Study of thermal comfort in the residents of different climatic regions of India—Effect of the COVID‐19 lockdown - Thapa - Indoor Air

Abbreviations

APV: air preference vote
ASV: air sensation vote
df: degree of freedom
HPV: humidity preference
HSV: humidity sensation vote
I_cl: clothing insulation (clo)
N: sample size
NV: naturally ventilated
RH: relative humidity (%)
s.d.: standard deviation
t: time (day)
T_out: outdoor mean air temperature (°C)
TP: thermal preference
T_rm(t): running mean of outdoor temperature (°C)
TSV: thermal sensation vote
T_t–n: mean temperature on n^th day before the day of observation (°C)
α: running mean constant
χ²: chi‐square

Practical implications

This paper focuses on the difference caused in thermal comfort conditions of sensation and preference and few other non‐thermal comfort factors supposed to be affected by the stressful condition of forced lockdown in the three different climatic regions of India. The findings from the present paper are helpful in developing protocols for designing the built environment where subjects live in a stressful condition, like a hospital.

1 INTRODUCTION

The emergence of COVID‐19 pandemic has changed the normal way of life in almost all the nations around the world. Though, the pandemic has still not come to an end and is taking life across the world, it has forced people to live indoors with an increase in the practice of “online work” and “work from home” culture across major economies.

This major shift in the work culture has led to a decrease in the air pollution both in India ¹ and across the world.² Most economies of the world underwent a lockdown, either partial or total, which was considered to help containment of the disease and prevention from its spread. Figueiredo et al ³ based on a study regarding incidence and mortality rates in Hubei and Guangdong report that the strict social distancing measure is effective in reducing the infection and mortality rates. The decrease in the office hours or commercial establishments and limitations posed on vehicular movement led to the decrease in both the energy demand and greenhouse gas emissions. Even though the occupants used to spend a large proportion of their time in the indoor environment of their work place before the lockdown had occurred, the forced lockdown indoors with restrictions to move freely affect the medical, physiological, and psychological well‐being ⁴ of a person, all of which can affect the ability of an individual to adapt to the environmental condition.

The studies on thermal comfort, both climate chambers‐based and field studies based, have gained much interest today, for effectively designing of buildings that are comfortable and energy conscious. The thermal comfort is “that condition of mind which expresses satisfaction with the thermal environment”.⁵ Secondly, the adaptive model of thermal comfort is based on the principle that, “if a change occurs so as to produce discomfort, occupants react in ways which tries to restore their comfort”.⁶ Subjects often utilize the available measures to make them more comfortable. In a restrictive environment, which poses constraints due to dress code, activity level and ability to move freely from one’s work space like in the case of institutional buildings or offices make this choice of available adaptive measures, limited.⁷ However in the case of residential buildings, subjects often exercise one or all of the measures available to them like, changing the posture, clothing level, access to windows or fresh air, moving to warmer or cooler areas, etc in order to make themselves more comfortable.⁸ It is seen from previous researches that, subjects often move toward sunny areas during the cold climate and take sun bask,⁹ whereas during the hot summers of outdoor environment people look for shade.¹⁰

However, even though residential houses may offer a higher degree of adaptivity to the occupants in comparison with the offices and institutional buildings, a continued and forced indoor lockdown can have physiological and psychological impacts, which ultimately affects the degree of comfort and overall satisfaction of the indoor environment. Thus, the forced lockdown imposed by several governments across the globe in order to combat the rise of COVID‐19 pandemic and the restrictions posed on the residents to stay indoors in their houses can pose constraints on the adaptation of the occupants. This ultimately affects the thermal comfort of an individual.

In an effort to contain the disease, the Indian government also announced the nationwide lockdown starting from the March 22, 2020. After a complete lockdown of 71 days, relaxation and phased unlocking began from 31st of May onwards and is still taking place at the time of writing this paper. Apart from the severe impacts caused on the migrant labor ¹¹ and economy,¹² the lockdown gave the government the much required time to update the health care facility.

Several field studies based research have been conducted in both the air‐conditioned buildings and free‐running buildings both across the globe ^{13, 14} and in India.^{15, 16} The ASHRAE Standard 55 ⁵ has included the adaptive model of thermal comfort based on the studies of de Dear and Brager.¹⁷ Despite all, the thermal comfort standards in India are not properly defined yet. A range of 23 ‐ 26°C and 21 ‐ 23°C for summers and winters, respectively, is recommended by the National Building Code ¹⁸ for all climatic regions and all building types. However, owing to novelty of the pandemic, researches on built environment and thermal comfort during such lockdown are scarce. ISHRAE ¹⁹ provides the guidelines regarding the procedure to operate air‐conditioning and ventilation systems which otherwise can be a driving force for the spread of the virus. For example, a lower relative humidity (RH) can make the lungs drier which lowers the body’s ability to fight against infections,¹⁹ whereas a higher RH can interfere with the evaporative heat loss causing hot and humid condition and thereby affect thermal comfort.

Previous studies have highlighted that people at their home and at office respond differently regarding comfort assessment even at an identical level of thermal conditions and clothing, which suggest the dependency of thermal comfort to factors like social context.²⁰ However, Humphreys et al ²¹ report that the difference could have arose due to the fact that it was difficult to precisely say whether the metabolic rates, both at residence and at office, were similar or not.

In this paper, we present the status of thermal comfort among the indoor occupants of three different climatic regions in India, that is, cold climate, composite climate, and hot and humid climate, who were forced to stay inside their houses due to the COVID‐19 lockdown. Previously, several field survey‐based studies were conducted in these Indian conditions of cold climate,^{10, 22} composite climate, ^{23, 24} and hot and humid climate,^{15, 25} separately.

India, being a tropical country the cold climatic region, is limited to the elevated regions, which are mostly along the Himalayan areas of the north and NE India (Figure 1). In the cold climate and elevated region of eastern Himalayas, Thapa et al ^{10, 26} had reported that the thermal sensation of the subjects varied significantly with the elevation. They also reported that the indoor comfort temperature of the subjects decreased with the increase in elevation of the location from the mean sea level. Similar findings were also previously reported by Rijal et al.⁹ Cena et al ²⁷ also had reported an inverse relation between an overnight mean clothing insulation and the air temperature inside the tents in Karakoram Range. The subjects had adjusted their thermal insulation throughout the night so as to maintain a thermal sensation which was close to neutral.

Locations in the climate map of India (Source: NBC 18)

The composite climate in India is typically marked by hot summers and cold winters.¹⁸ In the composite climate of Jaipur, Dhaka et al ²³ had found a neutral temperature of 25.6°C, 27.0°C, and 29.4°C during winter, moderate, and summer seasons, respectively. They had used the data obtained from naturally ventilated (NV) residential and office buildings. They also reported that the air movement was crucial toward adaptation in the hot condition and the subjects preferred for a higher air movement during these hot conditions.

