The Deviation between the Field Measurement and ENVI-met Outputs in Winter- A Cases Study in a Traditional Dwelling Settlement of China

: 11 Background: Numerical simulation has been used for assessing the process of outer environment 12 performance calculations accurately. The ENVI-met has been used in studies in simulating the 13 microclimate and energy cost in hot summer. On 27th November,2019, the latest ENVI-met v4.4.4 14 has been published, which is the first edition for forecasting the outdoor microclimate in winter, therefore, its accuracy is still needed to be discussed. Methods: This study aims at evaluating the 16 effectiveness of the latest edition on the prediction of the thermal environment of courtyard and outdoor space in a traditional dwelling settlement in cold climate zone of northern China by means 18 of validating the measured results against simulated results. Results: The final result shows a small 19 deviation between simulated and measured results for assessing the microclimate in the research site. Conclusions: For this reason, the new ENVI-met v4.4.4 can be a reliable tool for forecasting the winter microclimate in northern China, and also this study also provides a basic database for improving resident’s thermal comfort in cold climate zone of China in future studies.


25
As society develops, architects and urban planners have faced the important task of designing 26 the suitable outdoor and indoor environment for human while saving energy [1]. In developed 27 countries, the building energy cost occupies 40% of the total energy used in humans' daily lives [2], 28 in China, the building energy consumption has increased by nearly 45% in the last two decades under 29 the rapid development of the urbanization [3][4][5]. To design and ensure a healthy and comfortable

37
The cave dwelling is an ancient residential form in northern China, which also has more than 38 4000 years history [7]. So far, there are still too many people living here, therefore, many governments 39 and civil organizations have taken some measures to improve living environment in such traditional-40 style dwellings. The microclimate in the cave-dwelling is affected by different kinds of elements such 41 as vegetation, water body wind and geometry of this region [8,9]. All of them will affect the 42 distribution of air temperature in the courtyard. In the aspect of the cave dwelling, courtyard can 43 play an important role in the design of zero energy cost buildings.

45
As technology develops, the utilization of simulated process based on the Computational Fluid 46 Dynamics (CFD) is becoming popular. Among the development, we can find Design Builder, ANSYS Fluent and ENVI-met are the most popular three. Design Build provides a platform for assessing 48 architecture energy consumption hat has been reported not to reflect the true air temperature in 49 outdoor courtyard [10]. ANSYS Fluent can be used to simulate the turbulence and fluid models, it 50 needs a not proportional amount of effort and time to be applied the researched target [11].Since we 51 aim at discussing the microclimate in the courtyard, the most suitable tool is the software ENVI-met, 52 which we will introduce as follows. This tool can be able to analyze the interactions between outdoor 53 microclimate and air, soil, vegetation and buildings. Based on this reason, it's been used in many 54 studies about urban microclimate, in which it provides a platform with a horizontal resolution from 55 0.5 to 5m and a time frame of 24-48h with a time step 1-5s. It also has been accepted for many studies 56 in the research about outdoor thermal environment in summer [12-20].          In this software platform, the fundamental equations from the physical model are corresponding 100 with 1) mean air flow, 2) temperature and humidity, 3) turbulence and exchange process, and 4) 101 radiative fluxes, in addition, this software also includes biometeorological and particle dispersion 102 models. In this study, the air temperature is described as the example:

103
The potential air temperature θ could be calculated by the utilization of the combined advection-104 diffusion equations: ℎ is the turbulent exchange coefficient for heat, is utilized to define the heat 109 exchange between plants and air, LAD is the leaf area density, ,ℎ is the heat flux, and Tf is the 110 foliage temperature, Ta is the air temperature.

111
The 1 is the change in air temperature under the effect of divergence of the long wave 112 radiation, and the long wave radiation fluxes can be simply written as: Stefan-Boltzman constant, which is 5.67 • 10 -8 W m -2 K -4 , the (sky view factor) is a 126 measurement for the whole sky seen from the center of a grid cell, and is the maximum cutoff 127 angle in spatial direction . 128 In the process of the simulation, the ENVI-met has the shortcoming. For example, the incoming 129 longwave radiation is an important factor for affecting air temperature, which is emitted by building,      The on-site measurement for collecting air temperature (Ta) and relative humidity (Rh) start 148 from 9:00am to 6:00pm, all the measured days are windless and cloudless. The utilized instrument in 149 this study is a stable meteorological station called HOBO MX2300, Table 3 shows the detailed 150 information of the used instrument.
151 Table 3 The measured variables and parameter of the instrument Hr (%) ±2.5% 0-100% 1min 0.01% After the field survey by researchers, the whole settlement is divided into 9 points ( Figure 6 Table 4). In addition, a hemisphere photo at 1.5m in each selected point is taken 156 by using a fish-eye camera, the measured SVF (Sky View Factor) is calculated by software Ray-man, 157 which will be validated against by simulated results in ENVI-met platform 158 Figure 6 The selected points in the on-site measurement

