Austin Cary (Gainesville)

Installation:  December 2009

• Installed Picarro Cavity RingDown Spectroscopy system CADS06 temporarily at site
• Tanks from the Round Lake (?) site used as reference tanks since the CADS06 tanks were low

December 8, 2010

• Changed system from CADS06 to CADS09

May 21, 2012

• In order to stabilize temperatures: redirected airflow from air conditioning unit & shielded instrument
• Added 24″ SS Nafion dryer with compressor for heatless dryer for purge gas

May 15, 2013

• Opened pressure relief valve on compressor tank to purge water.  Tank may have been 1/4 to 1/3 full.  Has been approximately 1 year since the compressor was installed.
• Cal 1, CA08089, 1600 psi, Picarro measured 347.78 ppm CO2
• Cal 2, CA06438, 650 psi, Picarro measured 394.47 ppm CO2
• Changed Cal 1 from 12 min to 6 min cal cycle.

Sept 16, 2013: Scott on site to replace dead pump. Installed MD1 pump. Emptied water out of compressor reservoir. Restarted system and seems to be working well.

Sept 17, 2013:  Gas flags are different here than at the INFLUX sites.  Doing 0 (sample), 4 (ref 1), 8 (ref 2).

October 3, 2013:  Ref gas error jumped from -.3 ppm to -2.2 ppm between DOY 240 and 260, at the time the pump was replaced.  Suspected nafion leak.

Jan 14, 2014: Scott on site.  CADS09 was failing, Picarro not able to fix remotely.  Put CFKADS2060 at the site and shipped CADS09 directly to Picarro for repair.  NAFION apparently was leaking because when I plugged the outlet of the dry air the flow meter did not decrease, so I took it out of line and we are no longer drying at the site.  I brought the nafion home and verified while on site that the inner tube and outer tube did indeed communicate/leak.

April 14, 2014:  A couple weeks ago, had problems: “CFKADS2060 is having an issue that we’ve not seen before. When we checked on it, it was measuring ~18 ppm for CO2, which doesn’t make any sense, of course. After restarting (and also resetting the analyzer, same effect), the pressure stayed at ~2-3 Torr with the outlet proportional valve pegged low at 21000. Ideas on what is wrong?” Danthu response:  “It sounds like there is no flow through the analyzer. It looks like there is some restriction upstream of the cavity that doesn’t allow flow into the cavity.”  Rosvel found clogged flowmeter, bypassed it, and the instrument seems fine now.

April 21,2014 about 2000 UTC:  Installed Nafion. Old one was leaking and had been bypassed.  After install water went from ~1.5% to 0.1%.  Emptied small amount of water from Porter cable compressor.  Verified plumbing from tower, seems to be like it has always been.  Dry air flow about 0.5 lpm.

 Decommissioned

Canaan Valley

Installation:  18 April 2006 at tornado shelter site
Off-line:  ~ June 2006
Re-installation:  7 Sept 2007 at new site with trailer
Pump replaced (S. Brooks): 10 Feb 2009 (clear tubing, not Bevaline used)
Pump replaced (SR/NM):  12 Jan 2010
Heat distribution fan replaced (SR/NM):  12 Jan 2010
System tested with Picarro, baby jar test (SR/NM):  15 April 2010
Inlet repositioned from trailer to 10-m AGL tower:  ~20 April 2010
LICOR-820 replaced (S. Brooks): 14 Oct 2010

22 October 2013:  NOTE:  Loggernet gets data from this site, has to be running.  CanaanJan222010.dat is copied from calibration computer in Scott’s office to /eddy/s0/AmeriFluxProject/data/CanaanValley/ using WinCP (something like that).  If nlm136 password changes, Archive.bat (scheduled task) will have to be adjusted.)

22 October 2013:  System  currently off-line (since DOY 263 in 2012)

Chestnut Ridge

Installation:  14 September 2006

  Returned for repairs:  18 October 2008

  Re-installation:  December 2009

  Pump replaced:  12 August 2010

  LICOR 820 replaced: 20 January 2011

  Lightning strike (LICOR 820 damaged):  11 July 2011

LICOR-820 replaced (M. Heuer; serial number 451 owned by ORNL): 9 April 2012

 Repair visit: 3-4 December 2012.  Notes:

• Moved the drying columns outside on right hand side of box
• Replaced LI820, new SN is CGA-720.  Old one was not ours and number CGA-451.
• Replaced galman filter with a new one.  old one was full of water.
• Replaced drying columns with new ones (the old ones were completely spent and perhaps the reason there was water in the filter/pump).
• C4 and Target beswik fittings (screws) broke in the process of trying to fix leaks.  Thus removed archive and Cal1 and just have cal 2, 3, 4 and target connected to manifold.  Will have to do 3 tank calibration. Screw shafts are in the manifold still and are therefore open to atmos.  However, those valves are not powered (unplugged the clip from the solenoid valve) so it should be just like having the guest cal open to atmos all the time).
• When sending replacement drying columns send them so a 1/8 swagelok can go on it (i.e., put the 1/4 to 1/8 adapter on them before sending them).

