Tuesday, May 12, 2015

Summary of XPIA BAO Site Visit, 11 May 2015

1) WC61 was offline for the most number of days the previous week. It's stage head was rotating upon arrival, but stopped as soon as a Remote Desktop session was attempted. System was rebooted via the power switch. It was then reconnected to Cellular data network after a Verizon software coverage update. JKL confirmed login through TeamViewer.

2) WC68 was taking data and able to logged into via Remote Desktop Connection. Cellular data connection and related software programs were restarted and all worked fine on the device.

All data corresponding to NWTC 4.0 pad Fall 2014 was copied and removed from data hard drive, including empty of Windows Recycle Bin. However, RAW .dsp files were for this time period were not copied over (very large), but were moved to Windows Recycle Bin and not removed.

The above actions reduced disk usage to ~64%, which is much lower than the usage upon arrival.

3) WC16 v2 offshore system is so quiet, it is difficult to know if it is running when making a site visit! Neither Mozilla Firefox (windweb) nor a ping from a command window to the IP address could connect to the system.

Power was shut down momentarily to the system, and then a reboot occurred. Both the windweb software and the FileZilla data transfer worked beautifully. Although the analyst will have to look into the data files, a cursory look into the system showed little to no missing data, except for the brief time of power down today on 11 May.

4) ~3.05 GB of data from the AXYS Vindicator LiDAR system Compact Flash Card was copied today. PTQ updated the Google Drive folder to include the additional data.

--Electrical tape was put over a few outlet spots. Power cords sustained the Colorado rain & snow deluge of the power week.
--All 3 WC and the AXYS lenses were manually wiped off.

Tuesday, April 14, 2015

Converting .rtd to .ncdf

Steps:

1. Install LabVIEW on local machine at http://www.ni.com/download-labview/ (Installation may take a while, and students can get student version for free by contacting OIT, please see more information at http://www.colorado.edu/oit/software-hardware/site-licenses/labview).

2. Copy ConvertRTD.exe from breeze at /data/fielddata/WC_Info/Software to local machine using scp, FileZilla, etc..

3. Copy .rtd files of interest to local machine from breeze.

4. Run ConvertRTD.exe on local machine.

5. Choose converting one file only ("Rtd file path", option on top) or converting .rtd files in a folder ("Directory" and "Destination Folder", option at bottom), by clicking the switch in the middle. "Current folder" is an option on the bottom right once you are in the directory.

6. Click "START" to convert .rtd files into "Converted_xxx.rtd" files.

7. Copy "Converted_xxx.rtd" files back to breeze, say, /home/student/your_id. Make sure different WC rtd files are in different folders, say WC61's in one folder "/home/student/your_id/XPIA/WC61", to avoid making multiple .nc files of same time but from different WCs.

8. Copy "rtd_to_cdf" from /data/fielddata/WC_Info/Software to same directory, say, /home/student/your_id

9. Run "rtd_to_cdf" by typing "./rtd_to_cdf".

10.1 Type in directory containing "Converted_xxx.rtd" files, say, "XPIA/WC61"

10.2 Type in prefix, say "XPIA_"

10.3 Type in observed height levels of windcube, which are typically 40 to 220 m of 10 different heights, so type "10"

10.4 Type in time adjustments. If WC was taking measurements at UTC times, then type "0"

11. The program should say "Finished!" at the end and you can find .nc files in directory.

Written by Joseph Lee
10.50 14 April 2015

Saturday, April 4, 2015

XPIA BAO WC61 & 68 Basic Maintenance and Troubleshooting

On 2015/04/03 between about 1500-1600 MDT, Paul Quelet serviced the WCv1 LiDARs running at the BAO during XPIA. Next day (2015/04/04), have checked both systems remotely. Seem to be taking more vertical data and have stronger cell signals.

