Average Seeing on Magellan Telescope in the past 4 months

It has been awhile since I wrote my last blog post. I am not sure why. Maybe I felt like 'blogging' was no longer exciting and novel, or maybe I just spent too much time consuming media and did not create anything. Then, someone commented on one of my old data visualization works that I did several years ago about Bangkok. Thank you to that person since it made me want to write a new one again.

Enough about excuses. Today, we will discuss about something that is near and dear to my heart, which is observational astronomy. During my PhD study, I got a chance to visit Magellan Telescope in northern part of Chile (I mentioned a little bit at the end of my last post). However, coming to the telescope is not enough for this kind of works. Weather also plays a huge role in this kind of works. And as we all know, we cannot really control weather. More often than not, astronomers travel to the top of the mountains, sit quietly inside the dome, and wait for the cloud to disappear. But, how often? 

To answer this question, we have to pick a specific place and time to get a finite answer. Naturally, I picked Magellan Telescope and within the last 4 months as a starting point. I put the method and all the resources that I used in the note below.

Seeing at Magellan Telescope from 3/17-6/17

This is a plot showing an average seeing on a given night. What is 'seeing', you may ask? Seeing is a measurable quantity to represent how clear the sky is at that particular moment. Technically, it tells you how far two stars/objects have to be so that we can distinguish them as two stars/objects. And, an average seeing tells you how good was the weather that night for observing. 1" is a standard to quantify whether it is a good night (At other telescopes, 1" might be the best that they can get. It all depends on the site's location). And a day without any box represents a day that the weather is so bad that we didn't even open the dome.

As we can see from the data, a terrible night (no data was taken that day) is ~22%, and a great night (an average seeing is less than 1.) is ~70%. This doesn't sound that bad, but we have to recognize that one group might be able to only get 2-3 nights at a time, and there is a non-zero probability that they will not get any data at all.

This data can tell you much more than 'how often we get a bad weather'. For example, we can see that almost all nights in March 2017 were amazing with every single night except one is a great night. On the other hand, in May 2017 ~40% (12 nights) were a bad night.

However, this is not the end of the story. A seeing is simply a result from a combination of various weather effects. In fact, cloud is not the most important factor in this story, instead it is wind. In the next post, I will discuss further about wind and how it can have a effect on the observation.

Magellan Telescope

Note: The most important thing for answering this question is data. I had to scape the data from the telescope's weather website. Fortunately, there is a weather calendar website that archives the previous weather. What I had to do is using Selerium on Python to collect the data by running individual day data from the website and converting it into meaningful numbers. After that, I just made a box-plot to show the data.