Station back Online

The La Parguera ICON station is back online following the removal of an underwater light sensor by Wess Merton of UPR. This means that all told the station was offline for about eleven and a half days. The removal of the "deep" light sensor appears to have solved the power problems and the station has in fact recharged itself much more quickly than expected. In addition, all other instruments on the station appear to have returned to their prior states of operation. That is to say, other than the two sensors which were known to be malfunctioning (the acoustic wind sensor on the Vaisala Weather Transmitter, and the pressure sensor on the Falmouth "shallow" CTD), all station electronics appear to have survived this adventure unscathed.

As a reminder, the timeline of the station's failure and recovery is as follows:

• Wednesday, April 14th: last station transmission sent at 11:22pm local time.
• Monday, April 19th: remote analysis suggests that the station has lost power due to the failure of its deep light sensor. Parts required for removal of the instrument are shipped to UPR from AOML.
• Friday, April 23rd: the failed light sensor is removed.
• Tuesday, April 27th: first post-recovery transmission is received at 2:22pm local time.

The station's battery levels normally oscillate between about 12.7 V (nighttime lows) and 13.9 V (daytime highs). Right now (see graph at right, which shows hourly battery levels over the past month) the battery levels are still low but they are recovering a little every day and are expected to return to normal levels after a few more weeks.

CREWs Deep BIC removal on April 23rd

On Friday April 23rd I went out and removed the Deep BIC as requested. Below are photos from the morning. The first is the sensor prior to removal.



In this photo the sensors cable had been removed from the spar by cutting the cable ties holding it in place. The deep CTD was wound up in the same orientation as the Deep BIC; this instruments cable was fixed back to its original position prior to the maintenance. Once the Deep BIC's cable was loose, the sensor was removed by unscrewing the two screws at the end of the mount.



Once at the surface, the sensor was gently laid on its side on a towel while I carefully removed the biofouling that had colonized the cable. Some mussels had managed to grow near the interface between the cable and the sensor possibly allowing a mode of entry for water to the sensor. After the biofouling was removed the cable was unplugged from the sensor and a female adapter plug was placed on the end of the sensors cable. This cable was wound up and placed on the spar as depicted in the photo below.

CREWs Station Surveillance

Surveillance was conducted this morning (Wednesday April 21st) at 9 am to verify that all in-water and above-water sensors were present, intact, and not hanging on by a thread for some odd reason. This survey was conducted after the station went unexpectedly offline last Thursday, April 15th. I'm happy to report that everything was present and visually intact, however an internal cause (i.e. failed cable, water logged sensor) may be the culprit of the outage and would not have been apparent during the visual census. Below are photos of the station from the surveillance.

Station Offline, probable power loss

[This is a slightly modified version of an email message sent out Monday morning, April 19th, 2010.]

The executive summary is this: our ICON station in Puerto Rico has gone offline as of last Thursday, April 15th at 10pm local. Judging from the data reported by the station immediately before the outage, I believe that we have a flooded instrument and a total loss of power. I do not believe that the station is continuing to store data locally. I do not believe the station will come back online on its own.

The faulty instrument is the "deep" light sensor, or BIC (see the timeline below for details on why I believe this to be the case). I would recommend that our local collaborators try to visit the station as soon as they can to verify that there hasn't been a more serious structural failure of any kind. They should also disconnect the deep BIC as follows:

1. Remove the sensor from its mount and retrieve it with its cable, still connected, to a boat at the surface.
2. Dry off the connector and disconnect the instrument. Cap the cable end with a "dummy plug." [This will require a female-style dummy plug, we may have to ship some to UPR for this purpose if they don't have any.]
3. As a safeguard, I would recommend coiling the instrument's cable and affixing it to the pylon above the surface of the ocean (just in case there was damage to the cable itself). There should be some long cable ties in our large red "NOAA" box.

Note that this procedure *may* allow the station to return to full power on its own, but it could take several weeks for the solar panels to recharge the batteries. In the Jamaica incident, it was about 10 days before transmissions resumed, and not all instruments came back online when the station resumed transmitting.

The following is a timeline of events before the station failure, reconstructed from the station's final data reports. All times are given are local times (Atlantic Standard Time, UTC-4, which is currently the same time as our EDT here at AOML).

• Wed Apr 14, noon: Very large rain event reported, the strongest such event reported since last summer. Winds were somewhat elevated but not unusually so (gusts stayed below 15 knots).
• Wed Apr 14, 7pm: By this time the Shallow CTD has reported a significant drop in salinity (the Deep CTD reports a smaller drop in salinity at about 4pm). This is probably related to the rain event, and may or may not suggest large movements of water (i.e., currents).
• Wed Apr 14, 8pm: Just after sunset, the Deep BIC shows a rapid drop in power before going permanently offline. The rest of the station appears to be operating normally at this time.
• Thu Apr 15, 9am: Sunrise begins to affect the station power levels, which begin to rise in their normal diurnal cycle.
• Thu Apr 15, 10am: The station experiences a sudden voltage drop. Depending on which reading you look at, it seems like a drop of 1 V in magnitude.
• Thu Apr 15, 10pm: Over the next 12 hours the power levels fall off steadily. When they reach about 11 V (at 10pm local), all transmissions cease and have not resumed. It appears as though at 11 V the station can no longer power the transmitter. It is likely that the datalogger continued to log data locally for several hours after this but at about 9 V all operations would cease.

See also the description of a similar event in 2008 at the Jamaica ICON station, and the follow-up message about restoration of power following the removal of the failed instrument.

Note: we are making arrangements to ship some female dummy-plugs (suitable for capping the ends of the underwater cables) to the UPR team.

regards,

Mike J+

CREW station basic cleaning in April

On Tuesday morning (~1015 am), April 13th, 2010 Wess Merten and Duane J. Sanabria conducted a basic cleaning and assessment of the CREWS Parguera. We conducted a surface inspection, underwater inspection/cleaning of lines, and an optical sensor cleaning. Everything was visually intact.

CREWS Station Complete Cleaning in March

A complete cleaning (CTD validation, cleaned all sensors, chains, lines, and spar) at the CREWS Station in La Parguera, Puerto Rico, was done Friday, March 26th, 2010 by Wessley Merten and Duane Sanabria. The validation CTD was placed at 8:45am and was removed at 12:15pm. All surface and subsurface sensors were visibly intact. All rigging, lines, baseplates, sensors, and the spar are very clean. New screens were placed on the deep and shallow CTD's. Upon arrival to the station, a bird was perched atop one of the surface sensors that flew away shortly thereafter (see photos below).

CREWS Station Complete Cleaning in February

A complete cleaning (CTD validation, cleaned all sensors, chains, lines, and spar) at the CREWS Station in La Parguera, Puerto Rico, was done Friday, February 26th, 2010 by Wessley Merten and Duane Sanabria. The validation CTD was placed at 8:45am and was removed at 12:15pm. All surface and subsurface sensors were visibly intact. All rigging, lines, baseplates, sensors, and the spar are very clean.