CHATS: trh Messages: 32 Entries..

The chronology of the S/N101 and S/N 502 TRH sensors for CHATS is below.  
Also see logbook entry 128.  These two sensors were both determined to 
require empirical corrections as described here:

 ------------------------------------------
 Date: Mon, 29 Oct 2007 12:37:24 -0600
 From: Tom Horst 
 To: Sean Burns 

 Sean:

 I think that the original TRH at 6m was S/N 502, which we confirmed in 
 the lab had an error of -0.3133 C.  Since we confirmed this error in the 
 lab, we should add an offset of 0.3133 from Mar 17 until Mar 21.  Note 
 that this is an unusual circumstance where we know the magnitude and 
 duration of an error.

 A similar correction can be made to S/N 101 which was on the horizontal 
 array (upper) and was found in the lab to have an offset error of -0.099C.
 Thus an offset of 0.099 should be added to the data for this sensor.

          [details of the location of S/N 101 are below]

 --------------------------------------------

Chronology of sensors:

----------------S/N 101:

 - started on horiz array (upper position)

 - on 4/3 at 14:30 PDT, sn101 removed from ha/upper
   and replaced with sn206 (LOG 93)

 - on 4/3 between 16:30-17:30 PDT, sn101 back in at
    ha upper position.(LOG 93)

 - on 4/3 at 18:00 PDT, sn101 moved to 6m on vt (LOG 93)

 - on 4/12 sn101 at 6m replaced with TRH sn502 fresh from recal 
   in Boulder, sn101 now becomes "spare" sensor (LOG 145)

 - on 5/2, sn101 was recalibrated in BOULDER.  
   (after this date correction is no longer necessary)

 - on 5/11 sn101 was put in at 14m (LOG 189)

 - sn101 stayed at 14m for the rest of the experiment.

---------------S/N 502:

 - started on vertical tower at 6m (LOG 128)

 - on 3/21, sn502 at 6m replaced with sn206 (LOG 31)

 - on 3/23, between 11:30-16:50 PDT, sn502 was put on the vertical tower 
            at 1.5m to test the humidity profile.  

 - on 4/9, sn502 was recalibrated in Boulder/NCAR.
   (after this date correction is no longer necessary)

 - on 4/12 sn502 deployed on vt at 6m (LOG 145)

 - on 5/22 sn502 was moved up to 7.5m (LOG 280)


   

TRH at 1.5m was out starting from about 4pm today.  Cycled power at both the
data logger and at the TRH at about 11pm.  Output from 'rserial /dev/ttyS12
lower' suggests that it is now working again.

6/6  17:57 PDT

Installed TRH702 on the tower at 3m to replace 
TRH004 which was moved to the array today.


702 is the unit that has previously shown a problem with the RH
values, but we have no spare at the moment.

The upper TRH (/dev/ttyS15 at west) died during the array lowering
at around 10;50 LT.

The lower TRH (TRH704) works in port 15, so the problem must
be in the cable or the TRH.  Ned pulled the TRH out of
the boom and reseated the connected, but still nothing.

Will work on it in the base trailer.

4:25-4:35: Replaced with TRH 004 from 3 m level on vertical tower

TRH Serial Numbers:

S12 lower (1.5 m): 703
S13 lower (3 m):   004
S14 lower (4.5 m): 007
S15 lower (6 m):   204
S16 lower (7.5 m): 502
S17 lower (9 m):   001
S11 upper (10 m):  008
S12 upper (11 m):  006
S13 upper (14 m):  101
S14 upper (18 m):  005
S15 upper (23 m):  701
S16 upper (29 m):  104

S14 west (bottm):  704
S15 west (top):    206

Moved TRH's to mid-canopy from 9:44-10:15, completing the configuration change. Boom heights = 5.41 & 6.41 m

30may07  15:40 ... 16:42 Center tower TRH fan


The noisy fan in TRH.top.array was replaced between 1540 and 16450LT.
After moving to narrow, and raising the array yesterday, we observed that
this fan became noisy.   Today after installing the pressure port on the
array, this little devil began to really complain, so at about 15:40 we
took it down to swap in the best available fan we could find.

