Technical Note 16

22.12.00

CALCULATING DEW POINT

FROM RH and AIR
TEMPERATURE

Introduction Error in the Estimation of Dew Point

Dew point can be measured directly, to a high Teten’s equation is an approximation of the true
degree of accuracy, using traditional devices such variation of saturated vapour pressure as a function
as cooled mirror hygrometers, etc. of temperature. However, the errors in using the
inverted form of the equation result in dew point
However, such devices are often very expensive,
errors much less than 0.1°C.
require regular maintenance and may need air
pumps. They are also heavy on power consumption.
The largest component of error, in reality, comes
An alternative method, described in this Technical from errors in the absolute calibration of the
Note, uses relatively inexpensive RH and temperature and RH sensor.
Temperature sensors, in conjunction with a
Campbell Scientific datalogger, to calculate dew Figure 1 shows how dew point varies as a function
point. While end results may not be quite as of temperature and humidity. It can be seen that the
accurate as traditional dedicated devices, they are response is non-linear with respect to both
acceptable for a wide range of applications. variables. Errors in the measurement of RH and
temperature thus form a complex function in relation
Calculating Dew Point to the resultant error in estimated dew point. In
practise, the effect of errors in the calibration of air
Dew point temperature can be calculated by temperature can be taken to translate to an
Campbell Scientific dataloggers as follows: equivalent error in dew point, e.g. if the air
temperature sensor is 0.2°C high, then the
1. Measure the relative humidity (RH) and air
estimated dew point is approximately 0.2°C high.
temperature (Ta; units °C).
Figure 2 shows the errors in dew point as a function
2. Compute the saturation vapour pressure (Svp; of a ‘worst case’ 5% error in the calibration of the
units kPa) using Instruction 56. RH sensor.
3. Compute the vapour pressure (Vp; units kPa) For sensors installed in the field there are additional
from Vp = RH * Svp / 100). errors associated with exposure of the sensor, e.g.
sensors in unaspirated shields get slightly warmer
4. Compute the dew point (Td; units °C) from the than true air temperature in conditions of low wind
inverse of a version of Tetens’ equation, speeds and high solar radiation. However, if the RH
optimised for dewpoints in the range -35 to 50°C: and air temperature sensors are installed in the
same shield and are thus exposed identically, the
Td = (C3 * ln(Vp / C1)) / (C2 – ln(Vp / C1)) estimate of dew point is not subject to the same
error as the measurement of air temperature would
where: be. This is because the temperature sensor will
C1 = 0.61078 measure the actual temperature of the RH sensor,
which is what is required for the derivation of air
C2 = 17.558
vapour pressure and thereby dew point.
C3 = 241.88

Campbell Scientific Ltd., Campbell Park, 80 Hathern Road, Shepshed, Leicestershire, LE12 9GX, UK
Tel: +44 (0) 1509 601141 Fax: +44 (0) 1509 601091 Email: sales@campbellsci.co.uk & support@campbellsci.co.uk
www.campbellsci.co.uk
 60

50 55°C
D
E 40
W 40°C
30
P
O 20 25°C
I
N 10
T 10°C
0
T 0 10 20 30 40 50 60 70 80 90 1002°C
E -10
M
P -20

º -30
°CC
-40

-50

RELATIVE HUMIDITY

Figure 1. Dew Point Temperature over the RH Range for Selected Air Temperatures

5
D
E 4
W
3
P
O 2
I
N 1
T
0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
E -1
R
R -2
O
R -3

º -4
C
-5

RELATIVE HUMIDITY

Error + 5RH at 2ºC Error - 5RH at 2ºC Error + 5RH at 25ºC

Error - 5RH at 25ºC Error - 5RH at 55ºC Error + 5RH at 55ºC

±5 RH Unit Error at Three Air Temperatures)

Figure 2. Effect of RH Errors on Calculated Dew Point (±

Program Example 2: Excitation with Delay (P22)

