2.3  Protocol

What is protocol?
Should the sensor be painted?
Why and how should the sensor be painted?
How many sensors should be used?
How is output from multiple sensors interpreted?
Where should the sensor be located?
What is the best compass and vertical orientation of sensor?

Next section scale
Return to table of contents

What is protocol?

Protocol is the uncertainty that arises when an instrument is used with different observers or methodologies for measurement (Russo et al. 1993). The following protocol considerations are discussed.

Should the sensor be painted? 

Yes. Painting improves both the precision and accuracy of the sensor, see the discussion in these sections. See also Gillespie and Kidd 1978 and Magarey et al. unpublished #2).
 

Why and how should the sensor be painted?

Grid sensors are often painted to make their response closer to that of real leaves. A latex based paint will absorb water allowing droplets smaller than the gap between the fingers of the grid to create a sensor response (Gillespie and Kidd 1978). Another important influence of painting is that the sensor more closely mimics the thermal response of a leaf. White is the most suitable color, although for use throughout a season, light gray is less likely to discolor than off-white (Gillespie and Kidd 1978). While painted sensors may be more useful for measuring wetness periods caused by dew, it is unclear whether they are the most suitable for recording wetness periods caused by rain.

The following service provider(s) has been used by people in our group.
 

How many sensors should be used?
 

Ideally, we can suggest three sensors. Penrose et al. (1996) found that three sensors would have a 92% chance of detecting a wetting event, whereas two sensors would only have a 60% chance. Magarey et al. (unpubliehed #2) found that for a single painted sensor, the precision of estimating a wetness event would be as close to 1.5 hours, whereas for three sensors it will be almost half that value.

Where should the sensor be located?

Most reports suggest the top of the canopy. The top of the canopy is the first to wet during rain or dew events and in most grape canopies there is a fairly uniform drying. In grape, the top may have greater SWD although in most instances there was no significant difference between different parts of the canopy (Magarey et al. unpublished # 2, Penrose et al. 1996). The location of the sensor in the canopy was of secondary importance because the sensor has a precision between 1.5 to 5 hours whereas variation due to canopy position was closer to 0 to 3 hours in the few cases where it was important (Magarey et al. unpublished # 2). The precision of the sensor also appears to be greater at the top. Other researchers have also suggested siting sensors at the top of the canopy in tomato (Potraz et al. 1994), bean (Weiss et al. 1989) and apple (Wittich 1995). Other suggestions for siting sensors include over adjacent turf (Potraz et al. 1994).
 

What is the best compass and vertical orientation of sensor?

Vertical orientation
 Recently Lau et al. (in press) found that vertical orientation did not significantly influence affect the accuracy or precision of painted sensors for dew estimation, another reason to paint the sensor.  We suggest the sensor should be tilted on it short axis at a 45o angle to prevent excessive pooling of water during rain.  However, other researchers have suggested the sensor should be mounted in a horizontal position.

Compass orientation
The sensor should face in the compass orientation likely to give the longest SWD duration. In the northern hemisphere a sensor orientated to the north may be preferred because it will receive less solar radiation than one orientated to the south. Another factor may be the prevailing wind direction. In Oklahoma, it is recommended that sensors should face east so that prevailing humid westerly winds do not prolong the drying of the sensor (Fisher et al. 1992).  
 
Next section scale
Return to table of contents