The eastern, SE, and SW parts of the country are represented by hot and humid climate, which is primarily due to their proximity with the Bay of Bengal and the Arabian Sea. The region is represented by high temperature and humidity during the summer months. The humidity peaks especially with the onset of monsoons.¹⁸ These conditions often make it difficult to achieve thermal comfort, as the ability to lose heat by the body in hot conditions is reduced, owing to a high RH.²⁵ Several adaptive mechanisms are employed by the occupants in this climate, behavioral like taking a siesta in the hot afternoon, which reduces the body activity level ²⁸ or even physiological adaptation like the perspiration pattern.²⁹ Houses bearing cross‐ventilation strategies are often helpful in hot and humid climate.^{30, 31}

1.1 Objectives of the study

Very few studies have previously dealt specifically on residential houses with restrictions to move outdoors. In this paper, we aim to study the thermal comfort status of the indoor occupants in their residences during the restrictions posed against moving outdoors due to COVID‐19 lockdown in India. An online survey was conducted to explore the thermal comfort condition of the occupants who were restricted in indoor environment in their residences during the lockdown in India arising due to COVID‐19. The paper explores the variations in clothing insulation, thermal sensation and preference, variation of window opening behavior, and operation of fans. In addition to the thermal comfort factors, we also studied some non‐thermal comfort factors like, self‐judged productivity, desire to go outdoors, gain or loss in weight, and effectiveness of working from home.

2 THE RESEARCH DESIGN

In order to understand the status of thermal comfort in these restrictive conditions, a Google questionnaire form was prepared consisting of 4 sections. The first section contained the identifiers of date, time, place, gender, age, and profession. This section also had a question about the approximate number of hours spent per week by the subjects in the indoor environment of their workplace before the lockdown. The section further contained question on the type of building the subjects lived in, that is, whether the house was multifamily or single family. Also, the general built form of the building in which the respondents lived in was asked, that is, concrete walls with concrete roofs, concrete walls with tinned roofs, and wooden wall with tinned roofs, respectively.

In the second section, questions regarding the thermal comfort responses were asked. The sensation votes regarding the surrounding temperature, relative humidity (RH), and air movement were asked on a 7‐point ASHRAE scale, while the preference votes for the temperature, RH, and air movement were asked on a 5‐point Nicol’s scale. Table 1 shows the scales of the thermal sensation votes (TSV), thermal preference votes (TP), humidity sensation votes (HSV), humidity preference votes (HPV), air sensation votes (ASV), and air preference votes (APV) as used in the survey.

Table 1.
Scales used for assessment of thermal comfort and other parameters in the study

Scale used	TSV	TP	HSV	HPV	ASV	APV	Productivity	Indoor comfort	Change in lunch time / bed time	Desire going outdoors	Efficiency working from home
‐5								Very uncomfortable
‐3	Cold		Very dry		Very low			Moderately uncomfortable
‐2	Cool	Much warmer	Dry	Much moist	Low	Much higher				Not at all desirable	Not at all effective
‐1	Slightly cool	A bit more warm	Slightly dry	A bit more moist	Slightly low	A bit higher	Lower than as usual	Slightly uncomfortable	Early	Not so desirable	Not so effective
0	Neutral	No change	Neutral	No change	Neutral	No change	As usual		On time	Somewhat desirable	Somewhat effective
1	Slightly warm	A bit more cool	Slightly moist	A bit more dry	Slightly high	A bit lower	Higher than as usual	Slightly comfortable	Late	Very desirable	Very effective
2	Warm	Much cooler	Moist	Much dry	High	Much lower				Extremely desirable	Extremely effective
3	Hot		Very moist		Very high			Moderately comfortable
5								Very comfortable

Abbreviations: APV, air preference votes; ASV, air sensation votes and; HPV, humidity preference votes; HSV, humidity sensation votes; TP, thermal preference votes; TSV, thermal sensation votes.

In the second section, questions regarding self‐judged productivity were also asked on a 3‐point scale of higher than usual, as usual, and lower than usual. The indoor comfort was asked on a 6‐point scale of very comfortable, comfortable, slightly comfortable, slightly uncomfortable, uncomfortable, and very uncomfortable. In the same section, two checklists were provided for the respondents to give their response, that is, first containing a list of the clothing that the subjects were wearing during the filling up of questionnaire and second was the list of activities that the subjects were engaged in the last 15 minutes.

The third section of the questionnaire had questions on environmental descriptions. Since this was an online transverse survey, where it was not possible for the researcher to attend the respondent physically and measure the environmental parameters as prescribed in the ASHRAE Standard 55 ⁵ Class II protocol, self‐judged probable environmental temperature and RH were asked from the respondents. The number of windows that were open in the room at the time of filling up of the response form was asked, next. The presence and absence of fans and the status of running fan with speed in a scale of low speed, medium speed, and high speed were also asked. Also, the presence and absence of room heaters were asked.

In the final section of the questionnaire, physiological description was asked. This included the number of hours the subjects slept before lockdown and presently (ie, during the lockdown). Also, the change in lunch time and bed time was asked in a three‐point scale of early, on time and late. The self‐judged perception of whether the respondent had gained or lost weight during the lockdown was asked in a four‐point scale as, have gained weight, neither gained nor lost weight, have lost weight, and no idea. The desire to go outdoors was asked in a five‐point scale, extremely desirable, very desirable, somewhat desirable, not so desirable, and not at all desirable. Finally, the effectiveness of using internet to work from home was asked in a five‐point scale, extremely effective, very effective, somewhat effective, not so effective, and not at all effective.

A total of 406 filled in responses were received from 16 cities across India between 48th and the 71st day of the nationwide lockdown in India. These cities are marked in climate map of India as shown in Figure 1. In order to have a substantial number of dataset in each group, these cities were grouped according to their location in the climate map of India. Thus, a total of 109, 109, and 188 numbers of filled responses were received from the subjects of cold climate, composite climate, and hot and humid climate, respectively.

The outdoor daily maximum and minimum temperature for the lockdown period and a period of 1 month before the lockdown were obtained from the website www.ogimet.com for each city involved in the study. The daily mean outdoor air temperature for each city was calculated as the average of maximum and minimum outdoor air temperature for the day. The average of outdoor mean air temperature for the three different climatic locations under study is shown in Figure 2A.

Daily mean of (A) outdoor air temperature, Tout (°C) and (B) relative humidity, RH (%) in the different climatic regions

The physiological thermal adaptation of individual in an environment depends largely upon the history of past temperature experience.²¹ Nicol and Humphreys ⁶ had earlier proposed the exponentially weighted running mean of the outdoor mean temperature (Equation (1)) to include this time dependence of the adaptation of the subjects.

