A cool day with light winds. Scattered low cloud in the morning, partly lifting during the day with scattered upper level clouds remaining.
Today we started off at ISS1 where Isabel got the CL61 ceilometer plots working and I was working on the wind lidar. There is an issue on the lidar with an incorrect naming convention, the Leosphere folks are working on a fix. In the meantime the data is being copied off the lidar but is not being plotted yet. I also set up some new scans with the aim to get the boundary layer depth algorithm working.
We stopped at ISS2 to pick up some parts, check the systems and add a pallet to the bottom of the stairs to make it safer to get in and out.
The main activity of the day was at ISS3 at Sedgwick. After some tidying, testing and prep work we hosted the second sounding class. This time there were six students (some repeats from last time) and it went very well.
We also worked on the wind profiler, installing the now repaired clutter screen panel, covering a RASS cable that some critter had chewed on, and relevleing the antenna. We also adjusted the parameter file so that now only the WA/1 vertical beam is sampled, since the WA/0 vertical beam had excess clutter signal. The profiler is working sufficiently well to collect winds, although the vertical winds and backscatter are somewhat noisier than the wind profiler at Rancho Alegre. The NIMA processing is doing a good job cleaning up the data.
We did find that the site had another power glitch last night that stopped the profiler and ceilometer operation. It appears that the power was out for only a few minutes, so it is a little surprising that it interrupted those measurements since they are both on UPS. We plan to redistribute the loads on the UPS and consider bringing out the spare to reduce these kinds of problems.
Sounding class at Sedgwick
A mostly cool cloudy & calm morning in the valley. The fog/cloud lifted around midday and it was partly sunny although windy during the afternoon.
We were mostly working at ISS3 Sedgwick today. We took the profiler antenna apart, temporarily removing the clutter screen and radome to clean off the antenna, which was covered in a thick layer of dust and dirt. We had wondered if this was affecting the performance of the antenna, however as John will report elsewhere, is wasn't clear if this helped. Taking it apart did allow us to inspect the surface of the antenna and it was noticed that the panels are a little warped and cracked which may explain the problems we are having with it. We mostly reassembled the antenna this afternoon, although John is making repairs to still one of the clutter screen panels so the profiler is missing that panel for now. Liz also replaced one of the relays that had been sticking earlier.
Isabel is making good progress on the data flow and many plots are appearing on the web. The first are now available at the SWEX ISS web page with more to follow soon. Erik's wind plot visualization tool plotting up the 449 MHz profiler and the link to that was also added to the web page.
We replaced the main SSD disk drive on the wind lidar yesterday with a new drive that arrived from Vaisala. This included a software update that should improve the stability of the system. It seems to be working well so far, however the file names are misnamed ("WCS999951_" instead of the usual "WLS200s-181_") which is causing a few problems with the downstream processing for now.
Cleaning off the antenna at Sedgwick and reassembling the clutter screen afterwards.
ISS1 profiler data and LIDAR data are now available in the TimeHeightPlot web visualization tool.
449 profiler: 5-minute and 30-minute consensus winds.
LIDAR: VAD (Velocity Azimuth Display) and 30-minute consensus winds.
http://datavis.eol.ucar.edu/time-height-plot/SWEX
Clear and sunny along the coastal region all day. Clouds/fog over the Santa Ynez valley in the mornings, clearing skies and windy (20+ mph) in the afternoon.
Started the day at ISS #1. Isabell joined us today to work on networking & software issues. Bill and Liz replaced the hard drive in the Lidar, which required Bill to update the client on the lidar laptop, which Bill was able to do after some texting with tech support in Europe (before they went to bed). Lidar is now working and Bill can comment/clarify if the data is being transferred as desired to the data manager. Grabbed some RF adapters/attenuators and two towels from the trailer for debugging and cleaning the ISS #3 wind profiler antenna.