WC61 (NCAR):

  • 2149 45-55 seconds (approximately), wiped lens manually. Should be NaNs in data files.
  • Lens wipe seems to have helped LiDAR remotely sense more vertical levels? (So far this seems to be the case...
  • Still running very strong.
  • Usually transferring data (although not always full file size).
  • Strong Verizon dongle signal (usually ~3 bars).
  • System requested Verizon software update upon login 2015/04/04. After saying "Yes", was not able to update, and said update will come back in two days. Should happen ~0830 MDT 2015/04/06?
  • Sun cover on.
WC68 (CU):
  • 2213-2215 installed WC49 cover on top of WC68. Put Verizon dongle upward to extend length and wiped lens off manually.
  • Lens wipe seems to have helped LiDAR remotely sense more vertical levels? (So far this seems to be the case...
  • Running well.
  • Manually downloaded data from 2015/03/20 to 2015/04/03 through remote desktop connection.
  • Has had trouble transferring data remotely or allowing TeamViewer login.
  • 2015/04/04 0000 UTC FTP data transfer seemed much improved with dongle position change.
Yes, I am aware that this is the non-matching sun cover number on the WC. However, it was quite a feat of strength to get the cover out of the Leosphere box and transport it to the site alone.

Tuesday, March 31, 2015

Summary of XPIA AXYS Vindicator LiDAR Deployment & WC Operations, 30 March 2015

AXYS Vindicator LiDAR Deployment
  • Yanis Gryshan led the way to install two AXYS Vindicator LiDARs just to the east of the LiDAR supersite at the BAO, across the dirt driving path from the two WC v1s from CU, the Halo scanning LiDAR, and the NRG/Leosphere v2 profiling LiDAR.
  • The Vindicator lidars stand on Lundquist group WindCube flat platforms.
  • Each system has a serial number, noted as "13" and "15", these being AXYS generation 3 systems.
  • Single power cord run across the road with orange plastic fencing noting the placement. Power cable connections secured with electrical tape.
  • 10-minute averaged data can be accessed in real time through AXYS user login website, which Yanis and colleagues will disburse.
  • Data is distributed through cell network in real time. Also system has satellite connections available if cell network goes down.
  • (Much can be done by AXYS remotely once systems are running.)
  • Systems have installed GPS and compass for position can orientation that can be viewed on website also.
  • (Technically, both Vindicators are facing directly south, but these adjustments can be made in post processing, and may show systems aligned property due to correction in the software according to Yanis Gryshan.)
  • 1-second data is available on compact flash card inside electrical box. Access to the CF card should be via coordination with AXYS and Julie Lundquist.
  • Data heights are 49, 60, 80, 100, 120, and 149.7m (where the start and ending heights are near the BAO tower sonic heights, but are dependent on the Vindicator system setup).
  • On-site system login is very similar to a WindCube LiDAR where the system IP Address must be specified and a Laptop can be used for a remote connection with an ethernet (CAT 5) cable. Would need to obtain IP addresses for each Vindicator LiDAR from Yanis or AXYS.
  • LiDAR pair is to be deployed ~2 months.
  • Washer fluid and compressed air tank should last ~6 weeks without changing needs. These components wash off the optical apertures if they are obstructed or dirty.
AXYS Vindicator Next Steps
  • Arrival of UMBC motion table on truck will be coordinated with BAO staff Dan Wolfe, safety officer Otto, Julie Lundquist, and Andy Clifton. UMBC should be bringing compressed air tanks so that Julie can return compressed air rental tanks to Broomfield ASAP. Note that two compressed air tanks should last the two lidars three months.
  • AXYS LiDAR closer to the road ("15") will be deployed to the UMBC motion table for testing.
  • To deploy, the LiDAR does not need to be powered down, but could be via the red turn switch on the west side.
  • There is ~4.5m cable that can be extended out to keep the RS power cables connected between the systems.
  • Care must be taken in taking the Vindicator onto the motion table as cords / tubes must be threaded out from the central stand to not be obstructed keeping the systems connected.
AXYS Vindicator Storage Boxes
  • The AXYS Vindicator storage boxes are put back together and stored under the BAO middle shipping container next to the tower.
  • The boxes are covered with tarps to keep dry from weather.
  • The top and front panels are held in place only with two screws.
  • Robertson drives (square heads) are required to seal / unseal these boxes.
  • All remaining Robertson head screws are in sealed ziploc bags inside the AXYS boxes.
  • Note: The Robertson (square head) bit still needs to be returned to the BAO since it goes with the AXYS tool set. (Might want to hold onto it until packing up the Vindicator LiDARs for opening the boxes.)
General Field Work Notes:
  • Halo scanning LiDAR at the BAO LiDAR supersite was observed to be moving and scanning upon further examination. Julie Lundquist observed very little change in aperture orientation. Later, Paul Quelet confirmed the change of scanning head position. Hopefully, this system is working well.
  • Created on breeze in /data/fielddata/XPIA/ the following folders: MGAUS_photos and BAO_LiDAR_supersite_photos, and AXYS_Vindicator_deploy_photos. These are the start of some folders where photos can be transferred, shared, and used.
CU / NCAR WindCube LiDAR Notes:
  • Morning of 30 March 2015, Julie Lundquist manually wiped off the apertures of the WC LiDAR systems. Should see momentary loss of data on the date 30 March in MDT morning.
  • As of 30 March, WC68 is about 58% full on the local instrument D: drive data storage. This is cautioned with yellow in the WindSoft software. (WC61 is only ~40ish % full, so going strong).
  • WC61 continues its great performance. Did not make any changes to the system. Cover is on and Verizon cell dongle hangs vertically from entrance door, but still under sun shield cover. Usually has about 3 bars of Verizon 3G cell service.
  • WC68 has been inconsistent in cell data connection for remote login or data transfer (TeamViewer or FTP). Usually only has 1 bar of Verizon 3G cell service.
  • Testing of WC68 Verizon dongle on 30 March revealed stretching out the dongle to top of WC68 produced greater cell signal strength. When put away without sun shield, dongle was stretched more horizontally, so (update) 1-2 bars of Verizon 3G cell service is now occurring.
  • WC68 continues to not transfer the full files in FTP. (Update: 31 March 2015.) WC61 did not transfer full files via FTP either.
  • Could consider installing WC68 sun shield and hanging Verizon dongle vertically for better remote connectivity. 
  • On site, Paul Quelet was able to remote desktop with ethernet cable to WC68. Transferred all current Alarm, .rtd, and .sta files. Copied them over to breeze.
  • Changed WC_FTP.txt (script) on WC61 and WC68 to dates of 2015_03_3* so that only March 30 & 31 files are transferred.
  • Must change WC61 and WC68 WC_FTP.txt scripts to 2015_04* at the beginning of April to transfer all April files.
  • WC16 (new v2 system) needs more careful steps to remote desktop into to obtain 1-second data files. Also, requires FileZilla according to Welly Cobden from NRG to complete this data transfer. The manual must be read carefully here and executed on site.
  • WC49 is unplugged completely. Still would need an external keyboard, monitor, and mouse ("old school" style) to troubleshoot problems with on board computer. (Not quite a priority at this point.)

