The week of January 24th to 28th, 2011 brought together a number of people at the Little Cayman Research Centre (LCRC) for an unprecedented level of cooperation and collaboration on the maintenance of the Coral Reef Early Warning System (CREWS) station. The station was installed in July of 2009 as part of a joint effort by the Central Caribbean Marine Institute (CCMI), which is LCRC's parent institution, and the National Oceanic and Atmospheric Administration (NOAA) lab in Miami, the Atlantic Oceanographic and Meteorological Laboratory (AOML). CREWS is a project under the big umbrella of the Integrated Coral Observing Network (ICON).From LCRC we had Robert Hedges, the research centre's new manager, Michael Rosen, a scientific diver who is also affliated with the Shoals Marine Laboratory, and Lowell Forbes, LCRC's longtime dive master and facilities engineer. Representing AOML was Mike Jankulak, CREWS field engineer and programmer and an employee of the University of Miami (UM). Also following the action remotely were researchers from UM's Rosenstiel School of Marine and Atmospheric Science (RSMAS), Dr. Peter Minnett, Miguel Izaguirre and Xiaofang Zhu.
Our goals, which were ambitious, were fourfold: to bring Rob H and Mike J together to go over the operation and maintenance needs of the CREWS station; to conduct the station's annual instrument replacement; to install some new sensors for Xiaofang Zhu's graduate research project; and to address some of the problems that had developed over the last year and a half. These problems included a suboptimally-installed air temperature sensor (in a spot which did not have free air flow), the failure of the research centre's land-based computer (which had been set up last year with a radio link to the offshore station), and the deterioration of the station's satellite transmitter.
The sea conditions were borderline; on Tuesday we were not able to tie up to the station at all, and our plans had to be rearranged on the fly. Wednesday was the calmest day and Thursday conditions had worsened but we were able to adjust operations to compensate (by mooring the boat to one of the support pins and only tying up to the pylon itself when absolutely necessary).
The station's regular maintenance schedule had been on hiatus since September of 2010, due to the departure of the previous LCRC manager and the resulting shortage of divers for this work. Also, no NOAA/AOML divers had visited the station since the pylon repair and straightening operations on April 8th, 2010 and July 6th, 2010. So we began with a survey of the station.
The underwater lines and chains were remarkably clean, considering their long period without maintenance. 
The spectra lines were somewhat "slimy" (see photo at left) but there was no significant growth (of barnacles or the like) on the spectra, chains, lashings or the pylon itself. Only the pins seemed to be in need of more aggressive cleaning (photo at right). It seems like this station is far less prone to bio-fouling compared to St. Croix or Puerto Rico.
The underwater instruments were fouled and in need of some cleaning -- notably, the copper screens were dissolving on the CTD and the sensing surface of the light sensor was clouded over (see photo). However, both instruments were removed on this trip and replaced with clean instruments. As reported in July, the station is noticeably twisted and this is most obvious when looking at the bottom lashings to the base plate. Whereas previously these lashings were found to be overtight (which may have contributed to the damage reported here in January of 2010), on this trip these lashings were extremely loose.Two structural concerns were observed above the surface as well. First of all, the interior of the control unit or "brain" chamber was found to be noticeably damp. There was condensation on the inside surface of the station lid and enough moisture to drip off the brain's fiberglass board when it was first removed. All four conduits into the brain were found to be sealed tightly by aquaseal so it is not clear where this moisture is coming from, but it bears further monitoring. A second problem was that the bolts pinning the mast of the Biospherical "BIC" surface light sensor had both worked themselves loose enough that this mast could swing freely from side to side in its mount. The bolts were still in place, threaded into the fiberglass, and were securely tightened at the end of our operations.
We'll begin with the new LCRC/RSMAS/NOAA collaboration. Xiaofang Zhu is doing a detailed study of sea temperatures near the ocean's surface at two CREWS stations, one in Puerto Rico and this one in Little Cayman. Her team had provided four self-logging Levelogger sensors which automatically record sea temperature and pressure every six minutes. 
These were installed on the pylon (on Tuesday) at depths of approximately 0.2m, 0.5m, 0.85m and 6.1m. These measurements were taken during low tide and the average instrument depths will be somewhat deeper than this. The installation was done by Mike Rosen and Mike Jankulak. On Thursday the location of the lowest sensor was fine-tuned by Mike Rosen and Lowell Forbes. Photographs and video were taken by Mike Rosen.