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0001$ (1)

In the above Equation (1), T_rm(t) is the exponentially weighted running mean temperature (°C) for a time interval of “t”, which can be in hours, days, etc, and T_{t – n} is the instantaneous temperature at the n^th time interval previously. The term, “α” is a constant such that 0 > α ≥ 1. Thus, a higher value of α suggests a stronger effect of past temperature on the thermal adaptation of an individual. McCartney and Nicol ¹⁴ had illustrated that a higher value of “α” does not affect the correlation between the comfort temperature and outdoor air temperature, till α ≤ 0.90, after which this correlation decreases. Thus, previous studies ^{14, 32} have recommended a use of α = 0.80 in adaptive thermal comfort studies. As per Humphreys,³³ who had previously identified the “running mean” as similar to half‐life decay in nuclear physics and medicine, stated that the half‐life period of a particular running mean temperature is approximately 0.69/(1 – α). Thus, the larger the value of α, the longer is the “half‐life” or reaction of the running mean temperature to a change in outside weather condition.¹⁴

The indoor environmental parameters of air temperature, mean radiant temperature, air movement, and relative humidity could not be measured due to the online nature of the study among dispersed individuals and the unavailability of accurate and uniform instrumentation in each case. Therefore, we could not calculate the comfort temperature (T_comf) for the study. As a result, the relations of comfort temperature with the different parameters for adaptation, like outdoor air temperature and clothing insulation, could not be obtained in the present study.

The clothing insulation and activity level were, however, obtained from the responses and were calculated in “clo” and “met” as per the ASHRAE Standard 55.⁵

The responses as received were extracted in Comma‐separated values (CSV) format and converted into Microsoft Excel spreadsheets. The automatically generated timestamp column was used to identify the date and time for filling up of the survey form. This was done to ensure uniformity. The descriptive responses as received were denoted the scale as described in Table 1, before the final analysis using SPSS version 25. The “Descriptive Statistics” tab present in the SPSS version 25 was used to find the measures of the central tendency, that is, mean value and standard deviation for each of the parameters. In order to test the null hypothesis that the mean values for the concerned thermal comfort parameter for each of the three different groups, that is, cold climate, composite climate, and the hot and humid climate (and denoted with the numeric values 1, 2, and 3 in SPSS version 25) are equal, we performed a non‐parametric test. Unlike, a parametric test, the non‐parametric test does not require the data to be from a normal distribution. So, in the K‐Independent Samples of the Non‐parametric Tests tab in SPSS version 25, we took each of the test variables and grouped them as per the climatic distribution (from 1 to 3 in the grouping variable). The Kruskal‐Wallis test, also called the one‐way ANOVA (analysis of variance), which compares two or more independent samples and which can take input having either equal or even different sample size was performed when we had to test the variable between different groups (ie, climate groups). Variables which had to be analyzed in reference to the lockdown period, like change in the lunch time and bed time, self‐perceived gain or loss of weight, and desire to go outdoors, were tested using the non‐parametric test, Kolmogorov‐Smirnov (KS) test. This test measures the one‐dimensional probability distribution, which can be used to compare the sample with a reference probability distribution.

3 RESULTS AND DISCUSSIONS

In order to understand the effect of COVID‐19 lockdown on the thermal comfort of the residents in different climatic conditions of India, the data that were collected via an online study were analyzed and compared in terms of different climate conditions. At the same time for the non‐thermal comfort parameters like change in the lunch time and bed time, self‐perceived weight gain or loss, and desire to go outdoors that were considered to be the indicators of the effect of lockdown, the comparison was made between the present level and that before the start of the lockdown. In the present section, we first discuss the parameters which have an effect upon the thermal comfort of individuals, that is, the outdoor environmental conditions, which include the outdoor air temperature (T_out) and relative humidity (RH), the details of respondents, which include their age, gender, the type of house they lived in, and the clothing insulation. This is followed by the discussion on thermal comfort, that is, thermal sensation and preference, humidity sensation and preference, air sensation, and preference. The discussion on self‐judged productivity and indoor comfort is done next followed by the occupant behavior of windows opening and fans running. Lastly, the effect of lockdown on some non‐thermal comfort parameters like change in lunch time and bed time, self‐perceived weight gain, or loss is discussed.

3.1 Outdoor environmental condition

The average daily mean outdoor air temperature in the different climatic regions for the second half of lockdown period between April 13, to May 31, 2020 in India, during which the study was conducted, are shown in Figure 2A. The mean outdoor air temperature was 16.8°C (s.d. 2.16°C, N 50), 27.5°C (s.d 1.53°C, N 50), and 31.1°C (s.d. 2.53°C, N 50) in the cold climate region, hot and humid climate, and composite climate, respectively. This difference in mean T_out between the different climatic regions was statistically significant, Kruskal‐Wallis χ² (df 2, N 150): 120.393, p < .001 with a mean rank of 25.50, 80.57, and 120.43 in the mean T_out between cold climate, hot and humid climate, and composite climate, respectively. Thus, the outdoor air temperature in composite climate was the highest followed next by that in hot and humid climate and cold climate, respectively.

The average of daily mean outdoor relative humidity (RH %) for the same period in the different climatic regions under investigation is shown in Figure 2B. The mean RH was 83.1% (s.d. 6.69%, N 50), 73.5% (s.d. 10.88%, N 50), and 52.1% (s.d. 9.61%, N 50) in the cold climate, hot and humid climate, and composite climate, respectively. This difference in mean RH % between the different climatic regions was statistically significant, Kruskal‐Wallis χ² (df 2, N 150): 95.323, p < .001 with a mean rank of 113.27, 83.67, and 29.61 for the mean RH % in cold climate, hot and humid climate, and composite climate, respectively. A high RH in the cold climate region is primarily due to the low temperature which decreases the maximum partial vapor pressure of water vapor in air.¹⁰

3.2 Respondents, their workplaces, and their residences

A total of 406 responses were received, 218 (ie, 53.7%) from female, and 185 (ie, 45.6%) from male, while 3 (0.7%) respondents did not prefer to reveal their gender. The minimum and maximum age varied between 17 years and 64 years, with mean age of 25.04 years (s.d. 9.158 years), respectively. Figure 3A shows the distribution of the age (in years) of the respondents.

Percentage variation in (A) age (in years) and (B) profession of the respondents who took part in the study

Figure 3B shows the profession of the subjects, who took part in the study. The respondents were also asked about the number of hours spent by them inside their respective workplace before the lockdown had begun. The mean number of hours spent by the subjects inside their workplace as per their occupation is shown in Figure 4.

