4.1. Relationship between measured and simulated results
In order to evaluate outdoor thermal environment, the index RMSE (root-mean-square error) is calculated for checking deviation. This index is a normally used in validating the gap between the observed and the predicted values, which is an important factor for testing the simulated results [60, 61]. If this index can approach or reach 0, the most accurate results are obtained. A lower RMSE values represent that the simulated data is closed to measured result. Figure. 10 shows the index-RMSE between the simulated and measured result.
The RMSE of air temperature in Figure.10-a shows the big deviation occurs at point-5, reaching 3.5℃, this may be caused by the position of the measured instruments. In order to consider the peoples’ safety, the instruments are not stood in the middle of the road and instead are fixed on the sidewalks. Figure. 10-b shows the big deviation also occurs at point-5 in final results of relative humidity. In the field of wind velocity (Figure. 10-c), the simulated model shows a favorable correlation with the wind velocity, with changing from 0.1 to 0.2 m/s.
Besides the calculation of the RMSE, the analysis of coefficient of determination between simulated and measured result is another very important process to assess the accuracy. A very good linear regression is found , as shown in Figure. 11-13, where the coefficient of determination (R2) of air temperature is from 0.741 to 0.9582, R2 values of relative humidity for these points range from 0.7469 to 0.9693, and the values for wind velocity are within 0.7724 and 0.9424. The deviation may be caused some uncertain factors during the working time. These deviations are smaller than in previous published studies [63–65]. Our final findings proved that ENVI-met tool is an accurate software to finish the task in our work.
4.2 The new thermal environment under new cases
As mentioned, the index PET is used for assessing people’s thermal sensation. Considering the final results of numerical simulation, the thermal environment during the two measured days is shown in Figure. 14 and Figure. 15. These two figures both show that the peoples’ thermal sensation during the daytime are standing ‘Very hot’ and ‘Hot’ stages in accordance with the distribution of PET values int the researched climate zone . Therefore, improving outdoor thermal environment is very necessary.
Based on the previous studies, it’s obvious that a stronger cooling effect will be got with a little higher background air temperature [66,67], also it has proved that the positive effect of vegetation in sunny ,hot days will be higher than cloudy , cold days , in addition, the effect of paving material with higher albedo will show a better contribution in reducing temperature in sunny days than in cloudy days .
In common sense, the hottest time mainly appears from 2:00pm to 4:00pm , in this study, the hottest period appears at 3:00pm, therefore, the PET on 3:00pm, Jul. 28th is selected for calculating further analysis. As Figure. 16 shows that all the selected points are suffering from a high heat stress, nearly, the hottest PET of all the points can reach 60℃PET at 3:00pm. Under this situation, it is very necessary for us to ameliorate the UHI effect and improve people’s thermal sensation.
To be mentioned, the story of the building in the commercial pedestrianized block won’t be exceeded three-story, and the coverage ratio of the vegetation shouldn’t be less than 25% of the total site in accordance with local design specifications . The coverage ratio of different parameters in existing scenario (Base case) is shown in Figure. 17, the total building coverage ratio
occupies 55.37%, in which three-story building is only 10.35%. In addition, the vegetation coverage ratio just occupies 3.96%. These factors can lead to a worse thermal environment in hot summer.
Based on existing scenario, new strategies are put forward. The new cases under the scientific hypothesis are shown in Table. 7.
In new cases (Figure. 18), case-1 aims at increasing grass coverage ratio to understand the cooling effect, where the coverage ratio of grass increases from 2.94% to 23.98%. Like case-1, the coverage ratio of tree is increased to 22.06% in case-2 and evaluate its cooling effect. In third case, replacing the ground surface in existing scenario with a new paving material with higher albedo in improving thermal safety and reducing energy cost. The last case (case-4) is through increasing
three-story building coverage ratio and building height to understand its function, where the coverage ratio of three-story building is up to 55.37%.
The cooling effect of different parameters is shown as:
Where PET is people’ thermal sensation under existing scenario, PETS is the new thermal sensation under new cases. Upon the new cases, Figure. 19 shows that the improvement of PET appeared in the research site. The new distribution of PETs at peak time (3:00pm) has shown that the increase in tree coverage ratio (case-2) can largely change thermal environment at daytime, especially in open space (Point-1 and Point-2), in which ∆PET ranges from 1.5 to 3.9℃, meanwhile, increasing tree coverage ratio can also make a contribution to reduce PET in canyon space, where ∆PET can be changed within 1.2 and 8.1℃. This effect can be attributed to transpiration and providing shadow of the leaf at daytime. In case-4, increasing coverage ratio of three-story building and average building height can obviously improve thermal environment, which can be ranged from 2.1 to 12.5℃ PET in canyon space, but the thermal environment can’t be changed too much (Point-1 and Point-2). Meanwhile, increasing grass coverage ratio can also reduce PET, but the result is limited. What’s worse, changing the paving material with higher albedo (case-3) may result in a worse thermal environment in open space (Point-2), even paving material with higher albedo may cool down the ground surface, which will also reflect more solar radiation on humans’ body, thus leading to
a worse PET, and the effect of grass (case-1) is not obvious.
The former results just display the distribution of the improvement of PET through synergistic effect under new cases in general. In order to provide a quantifiable effect of different parameters, a more detailed analysis about the cooling effect is shown in next part.
4.3. Detailed correlation between new case and people’s thermal sensation
The whole selected block is composed of canyon space and open space. The detailed effect of new cases in open space is displayed in Figure. 20. The index R2 between different parameters and ∆PET demonstrates the proportion that can be interpreted by various regression analysis. In this figure, we can find a strong positive correlation between tree coverage ratio and ∆PET, where it can be observed that a 5% increase will reduce 0.4℃ PET. Meanwhile, it’s obvious that an invalid correlation is found between three-story building height and ∆PET, with an irregular R2 0.4844, in addition, an ascension in average building height will contribute to a lower SVF and higher H/W, it’s shown that the relationship between the H/W and ∆PET tends to be irregular (R2＝0.2971), what’s worse, as SVF develops, a negative correlation between these two parameters will appear. After increasing the percentage of grass, it’s found that a 5% increase in it will reduce0.15℃ PET. But a 3% changing with the new paving material will lead to a 0.2℃ PET increase.
Different from the thermal environment in open space, the most essential strategy (Figure. 21) in reducing PET in canyon space is increasing coverage ratio of three-story building, in which a 10% increase in percentage of three-story building will reduce 0.5℃ PET, in addition, a 0.1 increase in SVF will lead to increase 0.11℃PET, while an ascension of 0.1 in H/W can reduce 0.15℃ PET at peak time. Like open space, increasing coverage ratio of tree can obviously reduce PET largely, a 5% increase will contribute to reduce 0.25℃ PET at peak time. According to the final correlation analysis of other two cases (Grass and paving
material), it’s observed that the cooling effect of these two cases are limited.
In order to help local manager and policy makers understand the cooling effect of different strategy briefly. A new figure (Figure. 22) is conducted to assess the comprehensive standard.