Thursday, March 26, 2015

XPIA WindCubes Internet Connections, Data Transfer, Availability, Data Viewer Update

1) Connecting via TeamViewer
WC61:
--Has been nearly immediate when trying to connect.
--Good speed rendering and data transfer if needed most of the time.

WC68:
--Goes online and offline in terms of being able to connect.
--Slow speed rendering once a connection is established. Fast connection maybe only one time per day.

2) Data Transfer:
WC61:
* Has transferred up to the most recent files
* Did not transfer the full size of files timestamped 19 March and 21 March via FTP.
Sometimes the scheduled data transfer is interrupted multiple days in a row. Usually by 4 days past, the full files have been transferred. Current FTP script moves entire current month (e.g. 2015_03*) files, but could be more specific (e.g. 2015_03_2*) to reduce number of files transferred, especially with intermittent internet connection.
* Must modify at beginning of April to 2015_04* files.

WC68:
* Is three days behind FTP file transfer when examined on 26 March 2015.
* Two most recent days' relevant data (.sta files) are in the D:\DATA EOLID\ARCHIVE\ folder instead of the D:\DATA EOLID\DATA\ folder where the FTP transfer usually occurs from. This does not have a good work around yet, and should be noted for obtaining data files.
* Must modify at beginning of April to 2015_04* files.

3) Data Availability
  • Both Wind Cubes have had trouble reaching remotely sensing altitudes between 140-220m for 22-24 March 2015.
  • Difficult to say how this is affected by scatterers. PTQ does not see direct correlation with the HRDL data in terms of Aerosol Backscatter Signal Strength (http://www.esrl.noaa.gov/csd/groups/csd3/measurements/xpia/hrdl/).
4) DataViewer Update:
WC61:
--Since the 19 and 21 March data files were only 22 minutes and 30 minutes of data in the files, went to WC61 via TeamViewer, and was able to transfer the full sized data files directly.
--Included all data in DataViewer figures.
WC68:
--Since 22 & 23 March were not transferred yet to breeze, transferred the files from the \ARCHIVE\ folder
--Included all data in DataViewer figures.