159
To be mentioned, all the points in courtyard are conducted in all the measured days. Different 160 from courtyard, each selected point in outdoor space is carried out in every measured day, the former 161 three points are conducted on January (29th, 30th, 31th), and point-4 is carried out on February 1st. Table 4 shows the detailed characteristics of each selected point.

170
In this study, the whole simulation is conducted in total 96 hours (4 days), because the limited 171 simulated period of this software is only 48 hours, so the whole process is divided into two parts. The

172
first part is over a 48-hour period, starting from midnight 00:00, Jan 29th, 2019 with calculations every 173 one minute, the simulated results were output on an hourly basis, and the initial input meteorological 174 data in the simulated process is shown in Table 5, 6 and 7.

186
The final results of the air temperature of the on-site measurement during the measured period 187 is shown in Figure 8. It is very obvious that the maximum air temperature appears from 13:00 to 188 14:00, ranging from 0.31℃ to 10.71℃, in addition, the minimum air temperature changes from -11.62 189 ℃ to 1.66℃ at 9:00. In the first measured day (29th Jan), the air temperature in courtyard space is

200
Different from the former two days, the changing trend of the curves tend to be smaller.

201
Meanwhile, the courtyard point (point-5) displays the best performance, in which the air temperature 202 can rise to above 0℃ from 13:00 to 14:00. In the last measured day (1st Feb), because of the snow in 203 the former day, the air temperature is still very low in the morning. Similar to the former mentioned 204 three days, the point-5 and point-7 have a higher air temperature at daytime.

205
The final measured results illustrate the geometry is an important factor, which can affect the air 206 temperature not only in outdoor space but also in courtyard space.  240 Figure 12 shows the R 2 of the courtyard space, in which we can obviously find that the latest 241 edition of ENVI-met (v4.4.4) can forecast air temperature in cold winter accurately, with R2 changing 242 from 0.7558 to 0.9835.

208
243 Figure 12 The comparison between measured and simulated air temperature in outdoor space

244
In addition, Figure 13 display the RMSE between the measured and simulated air temperature,

245
in which point-4 has the largest error at daytime, reaching 1.81℃, and the point-3 has the highest 246 accuracy.
247 Figure 13 The RMSE between measured and simulated data in courtyard

248
Meanwhile, the values of RMSEs and RMSEu are conducted for assessing the accuracy of the 249 simulation, the detailed information is shown in  Figure 14 shows the final results, it is obvious that a highly accurate simulated results of the  Table 9.
257 Figure 14 The correlation and RMSE between measured and simulated data in courtyard 258

260
In the ENVI-met platform, the potential air temperature will be largely affected by the longwave 261 radiation and the heat exchange between plants and air. But in the real world, even the data logger 262 with thermometer screen, the air temperature will also be affected the shortwave and diffuse 263 radiation. In such situation, ENVI-met will underestimate the real weather conditions. In this study,

267
Regarding the value of RMSE=2.12℃ of point-5, slightly higher than other points, it's also considered 268 as valid in previous study [38].    Figure 1 The location of the Mi Zhi city. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 2
The climate zone of the Mi Zhi city Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 3
The ancient cave dwelling settlement in Mi Zhi city The existing dwellings in Mi Zhi city Figure 5 The schematic overview of the ENVI-met model Figure 6 The selected points in the on-site measurement The detailed information of the selected points  The simulated results of the selected points at 9:00am on January,31th, 2019

Figure 10
The comparison between measured and simulated data on January31th Figure 11 The comparison between measured and simulated data on February 1st Figure 12 The comparison between measured and simulated air temperature in outdoor space Figure 13 The RMSE between measured and simulated data in courtyard Figure 14 The correlation and RMSE between measured and simulated data in courtyard