Note: once apart, it was apparent that the fan blades had come off
and were in the rather dusty exhaust tube.
This may have happened during the move yesterday because the fan had
appeared ok, and Greg hadn't seen any issues in the data.

By 16:40 the sensor was back up with a good fan.

Ned noted a funny fan sound in T/RH.6m. No obvious impact on data and we are
monitoring.

23may07 10:30-11:00;   12:05-12:20 Tower Climbed

6m TRH swapped out, back in, filter installed:
Originally at 7.5m until last night, trh had been showing 
a broad-band frequency response that at first was thought
to be a feature in the canopy/atmosphere interaction.
That sensor was swapped with 6m last night, but Greg and Ned
observed that it retained the 'feature' so it was removed today.
However, the 'new' unit wasn't working, so we put the original
unit back up.......meanwhile.....Geez Louise....
....noticed gortex filter missing on culprit trh!
Don't know how long that was the situation, but now it has a
filter from the 'spare.'
We'll keep an eye on this to determine whether the missing filter
was the only problem, or if it has been effected by operating without
a filter.


Andy Turnipseed's 14m inlet
Briefly installed a o3 filter for him to check his instrument,
closer to 10:50, for roughly 1-min

On the basis of entry 280 in the logbook we will proceed with using a TRH spare
at TRH.6m on 23 May.

Per ongoing T/RH investigation and Tom's email we feel that there may be a 
non-physical low frequency undulation in T.7.5m including a time lag. We were
unable to detect this behavior in RH.7.5m. 

Today at 1455LT we exchanged the 6m and 7.5m TRHs to see if the feature 
currently at 7.5m changes position.

Start of tower 1455LT
End of tower 1515LT

Result: The feature moves. Not notable in RH measurement. Notable in T however.
(spectra and ts). Will change it out tomorrow (the 6m TRH).

21may07 11:45-11:50

TRH swapped out with the newly calibrated / delivered unit from Boulder.
The one removed had been the culprit at 3m with the high-freq. noise.

S-plus shows that the 3m noise remains in place. Thus we conclude that this is
one of the sensors (50Ys) that Steve S. suggests have this behavior and that we
put another one in.

However, the RH and MR values (profiles)
are much more physically realistic and now, consistent with two separate
h2o profiles from ACD (1.5 and 4.5 m) with height. The 'notch' from yesterday
is gone and the 'bulge' from previous days is gone and replaced with a smoothly
varying h2o profile!

Basically, the 3m RH seems to have been high by ~2% until this time.

Between 1540 and 1610 the 1.5m and 3.0m TRHs were exchanged. The purpose of 
the exchange is to discern whether the mixing ratio maximum showing at RH.3m 
(relative to 1.5m in particular, where soil moisture was expected to contribute
 moisture more effectively) is real or sensor-related. We expect a 'notch' to
to develop in the profile if the original 3m sensor was reading high or if the
1.5 m sensor was reading low (in creating the current condition).

Result: The notch was created. The exchanged sensors read within 2% RH however,
which is the accuracy of the sensor and so little can be done (unless the 
calibrations are unified to eliminate the bias).

Ned checked with the ACD team regarding their h2o profiles to see if the low
elevation peak at 3 & 4.5m in mixing ratio was real. It was confirmed in their
profiles. This appears to indicate a convergence of water below canopy and 
above ground. Very interesting ... is this counter-gradient transport? Given the
reverse shear going up to the canopy, I don't think so, it is just that the
shear is reversed from our expectation.

In the course of an investigation of the low mixing ratio at 1.5m on the tower
(compared to the 3 and 4.5m levels) we looked at T at those 3 levels.

High-rate plots (1Hz) showed highly noticeable noise about the T.3m signal 
(not in the RH.3m signal). We tried one idea related to Xbee radio 
interference to no avail.

Given the earlier fan problem in TRH.3m we are changing the fan. This resulted
in an outage from 1740-1810LT and had no effect.

A search back to 15 April confirmed that this noise has been present since at
least then (data storage here is limited to 15 Apr - the rest is at home). 

From the above we conclude that the sensor itself is unstable and should 
be replaced. Check with Tom.

1445LT re-install the intake for the tower T.9m
   - T.9m now operating well

The fan had quit. It turns out that it was caught and by re-rounding the intake
it worked again. Will monitor for intermittent behaviour.Installed 0940LT. Data
looks great.

