Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II

WCECS 2011, October 19-21, 2011, San Francisco, USA

Estimation of Black Globe Temperature for

Calculation of the Wet Bulb Globe Temperature
Index
Vincent E. Dimiceli, Steven F. Piltz, and Steven A. Amburn

AbstractThe wet bulb globe temperature (WBGT) index is

used in industry, sports and other areas to indicate the heat
stress level for humans and animals. One of the values needed
to calculate the WBGT Index is the black globe temperature.
The black globe temperature is measured using a Black Globe
Temperature Sensor which includes a black globe with a
thermometer inserted in the center. However, the Black Globe
Temperature Sensor can be costly and many of these
instruments may be needed to get measurements in many
locations. The lead author has derived a formula to estimate
the black globe temperature using readily available data
collected by the National Weather Service (NWS). The
formula was derived from a formula suggested by Kuehn [7],
which was based on heat transfer theory. The resulting
equation was a fourth degree polynomial in terms of the black
globe temperature. It was determined that the black globe
temperature can be very accurately approximated by taking a
fourth degree polynomial in terms of the black globe
temperature to create a linear approximation for black globe
temperature. Some preliminary tests indicate accuracy within
0.5F.
Index Termsblack globe, heat stress,
measurement, wet bulb globe temperature

temperature

I. INTRODUCTION

ne of the government regulations

instituted by the Occupational Safety and
Health Administration (OSHA) is heat
stress management [8]. The manual states in
Section III: Chapter 4 the second paragraph of the
introduction:

Outdoor operations conducted in hot weather,

such as construction, refining, asbestos removal,
and hazardous waste site activities, especially
those that require workers to wear semipermeable
Manuscript received October, 2010. Manuscript revised July 28, 2011.
V. E. Dimiceli is an Associate Professor of Mathematics with Oral
Roberts University, Tulsa, OK 74171 USA (phone: 918-488-6700; fax:
918-495-6660; e-mail: vdimiceli@oru.edu).
S. F. Piltz is Meteorologist-in-Charge with the National Oceanic and
atmospheric Administration, National Weather Service, Tulsa, OK 74128
USA (e-mail: steven.piltz@noaa.gov).
S. A. Amburn is Science and Operations Officer with the National
Oceanic and atmospheric Administration, National Weather Service, Tulsa,
OK 74128 USA (e-mail: steve.amburn@noaa.gov).

ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

or impermeable protective clothing, are also likely

to cause heat stress among exposed workers.
A rating is calculated which indicates the safe
amount of time a person can work outside on a hot
day. This quantity is called the Wet Bulb Globe
Temperature Index (WBGT). In the past, WBGT
data has been collected manually using a portable
instrument. The OSHA manual includes the
following formulas for the WBGT:
1. For indoor and outdoor conditions with
no solar load, WBGT is calculated as:
WBGT = 0.7NWB + 0.3GT
2. For outdoors with a solar load,
WBGT is calculated as
WBGT=0.7NWB + 0.2 GT + 0.1DB
Where: GT= globe temperature
NWB=natural wet-bulb temperature
DB=Dry-bulb temperature
This index is important to the military, sports
teams, construction workers, and anyone who will
be exerting effort in hot weather. The American
Academy of Pediatrics references this index [1]
for child safety in hot temperatures. Also, athletes
should cancel any outdoor training activity when
the WBGT Index is above about 82 degrees
Fahrenheit (about 28 degrees Celsius) [2]. The
value at which marines will cease all outdoor
training activity is about 90 degrees Fahrenheit
(about 32 degrees Celsius).
The International Standards (ISO) number 7243
is also based on the WBGT. Parsons [9] gives a
good description of how the WBGT Index can be
applied globally by considering ISO 7243. One
drawback can be the cost of the WBGT device, as
well as having personnel trained to use the device.
WCECS 2011

Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II
WCECS 2011, October 19-21, 2011, San Francisco, USA