Stopped by ISS #2 after lunch to check on the systems and for Liz and Bill to work on networking/software issues. The following images show the current draw of the final amplifier and the returns being generated by the 'good' 915 MHz radar for comparison with returns with the 'bad' radar at ISS #3.
ISS2 (RFE #2) Amp Current
ISS2 WA0 X-vert ISS2 WA1 Y-vert
ISS2 WA2 X+ East ISS2 WA4 X- West
ISS2 WA3 Y+ North ISS2 WA5 Y- South
We then continued on to ISS #3. When we arrived, the profiler, its computer, and the ceilometer computer were all off. It appears that the site lost power around 4-5 am local time this morning and was off for at least long enough for the CyberPower UPS's battery to completely drain and the UPS turned off. The UPS did not power back on when line power was restored, but it did recharge its batteries. John looked at the manual and datasheet for the UPS and did not find any settings that would change the UPS's behavior. The datasheet also showed that the UPS would power everything plugged into it for ~17-20 minutes at the present load levels. We will look at moving the ceilometer computer and monitor over to the UPS that the sounding computer is plugged into, which either did not deplete its battery completely or was able to turn back on its load outlets when line power was out since the sounding system was on when we arrived. After powering everything back up, John and Liz measured AGL sensor heights while Bill and Isabell did more work on networking and software (and other things?).
On the 'bad' wind profiler front, for comparison purposes, the final amp current of the ISS3 system (RFE #1) was 3.52 Amps, comparable to the power output of the 'good' system at ISS2. Images of the returns being generated were also taken, but alas, they were accidentally deleted. John will take images tomorrow and add them here, though atmospheric conditions may not be identical. Bill has determined that the NIMA processing is able to clean up the 'bad' returns well enough that the antenna will likely not need to be replaced at this time. However, the returns used for this analysis were under windy and moderately humid conditions that produced returns strong enough so as to minimize the effects of the antenna's clutter signal. We are interested to see how the processing performs when the signal returns are weaker, i.e. drying conditions and/or lower winds, some of which we may see tomorrow. Since SWEX is primarily interested high wind conditions, potentially poorer performance under low wind conditions may not be an issue.
It was quit windy (~20 mph) at the site so it was decided to postpone removing the clutter fence and radome until tomorrow when winds are forecasted to be less than ~10 mph. Once apart, the dust/dirt layer on the antenna will be removed and the RFE modified to accept an external local oscillator (LO) frequency from the signal generator that John brought out. Antenna input return loss measurements from yesterday indicate that the antenna's performance, across all beams, can slightly be improved by lowering the operating frequency to ~912.8 MHz from its nominal 915 MHz. We should have more data after tomorrow's work on the antenna's and system performance.
Overcast, rainy, and cooler today with highs only in the lower 50's.
More work at ISS #3 today. Due to the wet conditions and and a sore lower back, it was decided at breakfast to not take the clutter fence and radome off the 915 radar at Sedgwick this morning. Bill emailed this morning that we may be able to use the existing antenna since he compared its performance to the data from the radar at ISS #2 and saw good correlation between the two even though the system at ISS #3 was running at a lower power level (see Bill's comment on the ISS setup 27 March 2022 blog entry).
The final stage amplifier's current, and therefore output power, varies a lot for small changes in 60 MHz input power. The system had been running with a lower power using 12 dB of attenuation (10 dB + 2 dB) and a final stage current of 1.02 Amps. I increased the power by decreasing the attenuation to 10 dB (removed the 2 dB attenuator), which produced a final stage current of 3.8-4.0 Amps. I wanted the current to be in the 3.5-3.6 Amps to match the power level of the system at ISS #2 and to make sure the final stage is not being overly compressed. I restarted the system and adjusted the 60 MHz Tx power level to get the desired current by adding a couple lengths of thinner coaxial cable in the line (no 1 dB attenuators available), producing a final stage current of 3.8 Amps upon initial startup, which is expected to decrease to near 3.6 Amps as the system heats up.
