�From:�	Johan Smit ("smit.js")
�To:�		SOWACS@aqua.ccwr.ac.za
�Date:�	Saturday, June 22, 1996  4:26 pm
�Subject:�	Sensor specifications

Trevor Finch has stipulated the specifications of soil moisture 
sensors under six headings. I agree to most of it, but havetrouble 
with the following:
>4.  It should be able to be installed reasonably easily intoundisturbed soil.
>
>6.  It should measure a sufficient volume of soil that readingsare not
>unduly influenced by preferential drainage paths caused by sensor
>installation (eg water running down tubes or cables), or by local 
>root activity.
It appears to be physical impossibile to satisfy the above two
criteria simultaneously. 
Consider that the field strength diminishes exponentially from the 
sensor, and thus the readings are mostly affected by physicalchanges 
against the sensor. This is physics, and all electronic sensorswill 
abide by that law. 
If the sensor then is  vertically implanted, it does not conformto item 6.
If a vertical cable is used to feed sensors at various depths,then item
4 is the problem.
Forgetting for the moment anything else, what would physically be
the ideal sensor group? 
A cable feeding seperate sensors?
How long should a sensor be, if the sensor is similar to a pieceof  
cable? This is assuming each sensor itself will be planted 
horizontally at the chosen depth.
How could you then install such a sensor group, regarding item 4?
Perhaps this is what should have been done before anything else,
specifying the sensor dimensions and only once that is refined
and clear, look at the next step.
Debate ?
Regards
Johan Smit