5) Few Notes:

A) Will not change FTP scripts in the middle of an experiment. This is best done during a development phase.
B) There are other Windows FTP free softwares existing, but not worth exploring until after XPIA (WC operating systems are supposed to go to UNIX-based I believe.)

Sunday, March 15, 2015

Corrections to Windcube updates

Thanks, Paul, for the summary in your post of 3/14/15. Some corrections are necessary.... the lidar supersite at the middle south guy wire consists of :
- 2 CU v1 Windcubes (WC49, WC 68)
- 1 NCAR v1 Windcube (WC61)
- 1 HALO photonics scanning lidar (not an AXYS system)
- 1 Leosphere v2 Windcube offshore 8.66 system (with motion compensation). Logging into that system is beautifully simple. Use the "laptop" ethernet port and then use a browser to go to "http://windcube/windweb". Please do not change any settings when accessing this system (currently it is set to collect data at 40, 50, 60, 80, 100, 120, 140, 150, 160, 180, 200, and 220m - nice to have 12 levels!). Here's an image of the windweb from the 8.66 manual:




The two AXYS systems (one for the ground, one for the motion table) are currently in boxes just east of the trailer in which the windcube storage boxes are stored near the base of the main tower. Current plan is for Yanis from AXYS to arrive on 30 March to install.

Julie's version of the status:

WC49:
  • Operating but not likely collecting usable data. Most recent access (25 Feb) showed all NaN. Not ftp-ing data anymore since modem disconnected.
  • Cannot log into this system with ethernet cable in person, perhaps because of a bad port.
  • Previous alarm files noted laser errors ("Alarm: EDFA").
  • To Do: Welly Cobden suggests take keyboard, mouse, and external monitor into field with side panel open to troubleshoot computer ports at some point. Julie attempted this on 13 March but brought USB keyboard, mouse, monitor cables. Need to bring old-school keyboard, mouse, and monitor cables. Not a priority since WC61 and WC68 functioning reasonably well.

WC61:

  • Has been continuously taking data.
  • Have intermittent difficulties logging in via TeamViewer (JKL: these challenges continue depending on modem connection, but seldom persist longer than an hour.)
  • Welly suggested that noise was related to fan and overheating. Will not address until system goes to NRG in April
WC68

  • More difficult to log into with Team Viewer probably due to modem issues. Data are being ftped to breeze (/data/ftp/incoming/instrument_data) every few days (scheduled daily but ftps fail if modem not connecting at just the right time). When logging in with Team Viewer, JKL sometimes gets a blank white screen but file transfer can still occur (hence my diagnosis of modem bandwidth issue)
  • WC68 was "stopping" itself due to Drive issues (see /data/ftp/incoming/instrument_data/WC68_ALM files) that seem to have stopped since Welly's adjustment of drive accuracy tolerance (from 1 to 3) on 12 Mar. These drive errors are due to the motor aging (which is interesting as this motor has been replaced a few times).
The HALO scanning schedule involves periodic stares at the 50m tower sonic. Please try not to obstruct its view when visiting the WCs. Its schedule is:
  1. 8-beam VAD once every 12 minutes (each scan takes ~40 sec), 
  2. 50 m sonic stare for ~10 minutes once per hour,
  3. Target sector scan (Erie HS Stadium light posts) once per day, 
  4. Vertical staring the rest of the time. 

Saturday, March 14, 2015

XPIA A2e BAO Windcubes Status Update

LiDAR supersite below inner south guy wire at the BAO is quite crowded with 3 CU Leosphere/NRG v1 Windcube LiDARs, AXYS scanning LiDAR, and a new Leosphere/NRG v2 profiling LiDAR installed on Thursday 12 March by Welly Cobden and Julie Lundquist.