The system was showing very strong returns when started, likely due to the high humidity level/rain and winds, so much so that the clutter seems to be suppressed in the display. Here are images of the display for each beam after the system was started showing the strong return levels.
Vertical Beams WA0 & WA1
North-West & South-East Beams WA2 & WA4
South-West & North-East Beams WA3 & WA5
Liz arrived today and Bill arrives later tonight. Tomorrow's forecast for Santa Ynez calls for sunny skies with a high near 67 and north-northwest winds of 10 to 15 mph with gusts as high as 25 mph. We will see how the system runs tomorrow under the clearer and drier conditions. Hopefully this will give Bill enough data to decide whether or not the antenna at ISS #3 needs to be replaced with the MISS antenna.
Cloudy and cool in the coastal region with light mist at times. Clear and windy inland of the Santa Ynez Mountains and at Sedgwick, with a high in the mid-60's
Spent the day testing the 915 MHz antenna ISS site #3 (which happens to be Antenna #3). When I first started up the portable VNA, the cursors remained at their last locations, and indicated that the last patch that I measured had a better input return loss (IRL) at 909 MHz. The VNA was calibrated from 900-930 MHz to get a wider look at any trends in the IRL measurements. The IRLs for each beam were measured looking into the main antenna port. There were no clear resonances or frequencies across the frequency range where the antenna's IRL was better than 16.7 dB, with most of the beams producing IRL numbers in the 6-14 dB range. A video of the measurements was taken but is too large to be uploaded on the Wiki, so I uploaded to the shared APG_team_drive Google Drive (https://drive.google.com/file/d/1NZghXvnDb9MmcyK15yRBfzkZDmG4tLk2/view?usp=sharing). These IRLs are quit poor when compared to the IRLs of the 'good' antenna (at ISS site #2) which were better than 22 dB for all the off-vertical beams. This is indicative that something is wrong with the antenna.
I then pulled off the access panel on the bottom of the antenna and then measured the IRLs looking into the eight 1:4 way splitters that feed the patch antennas, both in the X and Y polarities. Measurements taken at these points are after the input connector and relay banks and just examine the splitters and the patch antennas themselves. Here are images showing how the splitters for the X polarization are numbered (Y polarization splitters numbered in the same manner) and an image of how the measurements were taken, with the VNA plugged directly into the splitter inputs.
The IRL's looking into the splitters did not look good either, i.e. no sharp/deep resonances where the IRL exceeded 20 dB. Since the IRL of a combiner can be no better than the worse IRL of the patches that are connected to the splitter's outputs, the poor IRL issue is with the antenna patches themselves. The issue also appears across all the splitters, so it appears to be affecting the entire antenna. Here are the IRL images looking into the X1-X4 and Y1-Y4 splitters and a summary of the measurements.
Splitter IRL at 915 MHz Frequency with best IRL & IRL value
X1 15.0 dB 909.9 MHz, 17.8 dB
X2 12.5 dB 911.7 MHz, 13.0 dB
X3 18.9 dB 914.2 MHz, 19.3 dB
X4 14.2 dB 911.9 MHz, 15.1 dB
Y1 13.5 dB 909.6 MHz, 15.9 dB
Y2 16.0 dB 913.0 MHz, 16.4 dB
Y3 18.5 dB 914.2 MHz, 17.8 dB
Y4 15.1 dB 912.4 MHz, 15.7 dB
It is noted that X3 and Y3 feeds are close to having their best IRL at 915 MHz and also have the maximum IRL values. Unfortunately, the four elements that each of these splitters feed are on different subpanels, as is true for the other splitters, so whatever is going on with the antenna elements is not isolated to one of the subpanels.
In summary, the antenna is not resonating well at any frequency between 900-930 MHz. The issue has been tracked to the output splitters and antenna elements, though these findings don't rule out other potential problems with the relays as we did find at least one 'sticky' relay during earlier testing on the 3/13/22. However, fixing the relay(s) will not solve the problem with the elements not radiating well, which must be fixed before the antenna will work with the system.
