How bad is it?

One of the miscellaneous systems in the observatory that I have inherited is the weather station. A critical set of gear that has been neglected far too long. Neglected to the point our telescope operators had been complaining, loudly, about a system that frequently gives erroneous data or provides wildly oscillating readings.

Weather mast covered with several inches of ice
The Keck weather mast covered with several inches of ice
The weather station is critical in protecting the all important optical surfaces of the telescope. The mirrors that gather light from distant galaxies depend on a thin coating of aluminum that is easily damaged. Snow, ice, fog or even simple dew can damage the coating and require the mirror segments to go through a laborious re-coating process. Thus the operators monitor the weather closely, when fog and humidity roll in, alarms go off, and the great shutters are closed to protect the telescope.

The first part I have replaced is the humidity and dew point sensor. In many ways the most important part of the system. The new unit is a modern sensor with a direct ethernet interface, simple to link into the observatory network. This is the same sensor used by the National Weather Service in their remote weather stations. All I had to do was spend a little money, and spend a day hanging off the weather tower on the observatory roof installing it. It was a beautiful, warm, sunny day up there, I got the job done, and got a slight sunburn in the process.

The new unit is in place and has been operating since September. The plan was to run the old and new system side by side and gain confidence in the new system before we use the data to make critical decisions. For the first several months there was simply no bad weather, no foggy or high humidity nights to allow testing of the new unit under the real conditions necessary to evaluate performance.

When the bad weather hit, it hit with a vengeance. With the strongest La Nina event on record in the south pacific we are getting strong winter weather on the mountain. Snow, ice and fogs have been common for much of the last month, with observing regularly canceled. Even when the weather has allowed the domes to be opened, the operators have had to keep an eye top the weather, ready to close again at any moment.

The new humidity sensor has proven to suffer from a major problem, once the shelter for the sensors is totally encased in ice the readings become unreliable. This was not a problem I really expected to worry about, I reasoned that if the conditions were that bad you really did not need a dew point sensor to tell you to close the dome. Not so, the observing staff wants to see good data even when some ice is present. Thus I need to fix the problem, I need a new instrument shelter.

The answer is a new housing for the sensors, one that more thoroughly protects the sensors and uses a fan to move an air sample past the sensor in place of simple air flow through the louvers. Such housing are standard, I just needed to order one… now sitting on the floor in my office. I go up next week to install it, maybe, as long as a blizzard isn’t raging, again. One thing is for certain, it will not be a warm sunny day for climbing the weather tower, and I will probably not need to worry about getting a sunburn.

Author: Andrew

An electrical engineer, amateur astronomer, and diver, living and working on the island of Hawaiʻi.

Leave a Reply

Your email address will not be published. Required fields are marked *