Individual Human Climate Analysis Applying the New Clothing Resistance Model

1. Introduction

The first clothing resistance models appeared in the 70s of the last century [1,2], they can be traced back to the work of Burton and Edholm [3]. Ács and co-researchers followed the philosophy of this model type, but the thermal resistance of clothing (r_cl,t) was calculated differently [4,5,6,7,8] with respect to the previous works [1,2,9]. We have shown that r_cl,t is also significantly dependent on human factors [5,6,8]. In these models, the sweating term as energy balance component was not used in the energy balance equation of skin surface, so it could not be used in climates or weather situations with a large excess of heat (the value of r_cl,t is negative in this case), when skin surface evaporation is an important component of the energy balance equation.

The above problem can only be solved if the energy balance of the skin surface is supplemented with the component of sweating. So far, we cannot simulate sweating based on the description of physiological processes. Its estimate is based on the measurement of the mass of sweat produced, this is the so-called regulatory sweating [10]. Consequently, in this new model, the rate of sweating is an input variable to be determined by measurements. Thus, the thermal resistance module of the model is supplemented with an evaporative resistance module. Both modules estimate the characteristics of comfortable clothing. Comfortable clothing is characterized by the fact that it always ensures thermal balance between the human body and its environment, and it always remains dry. Of course, this is an imaginary clothing that does not exist in real life circumstances, and due to the imposed requirements, it is suitable for characterizing the thermal and evaporative load of the climate or the weather. Since sweating is a distinctly individual-specific process, the clothing parameters characterizing the climate and weather are also individual-specific, and from this point of view, the weather and climate characterization is also individual-specific [11]. The question is to what extent deviations vary from person to person?

The measurements and observations were carried out by a single individual. The location: Martonvásár (a small town in the Hungarian lowland, Carpathian Basin, Central Europe), period: summer 2023 and winter 2024. There was a total of 115 measurements, observations. To the best of the authors' knowledge, no similar observations and measurements for environmental, human thermal and evaporative load estimation have yet been carried out.

The goals of this study are as follows: 1) presentation of the new clothing thermal and evaporative resistance model, 2) description of the data required to run the model and description of the data collection methodology, and 3) description and analysis of the model results.

2. The Clothing Resistance Model

The model uses the energy balance equations of the clothing surface and the skin surface, it is a 1-layer clothing resistance model [12,13]. The clothing adheres closely to the human body, from below it is affected by metabolic heat flow density and water from sweating, while from above it is affected by the heat load of the environment. The change in heat and water content of the clothing was not taken into account. The clothing's parameters are: its thermal insulation capacity (resistance to heat transport), moisture insulation capacity (resistance to water vapor transport) and properties determining radiation transfer (albedo, emissivity). The former two parameters change depending on the environmental thermal load and the rate of sweating, characterizing the relationship between the individual and the weather. The radiation properties of the clothing are the input parameters of the model.

Clothing should be comfortable, that is, we should feel comfortable wearing it. This means that it must always remain dry, regardless of the rate of sweating, and we must not feel either an excess of heat or a lack of heat in it, regardless of our activity and the environmental thermal load. There is no such ideal clothing in reality, so the clothing parameters estimated with the model are the parameters of an imaginary clothing. A schematic diagram of the heat transports and state variables of the imaginary clothing is shown in Figure 1.

2.1. Basic Equations

The environmental heat surplus or heat deficit strongly depends on the radiation conditions, so the radiation balance of the clothing surface is also a determining environmental factor. The two energy balances and the radiation balance can be written as follows:

$$M-H_S-\lambda E_S-W=0,$$

(1)

$$R_n-H_{res}-\lambda E_{res}-H-\lambda E+H_S+\lambda E_S=0,$$

(2)

$$R_n=R_{ni}-4\bullet {\epsilon }_c\bullet \sigma \bullet T_a^3\bullet \left(T_{cl}-T_a\right),$$

(3a)

$$R_{ni}=S\bullet \left(1-{\alpha }_c\right)-\left({\epsilon }_c-{\epsilon }_a\right)\bullet \sigma \bullet T_a^4,$$

(3b)

where M is the metabolic heat flux density, W is the flux density of the work of the muscles used during the activity [2], H_res is the respiration sensible heat flux density, λE_res is the respiration latent heat flux density, R_ni is the net isothermal radiation energy flux density on the surface of the clothing, S is the solar radiation energy flux density, α_c is the albedo of the clothing, ε_c is the emissivity of the clothing, ε_a is the emissivity of the troposphere and σ is the Stefan-Boltzmann constant. M is parameterized according to [14]. Its parameterization is described in detail in the work of Kristóf et al. [8].