1: 1 Ex Channel
;{CR10X} 2: 0 Delay W/Ex (units = 0.01 sec)
;Program: Demonstration of DewPoint 3: 100 Delay After Ex (units = 0.01 sec)
;calculation 4: 0 mV Excitation
;This example uses an HMP45C Relative
;Humidity and Temperature probe/ ;Measure RH and Temperature mV and multiply
;readings by 0.1
*Table 1 Program 3: Volt (SE) (P1)
01: 60 Execution Interval (seconds) 1: 2 Reps
2: 5 2500 mV Slow Range
;First turn on power to the probe 3: 1 SE Channel
1: Do (P86) 4: 1 Loc [ RH ]
1: 41 Set Port 1 High 5: 0.1 Mult
6: 0.0 Offset
;Use Instruction 22 to force a 1 second settling delay

2
;Turn off power to the probe 13 Z=X-Y (P35)
1: 8 X Loc [ WORK_2 ]
4: Do (P86)
2: 6 Y Loc [ WORK_R ]
1: 51 Set Port 1 Low
3: 8 Z Loc [ WORK_2 ]
;Subtract 40 from Temperature to scale to Celsius
14 Z=X/Y (P38)
5: Z=X+F (P34) 1: 7 X Loc [ WORK_1 ]
1: 2 X Loc [ Air_Temp ] 2: 8 Y Loc [ WORK_2 ]
2: -40 F 3: 3 Z Loc [ DEW_TEMP ]
3: 2 Z Loc [ Air_Temp ]
;And now, as an example, store the time and hourly
6: Saturation Vapor Pressure (P56) ;average
1: 2 Temperature Loc [ Air_Temp ]
15 If time is (P92)
2: 4 Loc [ Sat_VP ]
1: 0 Minutes (Seconds --) into a
2: 60 Interval (same units as above)
;Now calculate Vapour pressure using
3: 10 Set Output Flag High
;VP = RH * Sat_VP / 100
;This equation can be entered directly for Edlog 6+.
16 Real Time (P77)
;Instructions 7 - 8 show the instructions required for
1: 110 Day,Hour/Minute
;older versions of Edlog or keyboard entry.
7: Z=X*Y (P36) 17 Average (P71)
1: 4 X Loc [ Sat_VP ] 1: 1 Reps
2: 1 Y Loc [ RH ] 2: 3 Loc [ DEW_TEMP ]
3: 7 Z Loc [ WORK_1 ]
*Table 2 Program
;Multiply by 0.01 (equiv. to dividing by 100) 02: 0 Execution Interval (seconds)
8: Z=X*F (P37)
*Table 3 Subroutines
1: 7 X Loc [ WORK_1 ]
2: .01 F
End Program
3: 5 Z Loc [ VP ]

;Now estimate dew point using the equation: Input Locations

;Dew_Temp = 241.88 * ln(VP/0.61078) / (17.558 -
;ln(VP/0.61078)) 1 RH 1 1 1
;This equation can be entered directly with Edlog 6+. 2 Air_temp 1 1 1
;Instructions 9 - 14 show the instructions required for 3 DEW_TEMP 1 1 1
;older versions of Edlog or keyboard entry 4 Sat_VP 1 1 1
5 VP 1 1 1
;Multiply VP by 1/0.61078 (= 1.6373) 6 WORK_R 1 3 2
7 WORK_1 1 2 2
9: Z=X*F (P37) 8 WORK_2 1 2 2
1: 5 X Loc [ VP ] 9 _ ________ 0 0 0
2: 1.6373 F 10 _________ 0 0 0
3: 6 Z Loc [ WORK_R ] 11 _________ 0 0 0
12 _________ 0 0 0
10: Z=LN(X) (P40) 13 _________ 0 0 0
1: 6 X Loc [ WORK_R ] 14 _________ 0 0 0
2: 6 Z Loc [ WORK_R ] 15 _________ 0 0 0
16 _________ 0 0 0
27 ______ 0 0 0
11: Z=X*F (P37) 28 ______ 0 0 0
1: 6 X Loc [ WORK_R ]
2: 241.88 F -Program Security-
3: 7 Z Loc [ WORK_1 ] 00
0000
12 Z=F (P30) -Mod
1: 17.558 F -Mode 4-
2: 0 Exponent of 10 -Final Storage Area 2-
3: 8 Z Loc [ WORK_2 ] 0