Recently the crews that were working on the

Gulf oil spill in the summer of 2010 needed this
index all along the Gulf Coast. However, in order
to collect the data necessary for calculating the
WBGT, a relatively expensive device was needed
for many different locations along the Gulf coast
making this task cost prohibitive. Consequently,
the National Weather Service (NWS) was asked to
provide the WBGT using only data that is
routinely collected by the NWS. The main
problem with this is that one of the variables in
the equation to calculate the WBGT index is the
"globe temperature." This temperature is found
by using a copper globe painted in black matte
paint with a thermometer inserted so that the bulb
is in the center of the globe. This temperature is
not routinely collected by the NWS.
Turco, et al. [12] derived equations to estimate
the black globe temperature based on
meteorological data. However, their model was a
statistical model, not a physical model. The
equations derived were regression equations
computed from meteorological data. Although the
equations were extremely accurate, a more
accurate model may be derived from the heat
equations for the black globe. According to the
authors:
The models developed resulted in great
performance to predict the black globe
temperature, allowing the estimation of
bioclimatic indices to assess the conditions of the
environment, to accomplish regional studies, and
to indicate best house designs for animals. [12]
This paper shows how the globe temperature
can be approximated using only data routinely
collected by the NWS. A fourth degree
polynomial equation is derived for globe
temperature with the coefficients dependent on
readily available data. Then, it is shown that the
fourth degree polynomial can be reasonably
approximated by a linear equation, thus making
computation less costly and time-consuming.
Finally, some experiments were done to verify
the accuracy of the estimate using the linear
expression in terms of temperature.
II. DERIVATION

exception of the constant (h) in the second term on

the right:
1

(1)

The coefficient in the second term on the right

side of equation (h) is from the convective heat
transfer coefficient. It was determined during
testing that this coefficient varied according to the
Solar Irradiance and the cosine of the zenith angle.
A multiple power regression was performed to
determine an equation for h in terms of S and
cos(z), where z is the solar angle to zenith. The
equation that approximates h is:
a(Sb)([cos(z)]c)

(2)

Where a, b and c are determined experimentally

from data using multiple power regression. Now,
putting all Tg terms on the left of equation (1), and
dividing by
we get:
.

(3)

The values of all variables except Tg are

either given or can be calculated from
available data from the NWS. The following
values are provided below.
Globe albedo for short and long wave radiation:
0.05 so 1
1
0.95.
Black globe emissivity: =0.95
Stephan-Boltzman constant:=5.67x10-8 is
used.
Albedo for grassy surfaces:
= 0.2.
When these values are entered into equation (3)
we get:
.

.
.
.

(4)

The following heat equation was taken from a

paper by Hunter and Minyard [5], with the
ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

WCECS 2011

Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II
WCECS 2011, October 19-21, 2011, San Francisco, USA

Hunter and Minyard, in their paper [5], show

that
, where z is the solar angle to

zenith. Putting this into (4), we get

.

(5)

Where S is solar irradiance,

is the direct
beam radiation from the sun and
is the diffuse
radiation from the sun. The Stefan-Boltzman
Constant is and h is the convective heat transfer
coefficient.
The convective heat transfer
coefficient is calculated experimentally as
indicated above. Finally, the ambient temperature
is represented by Ta and the wind speed by u in
meters per hour. All of these are given or may be
calculated directly from data given by the NWS.
The last parameter on which the globe
temperature depends is the thermal emissivity, .
According to Hunter and Minyard [5], thermal
emissivity can be calculated using

0.575

1.0007

0.00000346

Figure 1. The graph of

C8,389,000 and t between 20 and 60.

and y1=Ct 7,680,000 for

Notice that the curve appears to be very close to

the linear graph. We compute the curvature for y
to see how close to a linear function y is. The
curvature of
is given by

(6)

Where ea is atmospheric vapor pressure, which

may be calculated by
exp

By doing this, we can treat (8) as a fourth

degree polynomial in terms of Tg. The values of
Tg in which we are interested are in the interval
[20, 60], since values below 20C are too cold to
cause heat stress and values above 60C, in
general, do not occur. Figure 1 shows a graph of
and y1=Ct 7,680,000 (the tangent
line approximation for the function y at t=40) on
the interval [20, 60] (C was calculated for a wind
speed of about 15 mph).