Time spent by the respondents in their workplace per week before the lockdown

The subjects were next asked for the type of built form and the residence type in which they lived at. The percentage distribution of built form and the residence type for the different climate under investigation is illustrated in Figure 5. A total of 397 responses were received for the residence type of the respondents and 32.0% lived in a multifamily or housing complex, while 68.0% lived in single family or detached houses. A total of 406 responses were obtained for the type of built form in which the subjects lived, 83.0% lived in concrete buildings having RCC slab roof, 10.3% lived in concrete walled buildings with tinned roof, and 6.7% of the subjects lived in wooden houses with tinned roof.

Distribution of built form and residence types in different climate under investigation

3.3 Clothing insulation

The subjects were asked to select the clothing that they were wearing during the filling of the response form. The clothing insulation values in clo for the respective clothing were taken from ASHRAE Standard 55.⁵ The mean clothing insulation was 0.36 clo (s.d. 0.216 clo, N 109), 0.28 clo (s.d. 0.109 clo, N 109), and 0.27 clo (s.d. 0.117 clo, N 188) among the subjects of cold climate, composite climate, and hot and humid climate, respectively. The Kruskal‐Wallis test revealed a significant difference in the mean clothing value, Kruskal‐Wallis χ² (N 406, df 2): 12.874, p = .002, with a mean rank of 234.73, 205.39, and 184.30 for the clothing insulation among the subjects in cold climate, composite climate, and hot and humid climate, respectively. During the similar warm months, previous studies conducted in the respective climatic conditions of India reports a mean clothing insulation of 0.81 clo (s.d. 0.301, N 1354) in the cold climatic region,³⁴ 0.52 clo in the composite climate, ²³ and 0.53 clo (s.d. 0.137, N 346) in the hot and humid region,³⁴ respectively. A relatively lower clothing insulation in the present study in comparison with that in the earlier studies in the similar climatic and weather condition, calls for a further investigation of the effect of lockdown on the thermal and mental well‐being of the respondents. However, it is pertinent to mention here that the previous studies referred above were not purely conducted with residents as the subjects, as done in the present study. Figure 6 shows the variation of clothing insulation with outdoor mean air temperature in the different climatic regions investigated in this study. As expected, there was a weak negative correlation (Equations 2‐4-2‐4) between the outdoor temperature and the clothing insulation. This shows that the subjects illustrated that they used clothing as a mechanism to adapt with the changing temperature, even in that short period of study.

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0002$ (2)

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0003$ (3)

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0004$ (4)

Variation of clothing with mean outdoor air temperature

From Equations 2‐4-2‐4, it is also noticed that the decrease in clothing insulation with temperature was sharp with a higher coefficient in case in cold climate than in comparison with the other two climates under investigation. This implies, that in cold climate, people use clothing insulation as a stronger measure of adaptation than in warmer conditions, where there is a limitation on the lower limit of clothing based on cultural factors.^{9, 33}

3.4 Thermal Comfort Responses

The following observations were made.

3.4.1 Thermal sensation and preferences

The study was conducted during the early part of the summer season in India, when the monsoon had not yet begun. The mean outdoor air temperature during the study period was observed as 16.8°C (s.d. 2.16°C, N 50), 27.5°C (s.d 1.53°C, N 50), and 31.1°C (s.d. 2.53°C, N 50) in the cold climate region, hot and humid climate, and composite climate, respectively. The thermal sensation regarding the present indoor condition encountered by the subjects was obtained in ASHRAE 7‐point scale (Table 1). The distribution of thermal sensation votes (TSV) in the different climatic regions is shown in Figure 7A. In cold climate, a maximum of 43.1% responses were obtained for neutral sensation followed by 26.6% warm, 11.9% hot, 9.2% cool, 8.3% cold, and 0.9% very hot, respectively. In the composite climate, an overwhelming response of 53.2% was for warm sensation, followed by 19.3% neutral, 18.3% hot, 8.3% very hot, and 0.9% cool, respectively. In the hot and humid climate, the maximum TSV were received for hot sensation (35.1%), followed by warm (34.6%), neutral (16.5%), very hot (9.0%), cool (3.7%), cold (0.5%), and very cold (0.5%), respectively.

Distribution of (A) thermal sensation votes (TSV) and (B) thermal preference (TP) votes during the lockdown in India in different climatic region

As seen in Figure 7A, the TSV was skewed the most toward the warmer sensation in hot and humid climate and least in cold climate. The mean TSV was 0.28 (N 109, s.d. 1.088) in the cold climate, 1.14 (N 109, s.d. 0.855) in the composite climate, and 1.26 (N 188, s.d. 1.049) in the hot and humid climate respectively (Table 2). This difference in mean TSV among the subjects of different climates was statistically significant, Kruskal‐Wallis test, χ² (df 2, N 406): 59.076, p < .001, having mean rank of 133.85, 217.22, and 235.92 for cold climate, composite climate, and hot and humid climate, respectively. It was interesting to note that among the two hot climates, that is, hot and humid climate and composite climate, the outdoor mean air temperature was higher in composite climate than in hot and humid climate (Figure 2A), whereas the mean TSV was higher in hot and humid climate than in composite climate. This is due to the fact that the higher RH brings about a higher level of discomfort among the subjects of hot and humid climate (Table 2).

Table 2.
Statistics of outdoor mean temperature (T_out), relative humidity (RH), thermal sensation votes (TSV), thermal preference (TP), humidity sensation votes (HSV), humidity preference votes (HPV), air sensation votes (ASV), and air preference votes (APV) in the different climatic conditions under investigation

Climate	Parameters		TSV	TP	HSV	HPV	ASV	APV
Cold climate	mean		0.28	‐0.33	0.04	‐0.05	‐0.11	0.01
	standard deviation (s.d.)		1.088	0.882	1.217	0.821	1.092	0.739
	sample size (N)		109	109	109	109	109	109
	Kruskal‐Wallis rank		133.85	158.23	225.46	212.56	229.19	242.67
Composite climate	mean		1.14	0.55	‐0.80	‐0.02	‐0.47	‐0.33
	standard deviation (s.d.)		0.855	1.014	1.034	0.561	1.214	0.639
	sample size (N)		109	109	109	109	109	109
	Kruskal‐Wallis rank		217.22	259.08	148.12	215.22	193.09	197.14
Hot and humid climate	mean		1.26	‐0.01	0.05	‐0.21	‐0.48	‐0.38
	standard deviation (s.d.)		1.049	1.039	1.423	0.910	1.252	0.829
	sample size (N)		188	188	188	188	188	188
	Kruskal‐Wallis rank		235.92	197.52	222.88	191.45	194.64	184.48
	Kruskal‐Wallis test statistics	chi‐square (χ²)	59.076	45.237	35.295	4.364	7.682	20.431
		degree of freedom (df)	2	2	2	2	2	2
		p value	0.000	0.000	0.000	0.033	0.021	0.000
Total data	mean		0.96	0.05	‐0.18	‐0.11	‐0.38	‐0.26
	standard deviation (s.d.)		1.092	1.043	1.325	0.809	1.209	0.774
	sample size (N)		406	406	406	406	406	406