Summary of past few days of WC v1 LiDARs:

WC49:

  • Appears to be taking data right now...? (Confirm Julie?)
  • Cannot log into this system with ethernet cable in person, perhaps because of a bad port.
  • Previous alarm files noted laser errors.
  • To Do: Welly Cobden suggests take keyboard, mouse, and external monitor into field with side panel open to troubleshoot computer ports at some point
WC61:
  • Has been continuously taking data this week.
  • Have had trouble logging into it sometimes with TeamViewer, but problem seems to have worked itself out.
  • Intermittent sound coming from side had to do with temperature controlling the system. Would probably stop of the heat dessicant was removed, but will not as continues to run fine taking data.
WC68
  • Logged on today (14 March 2015) and discovered the need for Verizon software update.
  • Restarted TeamViewer, should be online just fine.
  • Previous issue with the stage rotating had to do with a setting of how far off the stage can get over time as the optical head moves. Was set to a very low value. Welly Cobden changed it to a slightly higher value. Should hopefully remove optical head stoppage and scanner errors.
  • As a result, some missing data from earlier this week (9 & 10 March 2015).
(Quelet)

Wednesday, March 11, 2015

Update: WC68 is working now. Just needed to login to TeamViewer again with the all Teamviewer Accounts login. This information is contained at the end of the updated WindCube procedures document which is on breeze in the /data/fielddata/XPIA/. (Quelet)

WindCube v1s Status Report

During XPIA, the WindCube v1 systems have had several times of down data. Not sure what is going on exactly, but here is the status report.

Visited the WindCube systems at the BAO 0915 MDT. The summary of each one is below:

WC49 (CU oldest, southernmost)
  • No electrical power, not operating.
  • Last post from Julie to this blog that I saw here said stage was rotating, but were not able to access it. Earlier in the experiment, the data were all NaNs.
  • The latest alarm files have been due to hard drive errors. Welly Cobden at NRG has been contacted, and should be spoken to on 2015_03_12.
  • Alarm files can be viewed on breeze. (Need password for cuwindcube@gmail.com).


WC61 (NCAR northernmost)
  • Making a low volume buzzing noise coming from the south side as the head rotates.
  • The sound went away after a while.
  • Rotating, running, and able to access data. (Good girl... :) )


WC68 (CU new, middle)
  • Scanner head was frozen in one position upon arrival. Rochelle Worsnop was able to log into Team Viewer and see that the scanner light was red.
  • Restarted the wind soft software and saw all the data collection came back.
  • Verified with team viewer with Rochelle and Paul Quelet iPhone.
  • Is this something to ask NRG about? Can a script be written to override this?
  • Update: 2015-03-11 at 1038 MDT local was not able to connect from PTQ laptop or iPhone. (Message says that "Team Viewer is not running on remote computer."


(Quelet)

Friday, March 6, 2015

NCAR Radiosonde Launching Procedures

NCAR Radiosonde Launching Procedures
Instructor: Tim Lim
Authors: Paul Quelet & Joseph C. Y. Lee
Last Updated: 6-Mar-15

*** Start 20 to 30 minutes early to ensure you can launch radiosondes as close to the right local time (8 am, 12 pm, 4 pm, 8 pm SHARP) as possible. The radiosonde time is shown at the bottom right corner in a grey box in UTC time. ***

Definitions & Facts
MGAUS = Mobile GPS Advanced Upper Air Sounding System

NCAR uses the Vaisala radiosonde system. It records at 1Hz temporal resolution.

Variables Measured through Soundings
1)     Temperature (T)
2)     Air Pressure (p)
3)     Relative Humidity (RH)
4)     Wind Speed (WS)—through the GPS
5)     Wind Direction (WDir)—through the GPS

Radiosondes are still often used for ground truth! Remote sensing can still be erroneous.

Radiosondes cost around $200.00 per device.

The radiosondes rise approximately 5.0 m/s as a standard. (This depends upon the balloon inflation size for buoyancy). See the instructions on the volume of balloon inflation.

We are launching 100 g balloons at the BAO, which can rise up to 20 to 25 km above surface.

Radiosondes can land 70-100 mi. from the launch point! The radiosondes launched in the XPIA experiment are not coming back. They do not have self-addressed stamped envelopes, but only the Vaisala website saying that it is save.

Do not worry about the swinging motion direction of the radiosonde as it hangs from the balloon this motion in the GPS is taken away in post processing because the length of the cord is known and the calculation can be done.