6.112exp

(7)

where P is the barometric pressure and Td is the

dew point temperature.
When we take into consideration the fact that all
parameter values in equation (5) are constants that
can be entered at constant time intervals, we can
reduce the equation to
,

(9)
In order to get an understanding of the
magnitude of the curvature, consider the graph
(Figure 2) of the function

for u between 1 mph and 40 mph (1690 meters per

hour to about 65,000 meters per hour) and t
between 20C and 60C.

(8)
Figure 2. Pictured above is the curvature of y for 20<t<60 and
1<u<76,000.

Where
.

and

ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

Notice that the curvature is on the order of

410-10 or less on the domain of interest. This
confirms the assumption that y is nearly linear for
values of t and u that make sense for this context.
It is therefore reasonable to use a linear
approximation for y to solve for t (=Tg). In other
WCECS 2011

Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II
WCECS 2011, October 19-21, 2011, San Francisco, USA

words we may use a linear approximation on the

left side of equation (8) to estimate the value of
the globe temperature.
Using differential calculus to find the
equation of the tangent to the curve at t=40 (the
midpoint of the interval [20, 60]), we find that the
left side of equation (8) may be substituted by
256000

(10)

7680000.

Putting this in place of the left side of equation

(8) and solving for Tg, we get
(11)
with B, C and Ta as defined previously. Now
we have an estimate of Tg dependent only on
values which are either readily available from the
NWS or may easily be calculated from data
available from the NWS. Also, the equation is
linear making for easier computation than what
was necessary to solve the original fourth degree
polynomial.
III. PRELIMINARY TESTS

In September, 2010, three preliminary tests

were conducted to test the accuracy of the globe
temperature estimate.
A Black Globe
Temperature Sensor for heat stress was created by
an employee of the NWS using specifications
from an article by Purswell and Davis [7]. A
picture of the unit is included in Figure 3. This
unit was used to get some preliminary readings to
check for accuracy of the equation.

Figure 3. Black Globe Temperature Sensor used in preliminary tests.

Figure 4. Black Globe Temperature Sensor for Heat Stress used to

verify equations at the Oklahoma Mesonet test facility in Norman
Oklahoma.

ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

The first preliminary test was done on

September 9, 2010 in front of the NOAA offices
in Tulsa, Oklahoma. The weather conditions were
hazy that day with air temperature 86F, and dew
point temperature 69F. The barometric pressure
was 30.08 in. of Hg (about 993 mb for pressure
not adjusted for sea level) and the solar irradiance
was 336 W/m2. The wind speed averaged around
5 to 6 mph during the measurement. The globe
temperature was measured to be 91F using a
black globe as described earlier in this paper.
Using Excel, a spread sheet was created to use the
derived equation to estimate globe temperature.
The equation estimated the globe temperature to
be about 91.434F.
Another preliminary test was performed on
September 10, 2010. The conditions were sunny
with air temperature 93F and dew point
temperature 76F. The barometric pressure was
29.75 in. of Hg (about 982 mb for pressure not
adjusted for sea level) and the solar irradiance was
754 W/m2. The wind speed was measured at
about 7 mph during the measurement. The globe
temperature was measured to be 103F using a
black globe. The equation estimated the globe
temperature to be about 102.757F.
The third preliminary test was performed on
September 17, 2010. The conditions were similar
to the conditions on September 10. The air
temperature was 94F and dew point temperature
76F. The barometric pressure was 30.05 in. of
Hg (about 992 mb for pressure not adjusted for
sea level) and the solar irradiance was 579 W/m2.
The wind speed was measured at about 3.7 mph
during the measurement. The globe temperature
was measured to be 105F using a black globe.
The equation estimated the globe temperature to
be about 105.175F.
The three preliminary tests indicate that the
formula used to estimate the globe temperature is
very accurate. If the estimate is within 1F, it is
sufficient to estimate the WBGT index. As can be
seen from the preliminary tests, the estimates are
within about 0.5F. The main problem with our
tests was that we had to estimate the wind speed
and the formula is very sensitive to the value of
the wind speed. However, the estimates for wind
speed should be within about 0.5-1 mph. Also,
these tests were all done at about the same time of
day. Therefore, the solar irradiance and the cosine
WCECS 2011

Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II
WCECS 2011, October 19-21, 2011, San Francisco, USA

of the zenith angle were about the same for all

preliminary tests.
IV. TESTS WITH OFFICIAL WBGT SENSOR

An official black globe was set up at the

Oklahoma Climatological Survey site in Norman
Oklahoma for data collection in June, 2011. The
data from this site was then analyzed to determine
accuracy of the black globe temperature equation.
After extensive analysis and the derivation of the
Multiple Power Regression Equation for the heat
transfer coefficient, h, equation (11) was
successful in calculating the black globe
temperature to within one degree Celsius. An
Excel document has been included in the appendix
indicating the accuracy of the equations for
different times of the day on two days in July.
The instrument was set to collect data every
minute. The data collected included the black
globe temperature, natural wet bulb temperature,
ambient temperature, and relative humidity. Wind
speed at 2 meters above ground, solar irradiance,
and barometric pressure were collected, as well.
The data was then retrieved via the internet.
As can be seen from the Excel document in the
appendix, the black globe temperature was
estimated to within 0.666 C in every case. Since
the factor for the black globe temperature in the
WBGT Index equation is 0.2, this error will
contribute less than 0.15 C to the WBGT Index
error.
V. AN ALGORITHM

In this section, an algorithm is created for the

calculation of globe temperature estimates. First,
we will consider the values readily available from
the NWS. These will be input values to be
entered at the beginning of the program.
1. The values to be entered are wind speed
(u in meters per hour), ambient temperature
(Ta in degrees Celsius), dew point temperature
(Td in degrees Celsius), solar irradiance (S in
Watts per meter squared), direct beam
radiation from the sun ( ) and diffuse
radiation from the sun (
).
2. The zenith (z) angle may be entered or
calculated. (The angle, z, must be in radians
for Excel.)
3. The thermal emissivity must be calculated
next. Using the following two equations.

ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

a.

exp

1.0007
6.112exp
b.

0.00000346
.
.

0.575

4. Now B and C can be calculated using

the following equations.
.

a.

b.
, where h is
.
computed using (2).
5. Finally the estimate for globe
temperature is calculated using equation
(10).
7680000
256000
VI. NATIONAL WEATHER SERVICE APPLICATIONS

The National Weather Service produces

forecasts and observations of numerous
meteorological parameters. One of these is the
apparent temperature or Heat Index, based on
work by R.G. Steadman [11]. In this work,
Steadman constructed a table which uses relative
humidity and dry bulb temperature to produce the
"apparent temperature" or the temperature the
body "feels." The OSHA uses a more detailed
approach based on the WBGT Index [8] which
provides guidelines to protect workers. In an
effort to provide decision makers with forecast
information on the WBGT, the National Weather
Service in Tulsa has implemented and is testing
the algorithm described earlier in this paper.
The algorithm is being tested for feasibility and
accuracy in the forecast and also the hourly
analysis of WBGT. In the current test phase, the
first goal is to provide real-time, hourly estimates
of WBGT across the WFO Tulsa forecast area of
eastern Oklahoma and northwest Arkansas. The
second goal is to make forecasts of WBGT out
seven days to provide decision makers with
important planning information. Ultimately, the
algorithm will be offered to all National Weather
WCECS 2011

Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II
WCECS 2011, October 19-21, 2011, San Francisco, USA

Service Offices across the United States and

elsewhere.
The National Weather Service uses a forecast
system called the Graphical Forecast Editor, or
simply GFE [6]. Virtually all weather variables
are included in the system in a gridded format at
either 2.5 km resolution or 5 km resolution. Grid
fields of forecast parameters such as air
temperature, dew point temperature, wind,
pressure, cloud amount, the probability of rain,
rainfall amounts, and many others are produced
through a forecast period of seven days. From
these parameters, an assortment of secondary
forecast variables can be created through various
algorithms. Some of these currently include
relative humidity, wind chill index, heat index,
icing index, wild fire spread index, and others.
The WBGT algorithm described here will be
tested and eventually added to the suite of
products issued by WFO Tulsa.
An example of the WBGT Index graphics page
is shown in Figure 4.
This indicates the
variability of the index across the forecast area
due to differences in the contributing parameters.
The darker areas indicate higher index values.
Graphical forecasts of the WBGT may eventually
be available to the public and decision makers,
along with point forecasts of WBGT for any
location on the map.