It was interesting to note that in the cold climate, the mean TSV for male respondents was higher than that in the female respondents, while for the composite climate and hot and humid climate, it was the opposite. The mean TSV was 0.40 (N 40, s.d. 1.194) for male and 0.20 (N 69, s.d. 1.023) for female respondents in the cold climate, 1.12 (N 67, s.d. 0.749) for male and 1.17 (N 42, s.d. 1.010) for female respondents in composite climate and 1.21 (N 78, s.d. 1.073) for male and 1.24 (N 107, s.d. 1.008) for female respondents in the hot and humid climate, respectively. This difference in mean TSV between the two genders in different climatic condition was not found to be statistically significant, which could be due to the fewer number of responses in each category.

The TSV of the subjects showed a moderately positive but statistically significant correlation with the outdoor air temperature, R = 0.429 (N 406, p < .001). Figure 8 shows the variation of mean TSV with the mean T_out during the period of investigation.

Variation of TSV with the outdoor mean air temperature, Tout

The thermal preference (TP) votes regarding the indoor environment were recorded in 5‐point Nicol’s scale (Table 1). The distribution of TP in the different climatic regions is shown in Figure 7B. In cold climate, maximum preference was for a bit warmer (45.9%), followed by no change (28.4%), a bit more cooler (20.2%), much warmer (4.6%), and much cooler (0.9%), respectively. In the composite climate, an overwhelming responses were for a bit more cooler (57.8%), followed by no change (14.7%), a bit more warmer (11.9%), much cooler (10.1%), and much warmer (5.5%), respectively. In hot and humid climate, the maximum preference was for a bit more cooler (38.3%), followed by no change (27.7%), a bit more warmer (22.3%), much warmer (10.1%), and much cooler (1.6%), respectively.

As seen from Figure 7b, the mean TP in cold climatic region were more toward a warmer preference (mean −0.33, N 109, s.d. 0.882), while that in the composite climate was for a cooler preference (mean 0.55, N 109, s.d. 1.014) and that in the hot and humid climate was close to no change (mean −0.01, N 188, s.d. 1.039), respectively. This difference in mean TP in the different climatic condition was statistically significant, Kruskal‐Wallis test, χ² (df 2, N 406): 45.237, p < .001, having mean rank of 158.23, 259.08, and 197.52 for cold climate, composite climate, and hot and humid climate (Table 2), respectively.

The TP votes for the male subjects were lower than that for female subjects in cold climate and composite climate, while it was the opposite in the hot and humid climate. The mean TP was −0.30 (N 69, s.d. 0.863) for female subjects and −0.37 (N 40, s.d. 0.925) for male subjects in cold climate, 0.57 (N 42, s.d. 1.016) for female subjects and 0.54 (N 67, s.d. 1.020) for male subjects in composite climate, and −0.19 (N 107, s.d. 1.083) for female subjects and 0.19 (N 78, s.d. 0.927) for male subjects in hot and humid climate, respectively. Though the difference in mean TP between male (mean 0.19, s.d. 1.013, N 185) and female (mean −0.08, s.d. 1.051, N 218) in the total data was statistically significant, t = −2.637 (df = 401, p < .05), this difference in mean TP between the two genders was not statistically significant in different climatic groups, separately. This is due to the fewer number of responses in each group.

3.4.2 Humidity sensation and preference

The humidity sensation votes (HSV) which were received in a 7‐point scale (Table 1) are shown in Figure 9A. In the cold climate, the maximum responses of 33.0% were for neutral, followed by 22.9% slightly moist, 17.4% slightly dry, 13.8% dry, and 12.8% moist, respectively. In the hot and humid climate, the maximum responses of 37.2% were for neutral, followed by 16.0% moist, 15.4% slightly dry, 13.3% dry, 11.7% slightly moist, 3.7% very moist, and 2.7% very dry, respectively. In the composite climate, the maximum responses of 41.3% were for slightly dry, 30.3% neutral, 17.4% dry, 4.6% very dry, 3.7% moist, and 2.8% slightly moist, respectively.

Distribution of (A) humidity sensation votes (HSV), (B) humidity preference votes (HPV), (C) air sensation votes (ASV), and (d) air preference votes (APV) in the different climatic region studied

It was interesting to note that in the composite climate only, the HSV were more skewed toward the drier sensation. The mean HSV was 0.04 (N 109, s.d. 1.217) in cold climate, 0.05 (N 188, s.d. 1.423) in hot and humid climate, and −0.80 (N 109, s.d. 1.034) in composite climate, respectively. This difference in mean HSV was statistically significant, Kruskal‐Wallis test, χ² (df 2, N 406): 35.295, p < .001, with a mean rank of 148.12, 222.88, and 225.46 in composite climate, hot and humid climate, and cold climate, respectively. Thus, it is seen that the subjects responded to a drier environmental sensation in composite climate than the other two investigated climates. However, it is to be noted that majority of the responses for the composite climate were received from the western part of India, which receives scanty annual rainfall and is drier, while almost all of the responses for cold climate and hot and humid climate were obtained from the eastern part of the country is receives a higher annual rainfall and is wetter. This is also seen from Figure 2B, where the outdoor RH in composite climate is significantly lower than that in the other two climatic regions.

In the composite climate, the preference votes for RH were overwhelmingly for a no change sensation, whereas the preference for a bit more moist condition in the hot and humid climate reflects that the subjects in these regions are accustomed with a higher humidity. The distribution of humidity preference votes (HPV) is shown in Figure 9B. The mean HPV was −0.05 (N 109, s.d. 0.891) in cold climate, −0.02 (N 109, s.d. 0.561) in composite climate, and −0.21 (N 188, s.d. 0.910) in the hot and humid climate, respectively. This difference in mean HPV between the different climatic regions was statistically significant, Kruskal‐Wallis test, χ² (df 2, N 406): 4.364, p < .001 with a mean rank of 191.45, 212.56, and 215.22 for hot and humid climate, cold climate, and composite climate, respectively. This higher preference for a moist environmental condition in the hot and humid climate than in the two other investigated climates, that is, composite climate and cold climate is due the fact that the subjects in hot and humid climate are accustomed toward a wetter condition, which requires a further analysis as till now the same were viewed only in terms of non‐thermal factors by researchers.