There is a coating of dirt/dust on the surface of the antenna, under the radome, that could be causing the degraded performance, especially if the dust has an ferrous material in it. However, this is a long shot and I suspect the issue is with the antenna substrate material, i.e. it has degraded in some way(s) over the years, such as material has dried out, or has developed cracking, or has otherwise become lossy. Tomorrow the plan is to remove the clutter fence and radome, sweep off the dirt/dust and then wipe the antenna surface clean to see if that has any effect. I have talked to Steve and it sounds like the surface group can help me in the morning to remove the clutter fence and radome.
More fun (and measurements/data) tomorrow!
A sunny warm day. ISS setup so mostly complete. Although there are some minor tasks yet to be done, the only major outstanding issue is the ISS3 wind profiler at Sedgwick. It is running however still has the clutter signal problem dominating the atmospheric signal. We are working on a plan to repair the system (probably involving tuning the local oscillator to better match the antenna) and will return with the necessary equipment to do this work the last week of March. John, Gary and I did some tidying up at ISS1 this morning and flew back to Boulder this afternoon. I will be returning on March 28 and John and/or Liz may be flying back sooner to work on the profiler. Thanks to the rest of team (John, Liz, Lou, Gary and Dexter) for their hard work during the setup.
Mostly sunny with some clouds in the valley this afternoon, light winds.
We worked at all three sites again today:
ISS1 (County HQ): installed the Lufft WS800 sensor on the three meter tower
ISS2 (Rancho Alegre): conducted a training session with the students and postdocs who will be the sounding operators for ISS2 and ISS3. They were a great class. Also installed the Water Vapor GPS sensor, PurpleAir aerosol sensor, and a USB camera for monitoring sky conditions. Gary did a bunch of network and computer figuration work.
ISS3 (Sedgwick): John continued his heroic effort to diagnose the 915 MHz wind profiler. He found that the second circulator installed yesterday was likely faulty so removed that, and the second limiter might be giving a marginal improvement. We will discuss options further in the next few days, but currently we are considering replacing the local oscillator with a signal generator which will enable the frequency to be tuned to better match the antenna and hopefully give a better return loss figures and thus less of the clutter like signal swamping the atmospheric returns.
First sounding class of SWEX2022
A sunny day although windy at times. We worked at all three ISS sites today, the main issue was the ISS3 profiler
ISS1 (Santa Barbara county offices): Connected and powered up the CL61 ceilometer, although the data ingest on the DM computer isn't up yet.
ISS2 (Rancho Alegre): John hooked up the ISS3 wind profiler front-end to the ISS2 profiler and made some diagnostic measurements. The ISS2 profiler worked well with the ISS3 front-end which suggests a problem with the antenna at ISS3. Also we installed the NR01 solar radiation sensor and Gary did DM and DSM work in preparation for the training sounding tomorrow.
ISS3 (Sedgwick): More diagnosing of the problems with the wind profiler. John did some return loss and frequency response measurements on the antenna and some of its elements. He found that the peak response of the antenna was around the 905 - 909 MHz range instead of the 915 MHz that is should be. It appears to have drifted over time for some reason. There is some dust and dirt on the antenna, but wiping a small section didn't help the unlying elements so perhaps the antenna substrate has degraded. It may be that this is causing the more significant return loss seen on this antenna as compared to the ISS2 antenna (something around 30% as compared to around 5% at ISS2). The stronger signals being reflected in the antenna may be bouncing back into the front-end receiver chain showing up as clutter like signals. He added a second circulator and limiter to try to suppress it, but that doesn't seem to help.
Mostly sunny, scattered cirrus clouds, and breezy. Mostly at the Rancho Alegre ISS2 and Sedgwick ISS3 sites today.
At ISS2 installed the sounding system and CL51 ceilometer, although neither are networked yet.