Devices & Hardware
·        The most expensive and sensitive part of the radiosonde is the temperature sensor along with humicap capacitor. These are the fine wires in the middle top.
·        The humicap is hydrophilic cap that changes the capacitance. In fact, since there are two humicaps, the way Vaisala designed the instrument, the measurement and reporting of data changes every 60 seconds. This is reflected in the data display. One is always being heated and dried out while the other one takes data, especially important in supercooled liquid water environments in condensation.
·        The radiosonde needs to be sealed because there can be condensation onto the radiosondes.
·        The helix metal portion of the radiosonde is the GPS antenna.
·        There are no parachutes on these radiosondes. When the balloon breaks, the device falls around 35 m/s in the thin upper atmosphere. The device falls ~ 9-12 m/s in the dense lower atmosphere. (This is comparable to a heavy snowball.) So far, no problems reported with human head strikes.
·        Labview is the software that NCAR uses on the MGAUS truck, derived from their dropsonde software.
·        Unlike in the original demonstration, there should be a table out at the BAO to set things up on.

** Radiosonde Launching Procedure **

Setting up radiosonde:

1)     Be sure to launch a radiosonde from flat ground level away from the truck, any buildings or other persons.
2)     You should see a LabView software running on the screen, click “Soundings / Start” (should be first button on the left)
3)     After starting the LabView software, enter the “Mission ID” following format here (in UTC): ‘MGAUS.VC.00.YYYYMMDD.HHMMSS’ Example: MGAUS.VC.00.20150309.120003
4)     Click Save & Continue; Now a file named as “MGAUS.VC.00.YYYYMMDD.HHMMSS” would be created
5)     Click “Start Channel 1”: The radio channel that connects computer, GPS and the radiosonde
6)     Get the radiosonde out of the back of the truck and take it out of the silver-color package, being sure to not drop it!
7)     At this point, in preparation of the radiosonde, follow the instructions printed on the LabView screen…
8)     Put the radiosonde onto the port stand. Insert the umbilical cord port to the bottom of the radiosonde correctly so that “UP” is facing up.
9)     Recondition the humidity sensor by putting the radiosonde on the GC25 ground receptacle. Use the arrows to toggle to the recondition the humidity sensor setup. Press Select (Finnish for “Press Enter”) to recondition. Give the radiosonde about 3 minutes to setup a timer countdown.
10) Take the battery out of the smaller inner packaging. Note the water jackets to keep the batteries at nearly constant temperature.
11) Slide the plastic pins receptacle (with 2 small holes) out of the battery pack *top* so they can be connected first. The battery back cannot usually be snapped on in one single motion.
12) Set the tune frequency on the GC25 ground device if needed. Press select if yes, you do want to change the transmission frequency.
13) For radio transmission 403.00 MHz is the default frequency. This frequency is usually fine as long as there are no other instruments transmitting at that frequency in the area. The last frequency digit MUST BE AN EVEN NUMBER if a change is made.
14) (If needed, change the frequency either by using the mouse or the clicking to toggle the frequency to AN EVEN number frequency.)
15) Select “Not Display Timer”.
16) Select “Disconnect”.
17) Now the radiosonde is reconditioned. Carefully disconnect the “UP” cord, take it off the GC25 ground receptacle.
18) Push the pins on top of the battery pack into the top of the radiosonde until they are tight.
19) Push the whole battery pack onto the back of the radiosonde, starting from the top to the bottom, so that it snaps tight. You should hear a click. Try to pull and separate the battery pack and the radiosonde to check if they stick together tightly.
20) On the computer, click “Continue” and now you should see a spectrum. Each time lock onto the peak transmission frequency for communication with the radiosonde. If you chose 403.00 MHz (default) as frequency, you should see a huge signal peak (strongest signal) at 403.00 MHz, matching well with the radiosonde. Drag the mouse to the signal peak and click to rescale y-axis if necessary.
21) Click Save & Continue to advance to the next screen.
22) Once you have finished the prelaunch procedures, at this point the radiosonde will start trying to take data. Up to that point, all the radiosonde values will be overwritten and not recorded.
23) By clicking Continue, all the data would be written to new file.
24) After the radiosonde starts up after transmission and communication, the calibration coefficients take ~30 seconds to become calibrated. At the BAO, Tim would provide a table with a fan blowing at the radiosonde to minimize moisture impacts for calibration.
25)  Select “Transfer Value”: By doing so, you are setting the current radiosonde measurements as surface measurement on the ground.
26) Click Continue. There is a 10-second time lag for data recorded from radiosonde to be shown on the computer screen.
27) Make sure the flashing lights for the temperature and number of satellites settles down. In order to do this, you will need to hold the radiosonde away from other objects and away from the middle of your body.
28)  DO NOT START PREPARING BALLOON UNTIL ALL THE FLASHING NUMBERS ARE GONE. Make sure the radiosonde is working probably BEFORE you inflate the balloon (you need at least one hand to hold the balloon).
29)  IMPORTANT: Slightly bend the temperature and moisture senor (the metal that sticks out) and click it to the side (between two little plastic hooks) on top of the radiosonde, so the sensors would be tilted. This is CRUCIAL because we want to take measurements off the aerodynamic path of the balloon when the radiosonde swings, as the balloon itself can reflect, absorb and release heat and affect the measurements.