2.2. Parameterizations

The following parametrizations were used to describe the heat transports in the model:

$$H_S=\rho \bullet c_p\bullet \frac{T_S-T_{cl}}{r_{cl,t}},$$

(4)

$$H=\rho \bullet c_p\bullet \frac{T_{cl}-T_a}{r_{Ha}},$$

(5)

$$\lambda E_S=\frac{\rho \bullet c_p}{\gamma }\bullet \frac{e_S\left(T_S\right)-e_{cl}}{r_{cl,e}},$$

(6)

$$\lambda E_S=\lambda E_{sd}+\lambda E_{sw},$$

(6a)

$$\lambda E_{sd}=\mathrm{3,05}\bullet {10}^{-3}\bullet \left(256\bullet T_S-3373-e_a\right),$$

(6b)

$$\lambda E_{sw}=\frac{1}{A_f}\bullet \lambda \bullet \frac{dm}{dt},$$

(6c)

$$\lambda E=\frac{\rho \bullet c_p}{\gamma }\bullet \frac{e_{cl}-e_a}{r_{Ha}},$$

(7)

$$\lambda E_S=\lambda E,$$

(8)

Skin surface evaporation (6a) is the sum of evaporation from dry skin and skin covered in sweat. In the dry skin evaporation formula (6b), T_S is given in °C (34 °C), while e_a is given in Pa [15]. The rate of sweating (6c) cannot be simulated based on the description of physiological processes. There is a simple estimate, which is (6c), in which A_f is the surface of the human body, λ is the latent heat of evaporation and dm/dt is the loss of body weight caused by sweating, i.e. the amount of water sweated out per unit of time. This can easily be determined from the body mass and time values measured before and after the activity. The literature calls this λE_sw "regulatory sweating". The application of (6c) is challenging, but in our opinion it is the only reliable method for estimating the rate of sweating. The basic assumption of the model is that eq. (8) is valid. This is the guarantee that the clothing remains dry (there is neither convergence (clothing gets wet) nor divergence (clothing dries)), because we only feel comfortable in dry clothing. This makes sense, because wet clothing would sooner or later become uncomfortable due to its cooling effect. A_f was calculated by using the formula of Dubois and Dubois [16].

2.3. The Clothing Evaporative Resistance Model

The two essential output variables of the clothing evaporation model are e_cl and r_cl,e. By combining (6a), (7) and (8), it is easy to express e_cl,

$$e_{cl}=e_a+\left(\lambda E_{sd}+\lambda E_{sw}\right)\bullet \frac{\gamma }{\rho c_p}\bullet r_{Ha},$$

(9)

and since both λE_sd and λE_sw can easily be estimated based on the input meteorological and human (body mass, body length) variables, e_cl can be calculated with ease. It can also be seen that by applying (6), (7), (8) and (9), r_cl,e can also be expressed as follows:

$$r_{cl,e}=\frac{\rho c_p}{\gamma }\bullet \frac{e_S\left(T_S\right)-e_a}{\lambda E_{sd}+\lambda E_{sw}}-r_{Ha},$$

(10)

The parametrization of r_Ha has already been described in several works [5,8]. The unit of r_cl,e is [s·m⁻¹], but in the literature there is also the [clo] unit for water vapor transport. 1 clo-e = 0.155 m²·hPa·W⁻¹, based on which 1 clo-e = 0.155·(ρ·c_p/γ) = 0.155·(1.2·1004/0.65) = 287.3 s·m⁻¹.

2.4. The Clothing Thermal Resistance Model

The output variables of the clothing thermal resistance model are T_cl and r_cl,t. Let's express them! H_S can be expressed from (1) and described as

$$H_S=M-\lambda E_S-W=Hu.$$

(11)

Here, the difference in human heat flux densities is denoted by Hu. By combining (2), (5), (8) and (11), we can write the following equation:

$$R_n-H_{res}-\lambda E_{res}-\rho c_p\bullet \frac{T_{cl}-T_a}{r_{Ha}}+Hu=0.$$

(12)

The final form of T_cl after substituting (3a) into (12) is:

$$T_{cl}=T_a+\frac{R_{ni}-H_{res}-\lambda E_{res}+Hu}{\frac{\rho c_p}{r_{Ha}}+4\bullet {ϵ}_c\bullet \sigma T_a^3}.$$