VII. CONCLUSION

The black globe temperature, until now, has

either been measured using a Black Globe
Temperature Sensor on site, or by using a
regression equation to estimate the WBGT. One
example of this equation is used by the Australian
Bureau of Meteorology, but does not take into
account variations in the intensity of solar
radiation or of wind speed [4]. Dimiceli has
derived a linear expression to estimate the black
globe temperature to within about 0.666 C, thus
estimating the WBGT Index to within about 0.333
C or .733 F. More tests have been performed
using an official Black Globe Temperature
Sensor. These tests provided data to get a multiple
power regression equation to estimate the
convective heat transfer coefficient (h) used in the
estimation of the black globe temperature in the
formula (11) by Dimiceli.
The formula has been added to an internal
experimental web page in order to allow
computation of the WBGT Index for a limited
area. This web page allows interested parties to
calculate the WBGT Index for their local area by
entering the ambient temperature (Ta) and
approximate average wind speed at their location.
A tool which will be available to anyone
interested in computing the danger of heat stress
at their particular location is being created at the
writing of this article. By doing this, the NWS
will give sports teams, construction companies,
military personnel, and other affected people, the
ability to determine the danger of working in the
heat at different times of the day. Eventually an
application for cell phones will be designed to
calculate the WBGT Index. It is the investigators
hope that this index will eventually be used
nationally (or even internationally as Parsons [9]
proposed in his article) by the NWS to indicate the
dangers of heat stress in any location in the
country.

Figure 5. Display of WBGT index from WFO Tulsa

Graphical Forecast Editor

ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

WCECS 2011

Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II
WCECS 2011, October 19-21, 2011, San Francisco, USA

APPENDIX
Table 1: Preliminary Tests
fdb
fdif
Ta F C
Td F C
P in. of merc, mb
S W/m2
z degrees radians
u mph, mtrs per hr
Rh
Tw C
ea
epsa
B
C
Tg C F
e
actual Tg F C
WBGT
dimiceli
actual
australia

9/9/2010
0.67
0.33
86
69
30.08
0.336
38.44
6
67.5
26.7755045
22.6486791
0.89793603
3614652151
1197110170
33.0188360
28.543833

30
20.556
992.83
336
0.6709
9654

91.434

9/10/2010
0.75
0.25
93
76
29.75
0.754
36.65
7
54.27
26.5755611
28.4870860
0.92784344
7098450550
1309071170
39.3095775
28.5844209

33.889
24.444
981.94
754
0.6397
11263

102.76

9/17/2010
0.75
0.25
94
70
30.05
0.579
41.41
3.7
52
26.34526836
22.48619082
0.897012895
5617678219
904405917.3
40.65288578
28.24707268

34.444
21.111
991.84
579
0.7227
5953.3

105.18

91

32.778

103

39.444

105

40.556

28.3466203
28.2984087
32.1677267

83.024
82.937
89.902

29.8536972
29.8806707
34.3886774

85.737
85.785
93.9

30.01670945
29.99724341
34.57109956

86.03
85.995
94.228

The table above is the Excel Document which

shows the results of the three preliminary tests run
in September.

ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

WCECS 2011

Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II
WCECS 2011, October 19-21, 2011, San Francisco, USA

Table 2. Some results from official Black Globe Sensor

12:30
fdb
fdif
Ta (C)
Td (C)
P mb
S W/m2
cos(z)
u (m per
sec, m per
hr)
Rh
h
Tw (C)
ea
epsa
B
C
Tg (C F)
e
actual Tg F
C
error

fdb
fdif
Ta (C)
Td (C)
P mb
S W/m2
cos(z)
u (m per
sec, m per
hr)
Rh
h
Tw (C)
ea
epsa
B
C
Tg (C F)
e
actual Tg F
C
error

7/3/2010 13:00
0.75
0.25
26.26
17.21
973.62
193.04
0.9714

6487.2

88.98574

15:30
(not avgd)

13500

109.0913

7/3/2010

14:00

0.75
0.25
28.8
17.57
973.71
448.7
0.9244

1.812 6523.2
57.5
0.101274459
20.94781261
18.55668613
0.872734593
1678945078
306605873.4
31.73445826 89.122
19.56661611