Problems have recently cropped up with the 9, 14, and 23 m TRH's.

9 m TRH is too high during the day, suggesting a fan failure.

14 m TRH has bad data (T=-40, RH=-1.6).  It appears as though there is a bad
connection between the sensor and the A/D.

23 m TRH is not communicating with the data system.  Perhaps a power recycle
will get it going properly.

Since this T has been 0.1 deg low, it was replaced by the sensor just shipped
back from Boulder.
0101 removed 15:16
0502 installed 15:34 (but wait until 15:40 for fan to cool it down)

It took a while because the replacement fan was plugged into the wrong side
of the connector, which obviously had to be fixed.

Please enter this record of the laboratory recalibration of S/N 502 in
the CHATS logbook:

S/N 502 was originally at the 6 m level on the vertical tower and was
replaced with S/N 206 on March 21 (entry 31).  Ned and I had detected
the -0.3 deg C error in the profile plots.  We will send it out to
replace the TRH currently at 6 m (S/N 101) which we removed from the
horizontal array (entry 93) because it appears to be in error by -0.1
deg C.  Send S/N 101 back for laboratory recalibration.

Thanks,
Tom

-------- Original Message --------
Subject: 	T check on TRH502
Date: 	Mon, 09 Apr 2007 08:48:37 -0600
From: 	steve semmer 
To: 	Tom Horst 

Here is the result of the temperature check on TRH502. It had a mean
error of -.3133 C (sensor - ref). If we want we could use this as a
correction to the data when this sensor was deployed. New coefficients
will be computed.

Steve

8/20/07:

The -0.1 degree temperature error in S/N 101 was also confirmed in the
NCAR Calibration Laboratory:



-------- Original Message --------
Subject: Post CHATS TRH info
Date: Tue, 07 Aug 2007 12:06:23 -0600
From: steve semmer 
To: Tom Horst 

Attached is .txt file showing the post cal results for the TRH sensors.
The RH were checked over 10% to 90%  (20% steps) at 25C.
The temperature were checked over the range of 0 to 40 C (10 degree
steps).

Steve

CHATS TRH POST CAL					

TRH	mean T error (sensor - ref)    mean RH
error (sensor - ref)
	apr/may	jun			
704		0.004			0.45
5		0.084			2.02
204		-0.052			1.64
206		0.016			1.39
502	-0.313	-0.062			0.76
7		-0.018			3.25
6		0.035			0.14
104		-0.030			1.85
702		-0.001			1.53
701		-0.061			0.08
1		-0.008			0.26
8		0.043			-0.31
4		-0.008			1.02
101	-0.099	-0.006			0.56
3		0.043			-1.01
703		0.251			0.05


All sensors were calibrated pre CHATS					
Sensors in the apr/may column were re-calibrated on the following dates:				
 S/N 502 -  4/9/07
 S/N 101 -  5/2/07
 

The TRH heights are 6.40 - 0.47  = 5.93 m and 5.41 - 0.47 = 4.94 m

Don and I compared the horizontal array TRH's to those on the vertical tower
in an attempt to determine which of the two might be out of calibration.
The results were somewhat inconclusive, but with a tendency for the TRH at
the greater height to perhaps be reading too low.

Perhaps a more definitive answer could be obtained by first filtering the
data by wind direction.

This afternoon, we swapped the 6 m TRH (s/n 206) with the top (10.6 m)
TRH (s/n 101) on the horizontal array:
14:00 removed 6 m TRH
14:15 mounted in at 10.6 m on horizontal array
14:35 mounted 10.6 m TRH from horizontal array at 6 m

Don and I concluded that S/N 206 is reading 0.1 degree too low

Removed 10.6 m TRH (101) from the vertical tower at 16:03.
Removed 9.6 m TRH (704) from the horizontal array at 16:11
Moved 6 m TRH (206) to 9.6 m on horizontal array at 16:16
Reinstalled 10.6 m TRH (101) at 10.6 m at 16:18
Installed 9.6m TRH (704) at 6 m on vertical tower at 16:23

The 9.6m TRH looks okay, so swapped sonics back to previous
configuration:

Removed 9.6 m (704) from vertical tower at 17:30
	Required some repair of joint between aspirator and electronics
Removed 10.6 m (101) from horizontal array at 17:35
Lost track of time
Moved 6 m (206) up to 10.6 m
Mounted 9.6 m (704) back at 9.6 m
Mounted 10.6 m (101) at 6 m on vertical tower
Finished at 18:05

Final configuration:

6 m on vertical tower = original 10.6 m on horizontal array (101)
which is ~0.1 degree too low.