(13)

Note that when estimating T_cl, we used both (1) (energy balance of skin surface under clothing) and (2) (energy balance of clothing surface), i.e., the magnitude of this temperature is influenced by both environmental and human factors. By combining (4) and (11), r_cl,t can be expressed,

$$r_{cl,t}=\rho \bullet c_p\bullet \frac{T_S-T_{cl}}{Hu}.$$

(14)

Equations (13) and (14) are new equations and characterize the thermal properties of the clothing that creates thermal balance (neither excess of heat nor lack of heat) between the human body and its environment. Since it is clothing that creates heat balance, this clothing is imaginary. The unit of r_cl,t is [s·m⁻¹], but the [clo] unit is also used for heat transfer. 1 clo-t = 0.155 m²·°C·W⁻¹, based on which 1 clo-t = 0.155·(ρ·c_p) =0.155·(1.2·1004) = 186.7 s·m⁻¹.

2.5. Operative Temperature Model

The operative temperature is a known heat load indicator. It is estimated as follows:

$$T_o=T_a+\frac{R_n}{\rho \bullet c_p}\bullet r_{Ha}.$$

(15)

As we can see, it depends on environmental factors (air temperature, radiation and wind). Its dependence on human factors is negligible (it is hidden in the parameterization of r_Ha).

3. Location of Measurements and Observations, Tools Used

The location of the observations: Martonvásár, on the vineyard road starting next to the family house of the first author. This location made it possible to minimize the difference between the two times of weight measurement and the duration of the walk. The town Martonvásár is presented in Figure 2 on a topographical map of Hungary.

The body mass was measured using an Orion digital weight meter, accuracy: 0.1 kg. When the mass values were read, the time was also recorded, to the minute. We estimated the duration of walking with two devices: an AppleWatch smartwatch and a JUNSO stopwatch, this was done for control purposes. Knowing the length of the road and the duration of the walk, it was also possible to estimate the average walking speed.

4. Data

Human and weather data were used.

4.1. Human Data

Human data provide information on 1) human state variables, 2) body mass immediately before and after walking, 3) the times of body mass measurements and thus the duration of observation, and 4) the duration of walking. The following human state variables are used: body mass (most important), body length, sex, age. These data of the observer are presented in Table 1.

The observer was always on foot. Walking speeds varied between 1.1 and 1.7 ms⁻¹.

4.2. Weather Data

The weather elements used to drive the model are: air temperature, air humidity, average wind speed, wind gust speed, cloud cover and relative sunshine duration. With the exception of the last two, all data were taken from the HungaroMet (Hungarian Meteorological Service Nonprofit Zrt.) website (https://www.met.hu/en/idojaras/aktualis_idojaras/megfigyeles/) and transcribed into the database. The beeline distance between the HungaroMet automatic station and the location of the observations is less than 3 km. These data are 10-minute averages. The values of cloud cover and relative sunshine duration were estimated by the observer for the selected 10-minute time interval. Cloudiness is estimated visually in tenths. Global radiation is an hourly value expressed in Wm^-2 (Mihailović and Ács, 1985 [17]) and refers to the hour interval in which the 10-minute observation took place. We performed a total of 115 observations in the period June 2, 2023 – March 9, 2024. The wheater elements observed ranged as follows: solar radiation varied between 0-870 Wm^-2, air temperature between -1-36 °C, average wind speed between 0.3-7.5 ms⁻¹, relative humidity between 34-100% and cloud cover between 0-1.

All the human data measured and weather data observed are pooled together in Table S1 of the Supplementary Material.

5. Results

Several results are presented and discussed. Since skin surface evaporation is a determining input variable, we analyzed its relationship with the rate of sweating, operative temperature and metabolic heat flux density. After that, we turned to analyzing the relationship between a) r_cl,t and λE_S, b) r_cl,t and T_cl, c) r_cl,e and λE_S and finally d) T_cl and T_o. It should be mentioned that the most important output variables of the model were listed for all 115 cases, they can be viewed in Table S2 of the Supplementary material.