3.4.3 Sensation and preference of air movement

The sensation for air movement, that is, air sensation votes (ASV), was obtained in a 7‐point scale and is shown in Figure 9C. In the cold climate, 43.1% of the responses were for neutral, followed by 20.2% as slightly low, 17.4% slightly high, 11.9% as low, 6.4% as high, and 0.9% as very high, respectively. In the composite climate, slightly low sensation got the highest response (33.0%), followed by neutral (32.1%), low (12.8%), slightly high (9.2%), high (8.3%), and very low (4.6%), respectively. In the hot and humid climate, 34.6% responses were obtained for neutral, followed by 21.8% slightly low, 19.7% low, 15.4% slightly high, 4.3% very low, 2.7% high, and 1.6% very high, respectively.

Table 2 also shows the mean ASV in the different climatic regions. The mean ASV was −0.11 (N 109, s.d. 1.092) in cold climate, −0.47 (N 109, s.d. 1.214) in the composite climate, and −0.48 (N 188, s.d. 1.252) in the hot and humid climate, respectively. This difference was statistically significant, χ² (df 2, N 406): 7.682, p < .05, with a mean rank of 229.19, 193.09, and 194.64 in composite climate, hot and humid climate, and cold climate, respectively.

It was interesting to note that the mean preference for a higher air movement was noticed in composite climate, −0.33 (s.d. 0.639, N 109) and hot and humid climate, −0.38 (s.d. 0.829, N 188), whereas the mean preference was close to “no change” in cold climate, 0.01 (s.d. 0.739, N 109). This difference in mean preference of air movement between the different climatic regions was statistically significant, Kruskal‐Wallis, χ² (df 2, N 406): 20.431, p < .001 with a mean rank of 242.67, 197.14, and 184.48 between mean values of TPV in the cold climate, composite climate, and hot and humid climate, respectively. Figure 9D illustrates the percentage response in the APV in the three different climatic regions under investigation. It is clearly seen in the figure that the preference of the higher air movement was the most in hot and humid regions and the least in cold climate regions. This shows the fact that in a hot and humid climate, air movement is an effective way of combating the discomfort,²⁸ whereas in cold climate, a higher air movement can cause chilling effect and thus not preferred for.²⁶

Distribution of productivity and comfort votes of the subjects

3.5 Productivity and comfort

Self‐judged productivity was obtained in a three‐point scale (Table 1), that is, lower than usual (−1), as usual (0) and higher than usual (+1), respectively. A total of 403 responses were received of which 47.4% of the respondents voted that their self‐judged productivity were as usual, while 41.2% voted below usual and only 11.4% of the responses were higher than as usual (Figure 10A). While the indoor comfort were asked in a six‐point scale (Table 1), a total of 406 responses were received for the indoor comfort, 32.8% voted for moderately comfortable, 26.4% voted for slightly comfortable, 22.9% voted for very comfortable, 10.6% voted for slightly uncomfortable, 4.7% voted for moderately uncomfortable, and 2.7% voted for very uncomfortable (Figure 10B), respectively.

Though this was not a longitudinal study, where the same subject is studied for an elongated period of time, it was still interesting to note that both the self‐judged productivity and the comfort votes declined with the increase in the number of days inside lockdown (Figure 11). A weakly negative trend was seen between the self‐judged productivity and number of days (R =−0.046, N 403), which, however, was not statistically significant. Similarly, a weak negative trend was seen between the indoor comfort and the number of days of lockdown (R = ‐ 0.213, N 406, p 5).

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0005$ (5)

Variation of self‐judged productivity and comfort with the increasing day of lockdown

It is understood that the lockdown, which affects the subjects psychologically, has a high impact to this general decline of the productivity and comfort votes with the increasing day of lockdown, causing restrictions in outdoor movement.

3.6 Occupant Behavior

Behavioral adaptation is an important part of human adaptation.³¹ In naturally ventilated buildings, apart from the personal adjustments in clothing and activity level, opening and closing of windows to let the fresh air in and switching on and off of fans for increasing the evaporative cooling holds a significant place ³¹ for providing thermal comfort.

3.6.1 Windows

In a NV building, the occupants open the windows in order to improve the air quality of indoor environment and to improve the comfort by reducing indoor temperature encouraged by air movement.³² In the present study, the subjects were asked about the number of windows that were in open state during the filling up of survey form. Figure 12A represents a scatter diagram showing the number of windows kept open during the investigation period against mean outdoor air temperature and T_out (°C) for the day of observation. A quadratic equation shows that the number of openings increases first with the increase in ambient temperature, reaches the maximum, and decreases. Equation (6) shows the quadratic variation of mean number of open windows with the daily mean outdoor air temperature.

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0006$ (6)

Distribution of open windows at (A) different outdoor air temperature, Tout (°C) and (B) different time of the day (dark circles in figure (A) represent the actual no. of open windows)

It is noticed from Figure 12A that at high outdoor air temperature (T_out), the number of windows kept open decreases, which is probably a measure to prevent the heat flux from reaching the indoor environment. This hypothesis is further supported by the fact that in Figure 12B, the mean numbers of windows kept open in each climatic condition were least during the afternoon. Such lower no of windows opening during the afternoon hours of summer season was previously seen in the study of Daaboul et al.³⁵ It is further noticed from Figure 12B that the maximum numbers of windows kept in open condition during the study were in hot and humid region, which help in cross‐ventilation strategies in these high humid areas. The numbers of windows kept open in composite climate were the minimum. A reason could be that due to the higher temperature encountered in composite climate than the other two climatic conditions studied (Figure 2A), this helps to prevent the heat flux from reaching the indoor environment. Secondly, the window to wall ratio in the houses of composite climate could differ from those in hot and humid climate and cold climate, as the former locations are mostly on the western part of the country, while the latter two are on the eastern part. This difference in the housing design in different climatic locations within India needs a further research.

3.6.2 Fans

Fans are often used in warm climates to combat with the effects of high indoor operative temperature. In one of the earliest studies of field surveys, Nicol ³⁶ had found that subjects were comfortable in a temperature above 30°C due to various adaptations like the use of ceiling fans, which compensates for as much as 4°C rise in indoor operative temperature.⁵ It is pertinent to mention here that as per Fanger PMV‐PPD,³⁷ any temperature above 30°C would have been impossible to be comfortable at. Fans facilitate forced convection, which increases the convective heat transfer coefficient between the skin and the air.³⁸ This is favorable for heat loss from the body and thereby facilitates cooling. Secondly, in humid conditions which impede evaporative heat loss, movement of air over skin replaces the saturated air thereby increasing the opportunity for evaporative heat loss.³⁹

In order to estimate the proportion of running of fans, logistic regression analysis which returns outputs in binary variables has been conducted in literature.³² The output with “off” state of fans is represented by 0 and “on” state of fans by 1, with the cutoff point above 0.5. The probability of fans running (p) with the outdoor mean air temperature is expressed as a logit function, as given Equations 7‐8-7‐8 below.

