervision TS. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets necessary for preparing ENVI-met simulations are available in\u0026nbsp;\u003csup\u003e27\u003c/sup\u003e supplementary materials. All ENVI-met-related files, extracted temperature layers from simulation results, R Scripts and data generated for computing models are available on request to Jacopo Iaria, email: [email protected].\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLandsberg, H. E. \u003cem\u003eThe Urban Climate\u003c/em\u003e. (Academic Press, 1981).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, P., Imhoff, M. L., Wolfe, R. E. \u0026amp; Bounoua, L. Characterizing urban heat islands of global settlements using MODIS and nighttime lights products. Canadian Journal of Remote Sensing 36, 185\u0026ndash;196 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, W. T. K. \u003cem\u003eet al.\u003c/em\u003e Economic valuation of temperature-related mortality attributed to urban heat islands in European cities. Nat Commun 14, 7438 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSantamouris, M., Cartalis, C., Synnefa, A. \u0026amp; Kolokotsa, D. On the impact of urban heat island and global warming on the power demand and electricity consumption of buildings\u0026mdash;A review. Energy and Buildings 98, 119\u0026ndash;124 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSantamouris, M. On the energy impact of urban heat island and global warming on buildings. Energy and Buildings 82, 100\u0026ndash;113 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, H. \u003cem\u003eet al.\u003c/em\u003e Interaction between urban heat island and urban pollution island during summer in Berlin. Science of The Total Environment 636, 818\u0026ndash;828 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWong, N. H., Tan, C. L., Kolokotsa, D. D. \u0026amp; Takebayashi, H. Greenery as a mitigation and adaptation strategy to urban heat. Nat Rev Earth Environ 2, 166\u0026ndash;181 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEdmondson, J. L., Stott, I., Davies, Z. G., Gaston, K. J. \u0026amp; Leake, J. R. Soil surface temperatures reveal moderation of the urban heat island effect by trees and shrubs. Sci Rep 6, 33708 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSusca, T., Zanghirella, F. \u0026amp; Del Fatto, V. Building integrated vegetation effect on micro-climate conditions for urban heat island adaptation. Lesson learned from Turin and Rome case studies. Energy and Buildings 295, 113233 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSusca, T., Zanghirella, F., Colasuonno, L. \u0026amp; Del Fatto, V. Effect of green wall installation on urban heat island and building energy use: A climate-informed systematic literature review. Renewable and Sustainable Energy Reviews 159, 112100 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSusca, T. Green roofs to reduce building energy use? A review on key structural factors of green roofs and their effects on urban climate. Building and Environment 162, 106273 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDunnett, N. \u0026amp; Kingsbury, N. \u003cem\u003ePlanting Green Roofs and Living Walls\u003c/em\u003e. vol. VII (Timber Press, Portland, 2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEnvironmental Protection Agency. Reducing Urban Heat Islands: Compendium of Strategies. Green Roofs. (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCastleton, H. F., Stovin, V., Beck, S. B. M. \u0026amp; Davison, J. B. Green roofs; building energy savings and the potential for retrofit. Energy and Buildings 42, 1582\u0026ndash;1591 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEnvironmental Protection Agency. Reducing Urban Heat Islands: Compendium of Strategies. Green Roofs. (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eManso, M. \u0026amp; Castro-Gomes, J. Green wall systems: A review of their characteristics. Renewable and Sustainable Energy Reviews 41, 863\u0026ndash;871 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLondon City Council. London Plan 2011. \u003cem\u003eLondon City Hall\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.london.gov.uk//what-we-do/planning/london-plan/past-versions-and-alterations-london-plan/london-plan-2011\u003c/span\u003e\u003cspan address=\"https://www.london.gov.uk//what-we-do/planning/london-plan/past-versions-and-alterations-london-plan/london-plan-2011\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIrga, P. J. \u003cem\u003eet al.\u003c/em\u003e The distribution of green walls and green roofs throughout Australia: Do policy instruments influence the frequency of projects? Urban Forestry \u0026amp; Urban Greening 24, 164\u0026ndash;174 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, Y. \u003cem\u003eet al.