9.6 m on horizontal array is original (704)

10.6 m on horizontal array came from 6 m on vertical tower (206)

Heights of TRH booms - height of inlet below top of boom (47 cm)

10.12 - 0.47 = 9.65
11.14 - 0.47 = 10.67

Yesterday Steve fixed the lower TRH by unplugging and plugging back in the
50Y at the board.

Lowered TRH's down to 2.46 - 0.47 - 1.99 and 2.46 - 0.47 = 2.99 m.
Finished at 15:25, started a little while after finished lowering horizontal
array and coiling ropes.

Lowered TRH's for horizontal array to 5.41 - 0.47 = 4.94 m 
and 6.41 - 0.47 = 5.94 m.  
Started at 4:00 PDT and finished at 4:50 PDT.
Marked previous TRH heights with cable ties at tops of boom mounts.

Installed TRH 502, with bad T, at 1.5 m to test RH.
Installed at 11:02 PDT.

16:55 PDT: Replaced original TRH, S/N 004, at 1.5 m.  
17:40: Took several unplug/plug sequences to get it going.

Steve replaced the 6 m TRH with the one removed from horizontal array, 
S/N 206.

The original 6 m shield had a loose inner tube, so moved the 1.5 m shield to
6 m.  The 1.5 m trh has no shield at this time.

3/22:  The temperature profiles look much better now.

At some point today Steve fixed the shield and installed it at 1.5 m.

The 6 m T appears to be ~ 0.3 deg C too low;

Either the 3 m RH is high or the 1.5 m RH is low....

note: the original entry should have been:

"Steve replaced S/N 206 with S/N 704.  Data look good."

TRH serial numbers were checked with "data_dump"
and confirmed that S/N 101 was in the upper position on 
the array, and was not swapped out on 3/18.  

update by SpB
Tue Nov  6 15:51:21 MST 2007


-------------------------------original logbook entry:

Plug at adam for lower trh was loose, but this did not fix trh.  Rserial does 
not contact trh, suggesting board is bad.

14:25
Steve replaced S/N 101 with S/N 704.  Data look good.

30 m tower heights - clone of entry 14
Heights are (m agl) to top of boom - 32 cm.
Note sonic is 5 cm above boom and TRH inlet is 47 cm below boom.

Port	Location	S/N	Height	Height(AGL)	Height	Height(AGL)
			Sonic		(+37 cm)	TRH	(-15 cm)
--------------------------------------------------------------------------
S10,up	29 m		0744	28.66	29.03		28.66	28.51
S9, up	23 m		1117	22.70	23.07		22.70	22.55
S8, up	18 m		0738	17.67	18.04		17.67	17.52
S7, up	14 m		1123	13.72	14.09		13.72	13.57
S6, up	11 m		0364	10.76	11.13		11.27	11.12
S5, up	10 m		0671	9.66	10.03		10.19	10.04
S2, up	9 m		0536	8.65	9.02		9.19	9.04
S7, lo	7.5 m		0540	7.25	7.62		7.80	7.65
S6, lo	6 m		0740	5.70	6.07		6.22	6.07
S5, lo	4.5 m		1120	4.17	4.54		4.69	4.54
S2, lo	3 m		0674	2.74	3.11		3.26	3.11
S1, lo	1.5 m		0673	1.17	1.54		1.70	1.55

3/22:

Measured heights of booms for TRH's on horizontal array:

10.13 m - 0.47 m = 9.66 m
11.14 m - 0.47 m = 10.67 m

We mounted the two TRH's on the center tower of the horizontal array yesterday.
We got data for some 10's of minutes for both, then the 9.6 m TRH stopped.
No T or RH data.  10.6 m is okay.