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0007$ (7)

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0008$ (8)

In the equations above, “b” and “c” are the constants usually determined by performing a probit regression for the Logit function, with the given conditions that the fans are open.^{31, 40} The variation of fans running with the outdoor air temperature (T_out) is illustrated in Figure 13. We performed the logistic regression for the proportion of fans running with the outdoor mean air temperature, T_out as shown in Equation (9) below.

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0009$ (9)

Proportion (p) of fans running with the variation in outdoor mean air temperature, Tout (°C)

The high Negelkerke R² in Equation (9) shows a strong adaptation of subjects with the changing temperature and they use fans as an effective measure in warm conditions (Table 3). Figure 13 also illustrates the previous relations regarding the opening of fans given by Thapa and Indraganti ³⁴ for hot and humid climate of north Bengal in India and Rijal ⁴⁰ for the Japanese residents. Unlike Rijal,⁴⁰ it is seen in both the present and earlier studies of the author ³⁴ conducted in the same hot and humid locations, the coefficient of correlation is high which suggest that the subjects use fans as a strong measure to combat heat. It is seen from Figure 13 that at mean outdoor air temperature (T_out) of 30.0°C, an approximately 95.6% of the fans were found running in the present study. However, at the same temperature, 71.8% of the fans were found running as per Thapa and Indraganti ³⁴ and 57.0% as per Rijal.⁴⁰ It is pertinent to mention here that Thapa and Indraganti ³⁴ had included the hot and humid climate of Siliguri, which was also included in the present study (Figure 1). A possible reason for this difference could be that the previous study of Thapa and Indraganti ³⁴ was conducted in a college campus, where as the present study in the residences, even though majority of the respondents even in the present study were the college students (Figure 3B). A second reason could be the effect of continuous and forced indoor stay due to the COVID‐19 lockdown as a result the occupants desired for an increased quality of fresh air. This, however, needs further investigation. Also, a large scatter in the actual proportion of fans running in the present study was seen in Figure 13. This could be due to the fact that in Figure 13, the fans running state of two different climatic conditions, that is, hot and humid climate and composite climate, thus having subjects with different degree of thermal adaptation are represented.

Table 3.
Classification table for binary logistic model used to predict the “switching on” and “switching off” of fans with outdoor mean air temperature (T_out)

	Predicted fan status			Percentage accuracy
		Off	On	Percentage accuracy
Observed Fan Status	Off	92	29	76.0%
	On	16	269	94.4%
Overall percentage accuracy				88.9%

3.7 Effect of COVID‐19 lockdown on few non‐thermal comfort factors of the subjects

Though the thermal comfort conditions of the occupants of NV residential buildings during the restricted movement of the outdoors were the primary objective of the research, we also analyzed some of the non‐thermal comfort factors among the subjects which were highly likely to be affected by the change in the lifestyle, caused due to the COVID‐19 lockdown.

3.7.1 Change in the number of sleeping hours

The respondents were asked for the number of hours they used to sleep before the onset of the lockdown and during the lockdown. This was done in a 6‐point scale, 1 (below 4 hrs), 2 (4 ‐ 6 hrs), 3 (6 ‐ 8 hrs), 4 (8 ‐ 10 hrs), 5 (10 ‐ 12 hrs), and 6 (over 12 hrs), respectively. Figure 14 illustrates the percentage responses by the respondents regarding their number of hours of their sleep before and during the lockdown. It was seen for the period before the lockdown 64.3% responded for 6 ‐ 8 hrs, 19.5% for 4 ‐ 6 hrs, 9.4% for 8 ‐ 10 hrs, 3.4% for below 4 hrs, 2.2% for 10 ‐ 12 hrs, and 1.2% for over 12 hrs of sleep in a day. Whereas for the period during the lockdown, 39.9% responded for 6 ‐ 8 hrs, 31.3% for 8 ‐ 10 hrs, 13.5% for 10 ‐ 12 hrs, 6.9% for 4 ‐ 6 hrs, 5.9% for below 4 hrs, and 2.5% for over 12 hrs a day, respectively. Thus, a higher percentage of respondents responded that they slept for an increased number of hours during the lockdown. This finding was statistically significant, χ² (df = 25, N 406): 232.728, p < .001 (2 – tailed).

Percentage response of no of sleeping hours of the respondents (A) before and (B) during the lockdown

3.7.2 Change in the lunch time and bed time

During holidays, it is a usual trend that the subjects follow a different time for the routine activity, which can be largely due to leisure. The change in the time for lunch and time to go to bed was asked in a three‐point scale, early (−1), on time (0), and late (1), and this is illustrated in Figure 15. For the change in the time of lunch, 58.1%, 26.6%, and 13.3% of the responses were for on time, late, and early than usual, respectively. The mean response was 0.15 (s.d. 0.630, N 406). The higher response for late lunch was statistically significant with Kolmogorov‐Smirnov test having χ² (2 tailed): 0.310, p < .001. Similarly, for the change in the time to go to bed, the responses were 49.0%, 44.1% and 6.9% for late, on time, and early, respectively. The mean response was 0.42 (s.d. 0.619, N 406). The higher response rate for people going late to bed was statistically significant with Kolmogorov‐Smirnov test having χ² (2 tailed): 0.315, p < .001. Thus, it is seen that both the time to take lunch and time to go to bed was shifted significantly toward later than usual, which is understandable due to less stricter norms at home.

Change due to lockdown in the time of (A) lunch and (B) going to bed

3.7.3 Self‐perceived weight gain/loss

It is often noticed that people during holidays and at home are lesser active, and if they are not self‐aware regarding health and exercise regularly, they can gain weight. In the present study, subjects were questioned regarding their weight in a three‐point scale, (1) have gained weight, (2) neither gained nor lost weight, (3) have gained weight, and (4) no idea. Figure 16 illustrates this self‐perceived response regarding the change in their weight during the COVID‐19 lockdown in India. Of the 406 responses received, 39.2% responded that they neither gained nor lost weight, a staggering 36.9% responded that they have gained weight, 11.3% responded that they have lost weight, whereas 12.6% responded that they did not have any idea. The higher responses regarding that the subjects gained weight during the lockdown was statistically significant, Kolmogorov‐Smirnov test, χ² (2 tailed, N 355): 0.272, p < .001.