At ISS3, worked continued on diagnosing the 915 MHz wind profiler apparent clutter signal issue. John and Liz carried out further tests and using a VNA (Vector Network Analyzer) to look at the reflected signal in various areas of the antenna, appeared to have narrowed the problem down to a particular switching relay. However after bypassing that relay, the problem persisted so it is possible there are multiple issues. They did repair a couple of faulty cables and are considering returning some of the relays to Boulder for tests in the lab.
Also at ISS3, Gary met with the Sedgwick network engineer and got a comms link running to the main office which will give us high speed internet. We also installed a webcam (the network camera) and an aerosol sensor (a Purpleair sensor).
Liz (under antenna with VNA) and John working on the wind profiler antenna and relays.
Another sunny day. We split up today, with John and Liz heading to Sedgwick to work on the ISS3 wind profiler, and Gary, Lou & I to work at ISS1.
John and Liz did a lot of testing and switching around of equipment to diagnose the wind profiler issue. This included swapping out all of the electronics with the ISS2 profiler and other tests. The quick summary is that there is a problem with the switching relays. It is not clear how to fix it, so they are considering their next move.
At ISS1, we got the wind lidar connected and running. I ran it for about 3 hours using the old database (at CFACT in Utah, the database on the internal computer became full and would not clear which caused difficulties running the system). I then performed the factory reset that the manufacturer had suggested (and had warned might take 3 days and might not work). Fortunately the operation went much faster than expected and how the system is much more responsive and has plenty of space for SWEX. The air is fairly clear so returns were somewhat limited, but with a 100 m pulse it was seeing 3 - 4 km horizontally at times, although less with the 50 m pulse. We also leveled the lidar trailer, worked on networking, system updates, reran some power, and started installing cameras. The Modular Profiler continues to work well. In the 2.7% duty cycle mode it is giving winds to 3 - 4 km AGL. It's drawing 2.60 A on the 50V and 1.671A on the 31V supply.
Yet another nice sunny day. Today we mostly worked at ISS3 at Sedgwick Reserve. The site is mostly in good shape thanks to Lou, Liz, and John's work there last week. Most of the equipment had remained on site since March 2020 when the first attempt at SWEX was shutdown due to the pandemic, but required quite a bit of work to get it out of the shutdown and storage state.
The major issue at the site is the 915 MHz radar wind profiler. Earlier in the week the computer had refused to boot up and it was eventually found to have a bad RAM card. Now there also is a measurement issue. This profiler is a DBS (Doppler Beam Swinging) profiler whereby the radar phased array antenna is electronically steered in 5 beam directions (one vertical and four oblique). There is a strong clutter like signal that shows up in one polarization of the vertical beam, as well as the northwest and southeast beam directions. It is not clear if the problem is with the relays that do the steering, the beam steering cables, steering codes, interface box, ribbon cable from the computer, or the software. John and Liz are investigating and will report separately.
There also appears to be an issue with the cell communications. There is a cell phone repeater, which does generate a 3-bar signal supposedly 4G/LTE however our cell phones and the cell modem had trouble connecting to it. Gary found that the model gave a better data link without the repeater (good enough for him to update the DM computer) however with the marginal cell service at the site, that is still not a good option. He is meeting with the Sedgwick network engineer on Monday and hopefully we can get a better data link.
In other tasks at ISS3, we positioned the CL31 ceilometer on the front of the trailer but are unfortunately missing a power cable. We will look for that tomorrow. We also worked on the sounding system, installed the NR01 4-component solar radiation sensor, patched together a broken RASS surround, and did other tidying up tasks.
We also transported 10 Helium cylinders to the ISS2 site.
Working the radar wind profiler and solar radiation at the Sedgwick Reserve site
Another sunny day with light breezes.
Today we started at ISS1 examining the air conditioner, after some jolting it around, it started working. It is suspected that the fan may have some debris or bearing issue, but it seems to be working for now and the temperature in the container dropped from around 30C to 18C later inn the day. To be safe we ordered a new unit and hope to get that Monday. The Modular Profiler is working well, seeing winds to almost 3km (still on just the 2.7% duty cycle), and is drawing 2.74A on the 50V supply and 1.68A on the 31V supply later in the day.