5.1. Sweating Rate and Skin Surface Evaporation: Dependence on M and T_o

We could see that skin surface evaporation is the sum of evaporation from dry skin and skin covered in sweat (eq. 6a). Our measurements and calculations provide insight into the relationship between these variables in the case of a walking individual. The average walking speed of the observer varied between 1.1 and 1.7 ms⁻¹. In this speed interval, the measured sweat rates varied between 0-0.8 kg/(30-45 minutes). The largest amount of water sweated out was 0.8 kg, in this case the rate of sweating was 0.8/(72 minutes). The latent heat flux density of sweating varied from 0 to 340 Wm^-2. In contrast, latent heat flux of evaporation from dry skin is 10-15 Wm^-2. λE_sw is the dominant component of λE_S, which can be seen in Figure 3.

We can see that the λE_S – T_o relationship emerges more distinctively than the λE_S – M relationship, but it should be emphasized that T_o varied within much wider limits (-7 – 70 °C) than M (133-183 Wm^-2). The large spread of points suggests that the variability of λE_S is influenced by many different and independent factors.

5.2. The Relationship between Clothing Thermal Resistance and Skin Surface Evaporation

The scatterplot of the relationship between the thermal resistance of clothing and the latent heat flux density of skin surface evaporation is presented in Figure 6.

λE_S varied between wide limits (0-360 Wm^-2), the obtained r_cl,t value range was 0-43 clo-t. r_cl,t strongly depends on λE_S. It can be noticed that r_cl,t is close to 0 for very low (λE_S < 70 Wm^-2) and very high (λE_S > 165 Wm^-2) λE_S values. In the range 70 < λE_S < 165 Wm^-2, the r_cl,t values are greater than 3 clo-t in the vast majority of cases. It can also be noticed that in this range the value λE_S = 120 Wm^-2 (in this case the rate of sweating is 0.2 kg/(35 minutes)) is the limiting value, separating the increasing and decreasing r_cl,t values. It should be emphasized that these high r_cl,t values are not physically based, but rather the consequence of the singularity of the method (eqs. (14) and (11)). Very high r_cl,t values occur when Hu is close to 0. It is obvious that if Hu is close to 0 and greater than zero, then T_S > T_cl (cases of environmental heat deficit), and conversely, if Hu is close to 0, but less than zero, then T_S < T_cl (cases of environmental heat excess). The point Hu = 0 is the singularity point of the methodology (eq. (14)).

In the case of ambient heat deficit, r_cl,t increases with increasing λE_S. In this case, the smallest r_cl,t values are for λE_S = 0. These values were between 0.5 and 1.6 clo-t and, given the lack of heat, should be considered as physically based values. Note that minimal sweating (0.1 kg/(30 minutes)) significantly increases r_cl,t values in the case of environmental heat deficit. In these cases, r_cl,t can reach up to 2.5 clo-t.

In the case of environmental heat excess, r_cl,t decreases with increasing λE_S. When the rate of sweating is 0.5 kg/(30 minutes), the r_cl,t values are below 1 clo-t, which clearly shows that sweating can create a heat balance between the environment and the human body even in weather situations with a large excess of heat. Thus, for instance, for T_o = 43.5 °C, when λE_S = 229 Wm^-2 (sweating rate 0.5 kg/(44 minutes), average walking speed 1.36 ms⁻¹) r_cl,t = 0.11 clo-t, which practically means thermal neutrality.

It’s to be mentioned that in cases of thermal comfort (r_cl,t close to 0 clo-t), the absolute value of the (T_S -- T_cl) difference is very small (0.8-2 °C), completely regardless of which λE_S range we are talking about.

The described results refer to the blue points in the figure, in these cases the method can be used, that is r_cl,t > 0 clo-t. In the case of the red points shown in Figure 6, the method cannot be applied because r_cl,t < 0 clo-t. We will deal with the analysis of non-applicable cases separately in Section 5.6.

5.3. The Relationship between Thermal Resistance and Surface Temperature of Clothing

The scatterplot of the relationship between clothing thermal resistance and surface temperature is presented in Figure 7.