\u003c/em\u003e Beating urban heat: Multimeasure-centric solution sets and a complementary framework for decision-making. Renewable and Sustainable Energy Reviews 186, 113668 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHigh-Resolution 3D Modeling of Urban Microclimate with ENVI-met Software. \u003cem\u003eENVI-met\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.envi-met.com/\u003c/span\u003e\u003cspan address=\"https://www.envi-met.com/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalany, F., Ng, A. W., Muttil, N., Muthukumaran, S. \u0026amp; Wong, M. S. Green Infrastructure as an Urban Heat Island Mitigation Strategy\u0026mdash;A Review. Water 12, 3577 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMazzeo, D., Matera, N., Peri, G. \u0026amp; Scaccianoce, G. Forecasting green roofs\u0026rsquo; potential in improving building thermal performance and mitigating urban heat island in the Mediterranean area: An artificial intelligence-based approach. Applied Thermal Engineering 222, 119879 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuter, I., Maksimović, Č. \u0026amp; van Reeuwijk, M. A neighbourhood-scale estimate for the cooling potential of green roofs. Urban Climate 20, 33\u0026ndash;45 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStewart, I. D. \u0026amp; Oke, T. R. Local Climate Zones for Urban Temperature Studies. Bulletin of the American Meteorological Society 93, 1879\u0026ndash;1900 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDiffenbaugh, N. S., Pal, J. S., Giorgi, F. \u0026amp; Gao, X. Heat stress intensification in the Mediterranean climate change hotspot. Geophysical Research Letters 34, (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKottek, M., Grieser, J., Beck, C., Rudolf, B. \u0026amp; Rubel, F. World Map of the K\u0026ouml;ppen-Geiger climate classification updated. \u003cem\u003emetz\u003c/em\u003e 15, 259\u0026ndash;263 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIaria, J. \u0026amp; Susca, T. Analytic Hierarchy Processes (AHP) evaluation of green roof- and green wall- based UHI mitigation strategies via ENVI-met simulations. Urban Climate 46, 101293 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlexandri, E. \u0026amp; Jones, P. Temperature decreases in an urban canyon due to green walls and green roofs in diverse climates. Building and Environment 43, 480\u0026ndash;493 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDjedjig, R., Bozonnet, E. \u0026amp; Belarbi, R. Modeling green wall interactions with street canyons for building energy simulation in urban context. Urban Climate 16, 75\u0026ndash;85 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, Jiachuan \u0026amp; Wang, Zhi-Hua. Physical parameterization and sensitivity of urban hydrological models: Application to green roof systems. Building and Environment 75, 250\u0026ndash;263 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTsoka, S., Tsikaloudaki, A. \u0026amp; Theodosiou, T. Analyzing the ENVI-met microclimate model\u0026rsquo;s performance and assessing cool materials and urban vegetation applications\u0026ndash;A review. Sustainable Cities and Society 43, 55\u0026ndash;76 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJamei, E. \u003cem\u003eet al.\u003c/em\u003e Investigating the cooling effect of a green roof in Melbourne. Building and Environment 246, 110965 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNg, E., Chen, L., Wang, Y. \u0026amp; Yuan, C. A study on the cooling effects of greening in a high-density city: An experience from Hong Kong. Building and Environment 47, 256\u0026ndash;271 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, H., Ooka, R., Huang, H. \u0026amp; Tsuchiya, T. Study on mitigation measures for outdoor thermal environment on present urban blocks in Tokyo using coupled simulation. Building and Environment 44, 2290\u0026ndash;2299 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJin, C., Bai, X., Luo, T. \u0026amp; Zou, M. Effects of green roofs\u0026rsquo; variations on the regional thermal environment using measurements and simulations in Chongqing, China. Urban Forestry \u0026amp; Urban Greening 29, 223\u0026ndash;237 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVaezizadeh, F., Rashidisharifabad, S. \u0026amp; Afhami, R. Investigating the Cooling Effect of Living Walls in the Sunken Courtyards of Traditional Houses in Yazd. European Journal of Sustainable Development 5, 27\u0026ndash;27 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCTI Comitato Termotecnico Italiano -. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cti2000.it/\u003c/span\u003e\u003cspan address=\"https://www.cti2000.it/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorakinyo, T. E., Lai, A., Lau, K. K.-L. \u0026amp; Ng, E. Thermal benefits of vertical greening in a high-density city: Case study of Hong Kong. Urban Forestry \u0026amp; Urban Greening 37, 42\u0026ndash;55 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorakinyo, T. E., Dahanayake, K. W. D. Kalani. C., Ng, E. \u0026amp; Chow, C. L. Temperature and cooling demand reduction by green-roof types in different climates and urban densities: A co-simulation parametric study. Energy and Buildings 145, 226\u0026ndash;237 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR: The R Project for Statistical Computing. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.r-project.org/\u003c/span\u003e\u003cspan address=\"https://www.r-project.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeileis, A. \u0026amp; Hothorn, T. Diagnostic Checking in Regression Relationships. R News vol. 2 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003emfrasco. \u003cem\u003e/Metrics\u003c/em\u003e. (2024).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4259407/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4259407/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eUrban heat island (UHI) can jeopardize urban inhabitants, but the installation of green roofs (GRs) and walls (GWs) can contribute to mitigating it. The present study provides a novel index to easily predict the spatially median variation in air temperature at pedestrian height related to the application of GR- and GW- -based scenarios on the hottest hours of a typical summer day varying building height (BH), coverage percentage (COP), and leaf area index (LAI). The index is meant to be applied to built areas with 0.3\u0026ndash;0.4 urban density in the Mediterranean climate and is derived from a linear regression model fed with the outputs of 269 simulations of three urban areas developed and run in ENVI-met software. The developed models are all highly significant. GR model shows that the mitigation is influenced by all three parameters, and it can estimate the mitigation with a mean standard error of 0.05\u0026deg;C. GW model shows that BH is not influential in decreasing air temperature compared to the other parameters. GF and living wall (LW) index can predict the mitigation with an error of 0.03\u0026deg;C and 0.04\u0026deg;C, respectively. However, for the LW model, further parameters should be considered to improve its reliability.\u003c/p\u003e","manuscriptTitle":"Development of predictive indexes for evaluating UHI adaptation potential of green roof- and wall- based scenarios in the Mediterranean climate","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-09 04:32:46","doi":"10.21203/rs.3.rs-4259407/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-06-10T05:32:52+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-06-07T16:34:23+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"61612302066193960131723285720656413890","date":"2024-05-21T11:13:27+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-10T17:06:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"d56ac49e-e972-4950-b656-83ef98a3bae8","date":"2024-05-07T11:38:29+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-05-07T06:37:54+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-05-03T13:13:54+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-05-03T13:09:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-05-03T13:06:53+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-04-12T18:48:07+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3345da60-d582-4653-bec6-7808973ec0d0","owner":[],"postedDate":"May 9th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":31657552,"name":"Earth and environmental sciences/Climate sciences"},{"id":31657553,"name":"Earth and environmental sciences/Environmental sciences"},{"id":31657554,"name":"Physical sciences/Engineering"}],"tags":[],"updatedAt":"2024-09-23T16:01:10+00:00","versionOfRecord":{"articleIdentity":"rs-4259407","link":"https://doi.org/10.1038/s41598-024-67567-9","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2024-09-17 15:57:20","publishedOnDateReadable":"September 17th, 2024"},"versionCreatedAt":"2024-05-09 04:32:46","video":"","vorDoi":"10.1038/s41598-024-67567-9","vorDoiUrl":"https://doi.org/10.1038/s41598-024-67567-9","workflowStages":[]},"version":"v1","identity":"rs-4259407","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4259407","identity":"rs-4259407","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}