22 October 2013:  NOTE:  Loggernet gets data from this site, has to be running.  TennesseeDec2009.dat is copied from calibration computer in Scott’s office to /eddy/s0/AmeriFluxProject/data/ChestnutRidge/ using WinCP (something like that).  If nlm136 password changes, Archive.bat (scheduled task) will have to be adjusted.)

March 2015:  Decommissioned

Missouri Ozarks

Installation:  1 October 2006

• Temperature in building appears to be stable at approximately 25 deg C.
• Sample line is Synflex 1300 into the box.
• Program in the logger is the same as used at WLEF for the Picarro tests. That is, everything is taken out except for the essential measurements to save space.
• Downloads will occur once a week and the current program has 11 observations per 10 second interval which gives about 10.5 days of storage available on the logger.
• All regulators set to approximately 9 psi.
• Labbox system and the MO system (called Box4) differed by about 0.5 ppm in the lab and in the field.
• I used a new technique to detect leaks into the sample line upstream of the manifold by using a small (~500cc) glass jar flooded with gas from a known cylinder. In this case I used an extra cylinder that I brought along to replace the target tank since it is low. I tested the two systems the last hour of the test4.dat files and saw the two different by 0.5 ppm again. One system is about 0.25 ppm high (labbox) and one system is about 0.25 ppm low (Box4 deployed at MO).

Maintenance visit: 28 April 2006

• Replaced target with 382.67 ppm tank
• Brought back 2 empty target tanks
• Performed bottle test – agreement within ~<0.2 ppm
• Had leak in delivery pressure – similar rate in all tanks (??)

19 Sept 2012:  instrument turned off awaiting repair

Dec 2012

• Replaced Cal 1, Cal 2, Cal 4, and Target.  Target tank pressure was 1000 psi.  Tanks CBO9177 and CBO9167 are on-site as well, but not being used.
• Replaced LICOR-820?? (CGA???), replaced mol sieve?? (mounted outside box), replaced pump??, replaced nafion??

24 April 2013

•  Installed different system

22 October 2013:  NOTE:  Kevin Hosman sends an e-mail (quasi-weekly) which gets filtered and the data is copied to /eddy/s0/AmeriFluxProject/data/Missouri/2013/.

13 April 2015:  Adam Halley replaced target tank with CB09177 (395.05 ppm).  Tank CB09167 also on site.

Mead
12 March 2006:  Installation
December 2011:  Sent new pump and new LI820
May 2012:  Sent replacement tanks Cals 1-4 and target (2 tanks)
May 2013:  UNL personnel replaced pump provided by PSU

Fort Peck

Installation:  6 December 2006

• The system is located inside a shed approximately 100 m from a 5-m tower with flux measurements.  This is a Tilden Meyers site like WV, TN, and MO. Data will be collected by Tilden’s group every hour and posted to their site daily.  Station time is checked and automatically synchronized to Mountain Standard Time.  The shed is NOT climate controlled.  Sample inlet is at approximately 10 feet AGL and run to shed 100 m to NNE. Sample line is Synflex 1300 into the box.
• Tanks used are from Scott-Marrin and just received and calibrated directly before installation. This is so that remaining ESRL tanks can be used with Piccaro Inc. systems during RING II as part of the Mid-Continent Intensive 2007.
• Program in the logger is the same as used at Chestnut Ridge and WLEF.  The target tank is sampled every 45 min after the hour.
• All regulators set to approximately 9 psi.
• To detect potential leaks into the sample line upstream of the manifold, a small (~500cc) glass jar flooded with gas from the target cylinder was sampled. The data for this test is the last hour of the MTtest2.dat file and the error less than 0.5 ppm.  See plot of difference between calibrated [CO2] and known value.
• The enclosure is not temperature controlled and may get too cold at times; it was not possible to connect the heater and temperature control module for logistical reasons.  Data is transferred from box in shed to flux tower via long serial cable which lies next to Synflex sample line on top of the grass and snow.  Sample inlet could be mounted on 10 m (?) SurfRad (?) tower but line has to be trenched because it has to cross a driveway (ground is obviously frozen during this installation). If that is done, the temperature-control system can be installed at that time.

Re-Installation:  12 December 2007

• Because the electronics enclosure is inside an uninsulated shed, the enclosure is exposed to extreme environmental temperature changes.   This was causing large variations in moisture levels (over time periods less than the calibration frequency) and thus CO2 mixing ratio errors greater than 0.5 ppm.
• Thus the system was replaced with a new one fitted with a heater/cooler unit manufactured by ISC to heat the box when the inside temperature drops below 60 F and cool the unit when the temperature rises above 78 F.  Minco heater not installed.
• Baby jar test performed with cal gas 2 (353.93 ppm)

De-Installation:  March 18 2009

• Due to the remoteness of the site and lack of a climate controlled building/enclosure, sufficiently accurate CO2 measurements could not be maintained from this site and it is therefore being decommissioned.