The large scatter of the points is caused by the mentioned methodological singularity (Hu → 0). Therefore, we will only look at r_cl,t values below 4 clo-t (the r_cl,t = 4 clo-t value can be taken as the upper limit of the heat deficit in the lowland regions of Hungary). In this case, the characteristics of the r_cl,t – T_cl relationship can already be clearly recognized. The r_cl,t – T_cl relationship can be divided into 3 ranges: 1) T_cl is much smaller than T_S, 2) T_cl is around T_S and 3) T_cl is much larger than T_S. The main characteristic of the 1^st range is the lack of environmental heat, accordingly r_cl,t gradually increases as T_cl decreases. This change is clearly visible. In conditions close to thermal neutrality, r_cl,t values are by definition close to 0 clo-t. Here there is also a point with r_cl,t = 1.3 clo-t, which is large despite the small (T_S – T_cl) difference (0.4 °C), because Hu is very small, it is only 2.2 Wm^-2. This result does not reflect the real thermal state, it is caused by the singularity of the methodology. In the 3^rd range, there are smaller or larger excesses of heat, during which the amount of sweating can also be smaller or larger. If the excess of heat is extremely high and so is the rate of sweating, then r_cl,t → 0 clo-t. If, on the other hand, there is less sweating in such cases, T_cl and r_cl,t will also be higher. If Hu becomes positive due to a small amount of sweating, then the method is unapplicable. If the environmental heat excess is smaller and the amount of sweating is also smaller, the r_cl,t values change similarly as in the previous case. Given that we are talking about excess heat conditions, r_cl,t values greater than 1.2-1.4 clo-t cannot be considered as physically based, but rather as caused by the singularity of the method.

5.4. The Relationship between the Clothing Evaporative Resistance and Skin Surface Evaporation

The scatterplot of the relationship between the evaporative resistance of clothing and the latent heat flux density of skin surface evaporation is presented in Figure 8.

The r_cl,e – λE_S relationship is determined by the requirements (eqs. (8) and (10)) for comfortable clothing. As we can see, r_cl,e decreases exponentially with increasing λE_S. The r_cl,e values varied between 0.1 (for λE_S = 353 Wm^-2) and 21 (λE_S = 14 Wm^-2, there is no sweating) clo-e. If there is already sweating (λE_S = 60-80 Wm^-2, sweating rate is 0.1 kg/(30 minutes)), the r_cl,e values are around 4-5 clo-e. The relationship between r_cl,e – e_cl is also exponential, like the relationship between r_cl,e and λE_S, so it is neither illustrated nor analyzed.

5.5. The Relationship between Operative Temperature and Clothing Surface Temperature

In addition to the estimation of Hu, the estimation of T_cl is also important for the success of the method. In the absence of direct verification, T_cl was compared with T_o. The scatterplot of the relationship between clothing surface temperature and operative temperature is presented in Figure 9.

The relationship between T_cl and T_o is linear, despite the scatter of points. Two ranges can be recognized: 1) in the range T_cl < T_S the T_cl values are larger than the T_o values, and 2) in the range T_cl > T_S the T_cl values are generally somewhat smaller than the T_o values.

5.6. Applicability of the Model

The model has limited applicability. Three cases can be distinguished: 1) the model cannot be applied because the values of r_cl,t < 0, 2) the model simulates large, non-physically based r_cl,t values (methodological singularity) and 3) the model simulates real, physically based r_cl,t values. All these cases can be seen in Figure 10, where the scatterplot of the r_cl,t – Hu relationship is shown.

Red points represent not applicable cases (case 1). The blue points with a very high r_cl,t values (case 2) are the consequence of the methodological singularity, they are mostly in the range -40 < Hu < 40 Wm^-2. The physically based r_cl,t values (case 3) are in the most negative (the latent heat flux density of the sweated water is much higher than the metabolic heat flux density and there is a large heat surplus) and the most positive (no sweating and there is a large heat deficit) Hu value ranges. In this figure, it is difficult to recognize the points representing the state of thermal neutrality (r_cl,t is close to 0 clo and the temperature difference (T_S – T_cl) is small, a few °C). These cases can be seen in Figure 11 for values of T_cl, which are close to T_S.

In summary: the method performs well in the range of thermal neutrality, as well as when 1) the environmental heat surplus is high and the rate of sweating is high, 2) the environmental heat deficit is high and there is no sweating.

6. Discussion

Our measurements and analyzes showed that the model is suitable for simulating and characterizing the individual human climate. Today, there are no individual human climate descriptions, only attempts [6] that are similar to this one. The reasons for this are as follows: 1) the concept of the "standard/average" person has spread in human meteorological studies. There is no simulation for individuals, so that human variability does not have to be dealt with, and it is also important that conducting longitudinal measurements with individuals is a very tedious and complex task. 2) We characterize a human climate instead of a human thermal climate by expanding the concept of comfortable clothing. Clothing is comfortable when it is dry and thermally neutral. The imaginary clothing used in the model fulfills these two conditions. The condition of thermal neutrality was achieved by applying the energy balance equation at the clothing-air environment interface, and the condition of dry clothing is achieved by equating the latent heat flux densities leaving the skin surface and the clothing surface. Thus, the clothing was characterized not only from a thermal point of view, but also from a moisture point of view, accordingly, the model consists of a thermal and an evaporative module. With the addition of the evaporative module and its coupling to the thermal module, we obtained a new clothing resistance model.

