FLATLAND96: RAD Messages: 19 Entries..

Calibration of Q7's previous to 95, as evidenced by serial
numbers prefixed by 94xxx, are off.

Need to adjust the calibration factors we have by:
	positive multiplier	1.159
	negative multiplier	 .9065

This is an issue for SCMS / MARS95 / MicroFronts also.
Very Nice Fellas....

Because we need to move the radiation stand further away from the
ETI / Campbell than what our cables are, Kurt discovered that for

Licor LI200SA:
A BNC extension having less than 75-ohms or so is OK, and 
that equates to a couple hundred feet of RG.

REBs Q7:
Additional cable of tens of meters is OK, provided the
WIRE INSULATION is NOT PVC

In order to place the radiation sensorsso that they would not be shaded
by the PAM mast, I determined the length of the shadow from curves that
John had prepared showing solar elevation vs azimuth for various dates.
I used the data for Aug 21 (longest shadows for locations north of the
PAM mast), latitude 40 deg N.

Length of shadow, r, for object of height h and solar elevation phi

r = h/atan(phi)

	azimuth	elevation	r/h
	------- ---------	---
	360	63 deg		0.51
	345/15	60		0.58
	330/30	57		0.65
	315/45	53		0.75
	300/60	45		1.00
	285/75	34		1.5
	270/90	19		2.9

The Everest infrared surface temperature sensors are located on the
crossarm at the top of the PAM mast.  The Everest is on the end of
the crossarm that points to the northwest.  It is pointed down at a
roughly 45 degree angle and looking to the northwest. 

I originally intended to point the IRT to the north as suggested by
Wilf Brutsaert.  However since the crop rows are oriented N/S or E/W,
it appeared preferable to look at the rows from an angle of 45 degrees
in order to not be biased by either the crop (looking normal to the
rows) or by the ground (looking parallel to the rows).

The field of view is 15 degrees.  Thus the footprint dimensions along
a line to the northwest of the station are from

	(10m)*tan(45-7.5) = 7.7m
to
	(10m)*tan(45+7.5) = 13m

The footprint dimension in the NE/SW direction is

	r*dtheta = h*tan(45)*(pi/180 deg)*(15 deg) = 2.6m

reinstalled Tsfc sensor on station 3 on lower crossarm.
saw 2 "good" readings, installed sensor, and then received only
-12 deg readings.
removed BPH to see if it would work on station 2

installed datalogger,
tested BPH from station 3 (OK)

6/18/96

Data logger is Campbell CR10 #5350
twh

Because the malcontent Tsfc seemed to work for awhile on station 1 this
morning, Kurt put it back on 3 later on, however, it was still showing
bad values.

We diagnosed the EVE A/D using a 1.3V battery on a 16-pin amp connector
for the tsfc pins, and swapped the battery box output to the AD1 port,
plus tried the Tsfc on channel 0 (AD0 port).  The EVE AD channel 4 is OK
with both the batt, and power box inputs as determined by the EVE ad
monitoring routine.  The Tsfc sensor was still bad, and appeared to be
having stray capacitance that was charging/discharging. 

We'll need to try to isolate this problem tomorrow on station 1 by
lowering the mast and trying to determine whether its the cable of the
sensor. 

Between 96/06/15 ~18Z through 96/06/16 16:40Z.
The Tsfc sensor at station 1 was swapped with the 'bad' sensor from station 3.
This was done to determine whether the sensor from 3 or else the electronic box
there was causing the erroneous values.
It appears that the sensor is OK.
It was mounted overnight on station 1 on the lower cross-arm, pointing SE past
the TRH.

The Licor pyranometer does not go to zero at night, but decreases from
zero at sunset to about -7 W/m^2 just before sunrise.

The net radiation at site 2 is higher than the other stations during
the day and site 3 is higher (negative, but closer to zero) than the
others during the night.  Perhaps this reflects differences in the
surface and vegetation.

Oops.  When i openned the door of the precip gauge to look at the logger the
Licor BNC cable got jostled and the conductor, which had tenuous connection to
begin with, decided to make itself totally broken of at the connector.  I'll
need to repair or replace the cable tomorrow.

The Soldn was rewired in at about 19:40Z, 96/06/26

I was unable to repair the RG58 BNC, so instead i moved the millivolt adaptor
to the sawhorse and soldered an extension from it to the Campbell.  the new
cable is about 50' long perhaps, so it'll be interesting to see if it effects
the voltage measurement.  It probably should be changed back at some point
because i only taped the mV adaptor and soldered wires as a water seal.

Just cleaned mud(!) off the Licor and also cleaned the Q7 dome while I was up
there.  The Licor was caked with lots of mud - I can only speculate that
birds have carried some up there while working on their meals.

I used toilet paper and saliva as the cleaning agents and think I did a 
reasonable job, but others reading this comment may want to let me know
if they think I should reclean the radiometers with something different.

It was a good view of the site from there - see the next comment.

		visit time: 15:15 - 15:45 GMT

  The licor was acting up again last night. It seemed to
work fine during the day time after cleaning the sensor.
However it read negative values at night. A simple test
was done by covering the sensor with a black cap. The output
went to -25 W/m^2. The sensor was cleaned again and the black
cap test was run again. The output was the same. Will monitor
the sensor over the next 24 hours.

During our site visits today, I checked the radiometers (by ladder) at 
PAM2 and PAM3.  Both nets were a bit dirty (3 had bird poop and 2 had mud
drops), so I cleaned them.  I also gave the Licor's a wipe, though they
looked pretty clean anyway.  

I also releveled both nets.  PAM3 had been pointing up to the SE by about 
1/2 bubble and PAM2 had up to the E, by an unremembered amount.  Both 
Licors were level.

Both radiometers on PAM1 were pointing up to the NE by about 1/3 bubble
when I cleaned them 3 days ago.

The polarity was reversed so + is now into the high channel.  The gain in
the Campbell was changed to +94.0 from -93.56.  If the logger loses power
and starts up with the old gain, we will know it because the signal will
be negative!

Hopefully this change will help our nighttime negative values.

It was 1/2 bubble up to the NE.

Because of the low values from the PAM1 licor, I cleaned it about 45 minutes
ago.  However, it looked pretty clean already, so dirt may not have been
the problem.  It is too cloudy now to know if the cleaning worked.

I'm out for the first time tonight and notice that the full moon (through
thin cirrus clouds) is reasonably bright - possibly enough to register on
the Licors.  Thus, don't treat values around now as a bias!