Next we'll talk about the instrument exchange. The replacement of the two underwater instruments was completed on Wednesday by Mike R and Lowell, and a third "groundtruth" sensor was connected and verified to be operational on Thursday. The aerial instruments and "brain" control package were removed from the station on Wednesday, replaced and reprogrammed on land, and reinstalled on Thursday.
This leads naturally into discussion of the advances made in our LCRC/NOAA collaboration. For the first time ever at a CREWS station, the "brain" installation and instrument connections were not done by NOAA/UM people but were instead conducted entirely by LCRC's new station manager, Rob Hedges (shown in photo at left). The importance of this development cannot be overstated. Previously LCRC was limited in the type of maintenance or intervention they could do, because nobody at LCRC was trained in the climbing of the pylon or even the simple act of powering the station up and down. These actions are more are now within the capabilities of the locally-based LCRC team. Whereas previously LCRC could not schedule major operations without coordinating with visiting AOML researchers (typically a once-per-year event), now they have assumed more complete control of their CREWS station.Lastly we'll examine what was done to address the problems with the station's functions. The air temperature sensor was removed from its enclosed spot, and a new sensor was installed in the open air under a specially-made radiation shield. The radio link had originally been set up in January of 2010 but the station's computer crashed a few months later. Some efforts had been made (in May and July of 2010) to remotely guide LCRC researchers through the steps of reinstalling the software on a new computer but these were ultimately unsuccessful. In this current visit, we were able to reinstall and upgrade the software used by the radio link to make it operational. The research centre's computer is now configured to automatically pull updated data every five minutes from the offshore station.
The problem of the satellite transmissions proved more difficult to solve. Normally a CREWS station might lose up to 5% of its hourly satellite transmissions to bad weather or other signal loss. The Little Cayman station lost 25% of its transmissions in the last three months of 2010. The station inventory did not include a spare satellite transmission antenna so we retrieved the station antenna and cleaned its connector and cable plugs, and then carefully reinstalled its co-axial cable to avoid any sharp bends. We also replaced the transmitter itself with the backup transmitter stored in LCRC inventory on land. Unfortunately, the replacement transmitter (installed on Thursday) does not transmit at all. Due to the lack of a spare satellite antenna it was not possible to test the transmitter on land ahead of time, and by Thursday afternoon the sea conditions had deteriorated to the point where further access to the pylon would have to wait at least three more days (and perhaps longer).
Fortunately the repair of the radio link provides us with another way of accessing the station's data in near-real time. The LCRC computer used for the CREWS project, as mentioned, updates its local copy of the station data every five minutes. This includes much more detail than was available by satellite transmission. By satellite, we only saw the 60-minute summary data, but the radio link allows access to every instrument's individual data readings, whether 6-minute (CTDs, CT, transmitter diagnostics), 1-minute (our integrated Vaisala "WXT" weather transmitter), 30-second (our Biospherical "BIC" light sensors, datalogger diagnostics) or 5-second (the standalone air temperature sensor, barometer, anemometer and electronic compass)!
We quickly put together some scripts to upload the hourly data to a NOAA ftp site, so that they could be processed at AOML as before. Some further scripting will be required on the NOAA side next week before these data feeds resume. These feeds include the one to ICON's G2/Ecoforecasting project and the one to NOAA's National Data Buoy Center (NDBC), which is the means by which the Little Cayman data are assimilated into the global weather models used by hurricane forecasters. Furthermore, a monthly job was created to copy some of the larger (6-minute, 1-minute) data files without unduly burdening the LCRC network connection. All of these data will remain recorded on the station's new 1GB flash memory card (an upgrade from 64MB used until last week) which can continue to log data without loss until late in 2013.
One final note about data archiving: after retrieval of the station's CTD which had been deployed for 18 months, we connected directly to the unit and downloaded the locally-stored data from its flash memory. This dataset has been confirmed to be intact and is available for the asking.
It's been a hugely productive week all around. LCRC's advances in assuming control of their CREWS station maintenance and configuration, the new collaboration between LCRC and RSMAS researchers, and the greater access to station data by means of the repaired radio link are each in themselves significant milestones. Congratulations to everyone involved in this very successful operation!
(signed)
Mike Jankulak
Levelogger photos and video taken by Mike Rosen; all other photos taken by Mike Jankulak.