Which are the most important results that can characterize the simulated individual human climate? We would highlight three results that refer to the environmental heat surplus, the environmental heat deficit, and thermal neutrality. In the case of high thermal loads, the r_cl,t values were between 0.1-0.5 clo-t, while the r_cl,e values were between 0.4-0.8 clo-e. The values of r_cl,t and r_cl,e close to 0 were caused by intense sweating (λE_sw > 200 Wm^-2). The former models [1,5] - as they did not simulate the process of sweating - were not applicable in these heat excess situations. In heat-deficient situations, the amount of sweating decreases, so the r_cl,e values increase, and they are the highest when there is no sweating and the evaporation from the skin surface is equal to the evaporation from dry skin. In such cases, r_cl,e values were around 20-21 clo-e (λE_sd 10-15 Wm^-2), r_cl,t values varied between 0.5-1.6 clo-t. It should be mentioned that the average winter r_cl,t values are around 1.3-1.8 clo-t [4,5] In conditions close to thermal neutrality (T_cl values are close to T_S), we can sweat while walking - especially if we walk at a faster pace. Thus, when r_cl,t values were 0.1-0.4 clo-t, the corresponding r_cl,e values were 0.8-3 clo-e. Sweating is considered to be natural when we walk for longer distances at higher speeds (> 1.1 ms⁻¹).

These parameters characterize imaginary clothing for an individual (Table 1). How close are the parameters of the imagined clothing to the parameters of real clothing? In this regard, we would like to make a comment about the r_cl,e values. We observed that the t-shirt worn was always wet when sweating, this was visible. This means that the sweated water flux density is convergent, that is, the clothing absorbs water, even if the amount is negligible. Real and imaginary clothing differ in this, that is, the r_cl,e values of the real clothing are obviously larger than the r_cl,e values of the imaginary clothing. Therefore, the r_cl,e values of the imagined clothing are to be considered as lower limit values.

One of the key variables in the model is the rate of sweating. λE_sw determines both the thermal and evaporative climate of an individual. In the cold season, λE_sw is minimal, close to zero; in the warm season it can be extremely high (200-300 Wm^-2), it depends not only on the environmental heat load, but also on the activity. In our study, the λE_sw – M dependence is weak (Figure 5), as M varied between narrow limits (135 – 189 Wm^-2), while the variability of the environmental heat load (-7 °C < T_o < 70 °C) was large. The relation of λE_sw to M is of decisive importance, which can be seen from equations (11) and (14) of the model. If Hu is close to 0 (there is a mathematical singularity at the point Hu = 0), the r_cl,t values are unrealistically high, that is, they are not physically based. Little is known about the interpersonal variability of sweating [18]. To establish this, several similar longitudinal experiments would be needed, which goes beyond the framework of an individual project.

The model can be applied to all activities if M and λE_sw characterizing the activity are known. In our previous models, we did not estimate sweating, the activity was walking at a speed of 1.1 ms⁻¹ [4,5,6,7,8]. Examining the applicability of the model in the case of lying down and running is one of the important goals of our future research. We can say that the model can be used in any climate, weather and activity. The determinant physiological variables of the model are M and λE_sw, the relationship between which is unknown in many respects.

7. Conclusions

In this study, we continued the development of the clothing resistance model type. Until now, these models [4,5,6,7,8] have only been used in the winter season and in cold climates, because the determination of sweating was omitted when estimating r_cl,t. We changed this concept and characterized both the thermal and moisture climates of a person by determining sweating. Since sweating is individual-specific, the model can be applied individually. Our main conclusions are as follows: 1) the human climate can be characterized not only from a thermal point of view, but also from a moisture point of view. Knowledge of sweating is necessary not only to estimate r_cl,e, but also r_cl,t. 2) Sweating is the determining variable in the estimation of both r_cl,t and r_cl,e, but little is known about its interpersonal variability. 3) The applicability of the model strongly depends on the relationship between λE_sw and M, the human variability of which we know little about.