Self‐perceived response regarding weight of the subjects

3.7.4 Desire to go outdoors and efficiency of working from home

The severity of the forced lockdown indoors could have a psychological effect on the occupants. Therefore, the desire to go outdoors was asked in a five‐point scale, that is, not at all desirable (−2), not so desirable (−1), somewhat desirable (0), very desirable (1), and extremely desirable (2), respectively. The response for the desire to go outdoors is illustrated in Figure 17a. It was interesting to note that of the 406 responses received, a staggering 28.3% found extremely desirable, followed by 27.3% somewhat desirable, 20.7% not so desirable, 18.7% very desirable, and 4.9% not so desirable. This higher response for the desire to go outdoors was statistically significant, Kolmogorov‐Smirnov test, χ² (2 tailed, N 406): 0.179, p < .001.

Percentage response regarding (A) desire to go outdoors and (B) efficiency for work from home

Similarly, the ease of working from home by subjects during this lockdown was asked in a five‐point, that is, not at all effective (−2), not so effective (−1), somewhat effective (0), very effective (1), and extremely effective (2), respectively. The response regarding the ease of working from home is illustrated in Figure 17B. 38.7% of the respondents found working from home somewhat effective, 20.7% not so effective, 19.2% very effective, 13.5% extremely effective, and 7.9% not at all effective. This difference in percentage response regarding working online from was statistically significant, Kolmogorov‐Smirnov test, χ² (2 tailed, N 406): 0208, p < .001. A significant amount of subjects claiming that the working from home was not effective could be due to variety of reasons like network issues, difficulty in working with computers or even the structure of teaching learning patterns prevalent in India, considering majority of the subjects were students and teachers. These hypotheses, however, require a further investigation.

The psychological effect of lockdown was hypothesized to vary with the duration for which people stayed in the restricted condition. Figure 18 shows the variation of desire to go outdoors and efficiency of working online with the days of lockdown. Though this was not a longitudinal study which studied the behavior of the same subject for the entire duration, rather was a transverse study with single time filling of response, it was interesting to note that both the desire to go outdoors and efficiency of working online significantly increased with the days of lockdown (Equations 10‐11-10‐11).

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0010$ (10)

$urn:x-wiley:09056947:media:ina12778:ina12778-math-0011$ (11)

Variation of desire to go outdoors and efficiency of working from home with the lockdown days

The increase in the desire to go outdoors with the increase in the number of days of lockdown suggests that the subjects were stressed with the continuous indoor stay. In comparison, the increase in the efficiency of working online from home with the increase in the number of days represents the habituation of working online. However, the psychological impacts of the lockdown which affect factors like these need a further study.

4 CONCLUSION

The emergence of COVID‐19 has led to disruption in not only the health and economy but the lifestyle itself, which has a huge consequence on the pattern of energy consumption and corresponding environmental impacts. This is also true regarding buildings as the occupancy in social places have decreased, while those in residences have increased.

An online survey was conducted during the latter half of nationwide COVID‐19 lockdown in India. The status of thermal comfort of the occupants inside their houses and other details were recorded. A total of 406 complete responses were received in a fortnight period, from different places, which were classified as per the three climatic regions of cold, composite, and hot and humid climates in India. The following were the major observations from the study.

The mean clothing insulation was 0.36 clo (s.d. 0.216 clo, N 109), 0.28 clo (s.d. 0.109 clo, N 109), and 0.27 clo (s.d. 0.117 clo, N 188) in the subjects of cold climate, composite climate, and hot and humid climate respectively.
The TSV was seen to vary positively and significantly with the outdoor mean air temperature.
The productivity and overall comfort declined with the increase in the days of lockdown, which signifies the psychological impacts of the restrictions.
Variations in opening of windows and running of fans with the variation of outdoor mean air temperature were observed. While a quadratic relation between the number of open windows with outdoor mean air temperature suggested that at high temperature condition like the afternoon, subjects preferred to close the windows, the logistic regression between the running of fans and outdoor air temperature showed that subjects preferred to use fans at a much lower temperature than that found in the previous studies.
There was an increase in the number of sleeping hours among the subjects during the lockdown. The lunch time and the bed time were postponed among the subjects. On an overall, the subjects felt that they have increased on their weight during the lockdown.
The desire to go outdoors and efficiency of working online increased with the number of days of lockdown.

The online survey that was conducted had some limitations. First, as the responses were received only for a few days, the numbers of responses were limited. Second, due to the inability to measure the indoor environmental variables as per the ASHRAE Standard II ⁵ protocols in the online survey the indoor operative temperature could not be observed. This led to the inability of the researchers to calculate the comfort temperature and as a result further analysis of thermal comfort was not possible. Nevertheless, the study is expected to highlight into the problems and the condition of the subjects during forced and restricted indoor stay. Also, the findings will provide the foundations for the development of thermal comfort standards for built environment, where the subjects live in a stressful condition, for example, a hospital. Third, though thermal comfort questions were asked in detail, the study did not include stress related discussion. Also, a second round of the survey needs to be conducted to exactly the same sample group after the forced stay home period is over to address the possible difference.

ACKNOWLEDGMENTS

We are thankful to Dr Madhavi Indraganti, Department of Architecture, Qatar University, Doha, Qatar for her immense help during the preliminary phases. We are also thankful to Prof. Ajay Kr Bansal, Central University of Haryana and Prof. Goutam Kr Panda, Jalpaiguri Government Engineering College who were the former supervisors of the first author. We also convey our thanks to Prof. Michael Humphreys, Prof. Hom Bd. Rijal, Dr RL Sawhney, and Prof. Jyotirmay Mathur for their technical advice, time to time. We also sincerely thank Miss Kriti Rai, Dehradun (India), Miss Prayana Subba, and Mr Bikash Agarwal of Salesian College, Sonada (India) for the help rendered by them during the collection of responses. Last, but not the least we also thank all the 406 respondents who provided the valuable responses.

AUTHOR CONTRIBUTIONS

Samar Thapa (Corresponding and Lead Author) involved in overall research framework, design of experiment, data collection, data processing, data analysis, and paper drafting. Ramkishore Singh and Mahesh Bundele supervised and drafted the paper. Sheetal Thapa, George Thadathil, and Yogesh Jakhar collected the data and drafted the paper.

Data available on request from the authors.

Study of thermal comfort in the residents of different climatic regions of India—Effect of the COVID‐19 lockdown – Thapa – – Indoor Air

Related Posts

How Machine Learning has impacted Consumer Behaviour and Analysis

Market Research: The Ultimate Weapon for Business Success

Unveiling the Hidden Power of Market Research: A Game Changer

7 Secrets of Market Research Gurus That Will Blow Your Mind

The Shocking Truth About Market Research: Revealed!

Quantitative vs. Qualitative Research. How to choose the Right Research Method for Your Business Needs

Global 1-Methylcyclopropene(1-MCP) Market Professional Survey Report 2020 – LionLowdown

Categories

Recent Posts

Welcome Back!

Retrieve your password