Inflating balloon and launching:

30) Now it is balloon time. In the back of the truck, put the nozzle up the neck of the balloon about 6 inches or so. Make sure to hold onto the balloon with your hand so it does not fly away during inflating the balloon.
31) Tim recommends starting to tighten a zip tie around the balloon neck and nozzle as you are filling it to make the final tightening easier.
32) To start the flow of helium in the balloon, open the large valve (on the top of the cylindrical gas tank) with a hand turn. Each helium tank should be able to fill up 6 balloons.
33) Do not unravel the radiosonde to hang down much before launching. Keep the balloon at some distance away from the truck.
34)  There would be a count near the set up. When the count increases by 20, then stop. For example, when the count start at 339, stop at 359.
35) Loop the balloon bottom around the radiosonde hook and pinch this off with another zip tie so it is securely fastened. This is a bit tricky as a single person and takes some dexterity and practice. Do not let the radiosonde go yet!
36) Make sure the radiosonde tether string is not caught on anything before launch.
37) Once again, check the data the radiosonde is taking to make sure it closely matches that of the WXT-520 and that the flashing lights have settled down.
38) If the above procedures are correctly completed, the radiosonde is already taking data…
39) LET HER GO!! J (like the song)
40) When you go to the sounding software, you can close the front screen that has the temperature and satellites.
41) The data can be monitored in real-time with the LabView interface that has the tabs. It is fairly self-explanatory if you play around with it.
42) Monitor the data appropriately for 40-60 minutes or so…
43) Cutoff the radiosonde data collection sometime past 200 hPa pressure.
44) After a sounding has finished, you hope to click “Good Sounding” but there are other options if something went wrong.
45) Name the sounding following this format: ‘MGAUS.VC.00.YYYYMMDD.HHMMSS.txt’ (In UTC.) Example: MGAUS.VC.00.20150309.120003.txt
46) Write down your name on the name line. Note: These soundings are archived in a large national database for many years.
47) Write any interesting significant weather type things or abnormal things in the notes section. Add a note saying “intended for boundary layer purpose” as well.
48) Click “Quit” in the lower right hand corner of the software.
49) Throw any trash away in a bag in the front passenger seat.

Data Storage & Communication
·        File browse to on board Computer C:\GAUSdata folder
·        Find the name of your folder as named in the “Mission ID” step.
·        Locate the D############_P.1 file, which is your sounding data that continues to be updated until the sounding is killed. The 999 values in the file are essentially NaN’s.
·        This is a columnar text file that shows the measurements from the radiosonde.
·        Keep multiple copies of the data!
·        Copy the data to the large (orange) external hard drive (should be at the BAO).
·        Import the sounding data into the Aspen software, on the WindCube laptop (or your own laptop). (See https://www.eol.ucar.edu/software/aspen for download.)
·        Write a summary of the sounding operations noting any interesting features (e.g. low/high inversion, Low Level Jet, etc.).
·        Prepare the .png sounding from the Aspen software, and the .png of the windspeed and wind direction.
·        Post the summary and images to http://a2expia.blogspot.com/ . (Note: May need to have a personal flash drive or similar available as internet access at the BAO will be limited, and the WindCube laptop should stay at the BAO.)
·        Additionally, copy the raw sounding data file to breeze at /data/fielddata/XPIA/MGAUS.

