Editor's Note:
Due to the COVID-19 pandemic, Unidata's 2020 summer interns did not travel to
Boulder to work on their projects in person. Instead, they interacted with Unidata
developers through Slack, Zoom, and other electronic means.
During the duration of this summer’s internship program I hit the ground running by learning how to code in Python. Before this internship, I had only opened Python a few times while attending classes and did a little coding in this language for a collaborative project at my university. I familiarized myself over the first month by using the workshops available in JupyterLab provided by Unidata. I learned everything from loops to using the THREDDS data sever to plot variables. Outside of the language, I got to learn how an online community works with GitHub to share and process software and data. I also learned about development environments, which I had no clue about before. I will definitely be using all of these tools moving forward.
I was assigned the project of creating a Wet Bulb Globe Temperature (WBGT) calculation to add to Unidata’s MetPy package. WBGT is a measure of the heat stress in direct sunlight, which takes into account, temperature, humidity, wind speed, sun angle, and cloud cover (solar radiation). WBGT is often a better indicator of heat stress than Heat Index because it considers additional variables that affect the way people experience heat, and the National Weather Service is looking at including a it in their gridded forecast products.
(Click to enlarge)
At the outset, the project seemed very self-contained as there is documentation that provides good examples of how to do the calculations. Given a closer look, the project became a bit more advanced than I expected. WBGT is calculated using wet bulb temperature, ambient temperature, and globe temperature. Globe temperature was the troublesome variable. It is measured using a hollow metal sphere (usually copper), painted black and fitted with an internal thermometer. Not only is this temperature measurement not collected by any data services, values for the same variable (Tg) vary in how they are calculated, some needing to be iterated. It is also very troubling testing calculations when you do not have a reliable way to check the results. Luckily, I have awesome people to work with like Ryan May, Drew Camron, and Sean Arms who were available every step of the way (despite working in different time zones).
Going through this internship remotely was less than ideal, but the amazing people at Unidata made it all the worthwhile. I felt completely welcomed into the organization. I was invited to staff meetings, tech meetings, workshops, and even happy hours. Since I entered the Unidata environment as a beginner, I had many questions about everything, but not once did I feel as if I didn’t have somebody who was happy to answer those questions. Although I was unable to fully complete the WBGT calculation for MetPy, I am hopeful that what I have done is a contribution that can be picked up by somebody in the community. Coincidently, I will be continuing work on this subject over the course of the next year. I have been given the opportunity to research heatwaves in Florida with a professor and I will be examining WBGT, Heat Index, and Environmental Stress Index data for my findings. So although the internship has come to an end, my research on this subject is just beginning. This was such an amazing summer that I will forever be grateful for. I am going to be such a step ahead for my final semesters in my undergraduate degree and I couldn’t thank Unidata more for this opportunity.