At ISS2 the storage pod was delivered (eventually, delayed by address confusion by the driver). We did more unpacking, set up computers, erected the 10 meter surface met tower and the 3-m Lufft sensor tower, although the sensors haven't been installed yet. John, Liz and Lou got the 915 MHz wind profiler working. It is running at about a 5% duty cycle, seeing signal up to about 2km and drawing about 4.4A on the power supply.
Tomorrow we are heading to ISS3 at Sedgwick.
Raising the 10-meter met tower and leveling the radar wind profiler
Another fine day in sunny Santa Barbara, light northerly winds at times.
The lidar was craned up on top of its container by Buck's Moving. It is in position but not powered yet.
Buck's also moved the ISS2 trailer from where it was dropped off at the Santa Barbara County HQ complex, over San Marcos pass along highway 154 to the Rancho Alegre scout camp site near Lake Cachuma. The Forest Service road up to the camp is a rough dirt road so we had Buck move it there rather than the truckers who brought the trailer from Boulder last week.
We started on setup of ISS2, unpacking and getting the profiler antenna into position and the clutter fence installed. We also started on the surface met tripod and on the power hookup.
Back at ISS1 late afternoon, I found that both the Modular Profiler distributed amp enclosure and the ISS1 container were getting rather warm, with the enclosure hitting 36C and the container reaching 33C. I turned down the duty cycle on the profiler (see below) although talking to John at dinner, he said that the amp enclosure should be okay at those temperatures. The container is more of a concern since it appears that the air conditioning in there has failed. I directed a fan at the computer rack to circulate air and hopefully keep the computers cooler. We will investigate further tomorrow.
On the Modular Profiler, I switched the operating parameters from a 5.3% duty cycle down to a 2.7% duty cycle. At 5.3%, the 50V power supply was reading 4.52A and the 32V supply was 1.864A. At 2.7% the supplies were 2.54A and 1.67A respectively. I toggled between both modes a couple of times 0130 - 0140 UTC (March 11). In both modes we seemed to be getting good signal up to about 2.6 km AGL.
Liz and the guys from Buck's craning the lidar into position at ISS1
John, Liz and Gary assembling the clutter fence on the radar wind profiler at the ISS2 site
ISS setup has been underway for a few days and has been making great progress thanks to Lou, Liz, John and Dexter who were the advance team arriving last week. Gary arrived on Monday and I arrived last night. The three trailers arrived last week with the ISS1 and lidar trailers coming directly from the CFACT campaign in Utah, and the ISS2 trailer which coming from Boulder.
The ISS1 site (the Santa Barbara county office complex on Cathedral Oaks road) hosts the ISS1 and lidar trailers (and the ISFS base trailer). ISS1 has mostly been unpacked and set up, although there is still computer configuration and tidying up work to do. The 449 MHz Modular Profiler started running this afternoon and is working well so far, getting winds to the 2 - 3 km level. There is some radio interference although so far it seems to be manageable. The surface met is hooked up but not yet recording and we only have a temporary mount for the Lufft. The CL61 ceilometer is mounted on the front of the ISS1 trailer, but not hooked up. The lidar trailer has been mostly unpacked and the lidar prepared for a crane lift tomorrow to the roof of the trailer. The GPS Water Vapor was also mounted and connected.
The ISS2 trailer is still packed up and currently at Cathedral Oaks but will be moved to the Rancho Alegre site tomorrow.
The ISS3 trailer has been on site at the UCSB Sedgwick Reserve since March 2020 when the pandemic shutdown the original SWEX campaign. Much equipment had been packed away, although the profiler was still assembled (minus the electronics) and the surface met power was still standing (with most sensors removed). Some work has been done restoring the site, and we will be working there again over the weekend.
Gary, Liz and Dexter working on the ISS1 surface met system
John working on the Modular Profiler.