Important Notes for XPIA
·        The new radiosondes are in a box in the back of the MGAUS NCAR truck at the BAO. They should stay in the truck so they do not take very long to be acclimatized.
·        Erie airport is an uncontrolled airport so that there is no need to communicate when a sounding is being launched. Since there is no tower control, only make sure an airplane is not directly above you when you are about to launch!
·        Note that the communication radio frequency is between 400-406 MHz. This is the part of the EM spectrum set aside by the NTIA for meteorological transmissions.
·        The 400-406 MHz frequency band should only ever be crowded if there are many meteorological devices in the area.
·        There will be a Vaisala WXT-520 short ~2m tower setup on a tripod at the BAO. This is an all in one weather sensor. This will give the “ground truth” to compare against the radiosonde when it is first turned on near the ground at the BAO.
·        Even though the WXT-520 has an audio distrometer, it is not being recorded. There is precipitation recorded at the LiDAR supersite for XPIA.
·        The XPIA project is using 10 gram balloons these are the smallest balloons NCAR uses for radiosondes, so that they have a very slow ascent rate for good vertical resolution in the Atmospheric Boundary Layer.
·        The radiosonde will hang a significant distance below the balloon so that the balloon solar reflectance does not affect the true ambient conditions of the radiosonde.
·        Make sure to launch the balloon as close to the top of the target hour, or a few seconds afterward, as possible. Use the UTC clock on the computer display to verify.
·        Make sure the radiosonde goes up to at least 200 hPa. It is somewhat uncertain where it will pop and start decent, but you have to kill the data collection at some point.
·        Call Tim Lim if there is a significant problem you cannot troubleshoot. 303-818-9906.


Wednesday, March 4, 2015

XPIA checkup

Nice to visit the WCs after a light snowfall. Tracks in the snow showed that a rabbit had also been visiting, but thankfully our visitor did not munch on power cables.

WC-68 has been turning itself off periodically (or else someone has been turning it off when checking via TeamViewer), but this afternoon was operating just fine and collecting data with visibility all the way to 220m. I cleaned off .dsp files so disk space is only 40% full now. Will continue to monitor to ensure it is collecting data now that the other XPIA instruments are collecting data for comparison to these observations.

Concerned that we still can't access WC-49, I turned on WC-61, which had been making an unappealing fan noise. Said noise has disappeared and the lidar operated well for the 2+ hours I sat there in the cold, and is continuing to collect reasonable data. I also wiped fall 2014 .dsp files (space now at 30% full), transferred .sta and .rtd files to the external drive, and made sure the time was set correctly to UTC. Pitch is slightly off (0.8 deg) but I did not think I could adjust well on my own in the cold.

I attempted to access WC-49 repeatedly without success. The stage was rotating. I scavenged WC-49's modem to loan to Valerio Iungo for use with their 200S since WC-49 had only come on line once or twice in the last week.

To do:
- transfer files from external drive to breeze
- set up script for automated plots to display to XPIA collaborators
- ensure WC-68 and WC-61 continue to operate
- get into WC-49, collect alarm files to give to NRG for advice on repairs.

Sunday, February 22, 2015

Moves and updates for XPIA

All three Windcubes have been moved to the BAO on 2/9 in preparation for the XPIA experiment. GPS coords for each system are listed below. According to Google Earth, they should all be at 40.004881N, 105.0039W, 1582msl although the individual measurements show more variability.
 
Paul and Clara with Windcubes 9 Feb 2015

Status as of visit 2/20:
WC68: (40.04885N, 105.00385W, 1581 msl; 40.04887N, 105.00385W, 1579msl on 2/9) operating fine, modem updated, connecting regularly via TeamViewer. To do:
- update ftp script
- connect water reservoir to wiper

WC61: (40.04886N, 105.00389W, 1583msl on 2/9) could operate fine but there is a very loud fan noise, so the system has been turned off. To do:
- Data drive 100% full, so files need to be removed (first remove .dsp files and then transfer .rtd and .sta files to external drive, then remove from WC61's hard drive) and stored on breeze. Verify date/time correct
- Update modem software 
- Update ftp script
- Consult with NRG

WC49: (40.04884N, 105.00385W, 1582 msl; 40.04885N, 105.00390W, 1583 msl on 2/9) stage head rotating, date/time OK, modem software updated, but data returning all NaNs despite a few reboots on 2/20. To do:
- Investigate performance again after several reboots, then check with NRG. 
- Update ftp script.

View to southeast on 9 Feb 2015