2.164 7790.4
50.7
0.132569543
21.45412447
18.59184046
0.872970592
3979819932
444878843.2
37.74139651 99.9345
20.00916235

31.55
-0.18445826

37.16
-0.58139651

7/10/2010
0.75
0.25
29.5
16.2
971.03
173
0.9689

13:00

7/10/2010
0.75
0.25
31.06
16.59
971.37
387.98
0.9226

14:00

2.734 9842.4
44.65
0.098524678
20.46198409
16.74872176
0.860047363
1506448296
378649115.5
33.47613079 92.257
18.34712008

3.86
41.86
0.127660528
20.86084154
16.93891306
0.86143581
3444243993
599274858.9
36.80444598
18.80503063

13896

33.58
0.103869211

37.47
0.665554019

ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

98.248

WCECS 2011

Proceedings of the World Congress on Engineering and Computer Science 2011 Vol II
WCECS 2011, October 19-21, 2011, San Francisco, USA

ACKNOWLEDGMENTS
The authors would like to thank Bruce Sherbon
of NOAA for his construction of an inexpensive
Black Globe Temperature Sensor. The authors
would also like to thank Jamie Frederick of
NOAA for programming the web page to
calculate the WBGT Index. Finally, V. E.
Dimiceli would like to thank Glenn Wiley of
NOAA for helping V. E. Dimiceli to get a
research sabbatical at NOAA/NWS in the fall of
2010.
REFERENCES
[1]

American Academy of Pediatrics, Committee on Sports Medicine and

Fitness. Sports Medicine: Health care for Young Athletes. 2nd ed. Elk
Grove Village IL: American Academy of Pediatrics; 1991: 98
[2] Binkley, Helen M, Beckett, Joseph, Casa, Douglas J., Kleiner,
Douglas M., and Plummer, Paul E. (2002). National Athletic
Trainers Association Positional Statement: Exertional Heat Illnesses.
J Athl Train. 2002 Jul-Sep; 37(3): 329343.
[3] Campbell Scientific (Canada) Corporation. (2010). Black Globe
Instruction Manual, Campbell Scientific (Canada) Corporation.
Available: http://www.osha.gov/dts/osta/otm/otm_iii/otm_iii_4.html
[4] Commonwealth of Australia (2010). Thermal comfort observations,
Bureau of Meteorology, ABN 92 637 533.532
[5] Hunter, C. H. and Minyard, C. O. (1999). Estimating wet bulb globe
temperature using standard meteorological measurements, U.S.
Department of Energy, Office of Scientific and Technical
Information, WSRC-MS-99-00757, 2.7
[6] Information Generation Section, National Oceanic and Atmospheric
Administration (Summer, 2007). Joe Wakefield, Systems chief.
Earth System Research Laboratory, web page. Available:
http://gfesuite.noaa.gov/EFTHome.html
[7] Keuhn, L. A., et al, (1970). Theory of the globe thermometer, Journal
of Applied Physiology, Vol. 25, No. 5, 750-757.
[8] OSHA Technical Manual, TED 01-00-015 [TED 1-0.15A], updated
June,
2008.
Available:
http://www.osha.gov/dts/osta/otm/otm_iii/otm_iii_4.html
[9] Parsons, K. (2006). Heat stress standard 7243 and its global
application. Industrial Health , 44 368-379
[10] Purswell, J. L. and J. D. Davis (2008). Construction of a low cost
black globe thermometer. Applied Engineering in Agriculture.
American Society of Agricultural and Biological Engineers. Vol.
24(3): 379-381.
[11] Steadman, R. G. (1979): Part I: A Temperature-humidity index based
on human physiology and clothing science, Journal of Applied
Meteorology, July 1979, Vol. 18 No. 7, pp. 861-873.
[12] Turco, S. H. N., T. G. F da Silva, G M de Oliveira, M. M. V. B. R
Leito, M. S. B de Moura, C Pinheiro, C. V da Silva Padilha. (2008).
Livestock Environment VIII, 31, August 4 September 2008, Iguassa
Falls, Brazil 701PO408.

ISBN: 978-988-19251-7-6
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

WCECS 2011