Global warming; weather extremities seen through more than just temperature
Each year, our planet falls into a deeper and further unsavable state environmentally. As we continue to consume our planet’s resources, we also grow in population and evolve technologically, requiring more resource demand from our planet year by year. Our technological demands also heavily tax our planet, including overall emissions from cars, energy production, and more, as well as the waste produced as a result of new technological discoveries. As this growth continues to occur, it is very important that we analyze the implications our consumption has on our planet, largely through our weather. Our planet’s weather is largely responsible for maintaining ecosystems, keeping us as humans healthy, but most importantly preserving the future of us and our planet. Global warming is a topic that oftentimes is heavily overlooked in our society, with many companies prioritizing monetary income over “claims” about our environment. We must consider how time and evolution affect our overall weather patterns, especially in recent years as we continue to ignore healthy emission guidelines and continue down the road of “big business”.
Earlier in the semester, I performed an analysis of weather patterns in Boulder, Colorado. I chose Boulder due to the availability of weather data but also because I go to college at CU Boulder and it felt much more meaningful to analyze how the weather has changed in the town I have lived in through the past four years. In this analysis, I took an in-depth look at the average daily temperature, minimum temperature, and maximum temperature from the start of 2010 up until late 2022. Through this date range, I looked at the overall temperature change in all three statistics, as well as the rate of change in which the overall temperature was changing. This statistic is commonly defined as volatility, which I believe gives valuable insight, especially with temperature. This not only allows you to see how much temperature has changed but also analyze the percent extremity in which temperature has changed between two dates. Through this analysis, I ended up finding that minimum, maximum, and average temperature volatility in Boulder all had positive percent changes greater than 25% from 2010 to 2022. Although this analysis is limited to Boulder, it went to show that the rate of change in temperature in Boulder had gotten much more extreme since 2010. This made me further concerned regarding both climate change and how much time we have left before this planet becomes unsavable. Moving forward, I figured that since I know temperature change is getting more extreme in Boulder, I could analyze further features of weather in order to determine exactly what else is getting more extreme along with average, minimum, and maximum temperature.
In my extension of my Boulder, Colorado weather analysis, I chose to look at four external weather variables aside from minimum, maximum, and average temperature. In the provided Boulder weather dataset, data was also provided on visibility, wind speed, STP, or standard temperature and pressure, as well as DEWP, or dewpoint temperature. In my updated weather extremity analysis, I chose to look at shifts in all four of these variables as well as their volatilities. Variables like wind speed and standard temperature and pressure are harder to directly interpret, but in my analysis, I was largely looking for extreme changes in the rate of change of these variables in order to further justify my initial findings with Boulder temperature volatility. As I mentioned earlier, these original findings included a rate of volatility greater than 25% within all three temperature statistics. I also realized how important it is to analyze volatility and not just overall change, as when I did perform this initial analysis there was just barely any variation in mean temperature from 2010 to 2022, although some variation was displayed in minimum and maximum temperatures (less than five degrees). So, learning from my previous analysis, I gathered data on these four new external weather statistics and began my analysis.
In order to start analyzing this new data, I firstly grouped everything I needed into an organized data frame. I dropped unnecessary columns, keeping the date, visibility, wind speed, standard temperature and pressure, and dewpoint temperature. In order to calculate accurate volatility readings, I then found the average of each statistic and cleaned each column, accounting for missing values and extreme outliers in the data. I then subtracted the means from each value of each column and calculated the squared sums, giving me the variance of each statistic. From variance, it was pretty straightforward to find the volatility of each column, utilizing the standard annualized volatility formula commonly used to analyze stocks. With each volatility calculated as well as some other meaningful statistics, I was ready to further analyze the extremity of weather in Boulder, Colorado, through comparison as well as some visualization.
The very first thing I noticed after calculating the volatilities was that the volatility of standard temperature and pressure (STP) wasn’t significant enough to analyze (a value of 0.4%), and the volatility of dewpoint temperature was pretty much identical to the average temperature volatility in my previous analysis. So, I decided to look much more closely at both visibility and wind speed, which I feel are more “extreme” features of weather. I was surprised to find that from 2010–2022 in Boulder, visibility had a volatility of 48.47%, and wind speed had a shocking volatility of 121.65%. I firstly zoomed in on visibility, which you can see in Figure 1 above. It is very visually apparent that visibility in Boulder has gotten worse from 2010–2022, with the shift in the visualization getting lower and lower each year. I also accounted for the volatility, which shows that visibility has been/is changing at an increasingly extreme rate. More shockingly, I nextly zoomed in on wind speed as I was shocked by the volatility calculation. As you can see in Figure 2 below, wind speed in Boulder over the date range tends to largely average out at around 5–6 mph. If we look at the data, wind speed in 2010 averaged out around 5.8 mph, while in 2022, the wind speed averaged out to be 6.9 mph. As apparent by the volatility calculation, the rate at which wind speed is increasing is getting MUCH more extreme over this time period. Although a one-mile-per-hour difference doesn’t seem like much over 12 years, it’s really the volatility calculation that puts things into perspective. For example, you must consider how low these speeds are to begin with, how I cleaned out extreme outliers in the wind speed data beforehand, and how volatility is accounting for the rate of change in which the rate of change is occurring. So basically, in 2022 in Boulder, we can say that wind speed is changing at a much more extreme rate than it was back in 2010. Volatility is a statistic that oftentimes doesn’t directly show in numbers until a couple of years down the line, which makes me think about how extreme wind speeds and temperatures could get in a decade. Analyzing this data really showed me what I didn’t want to see; I concluded that the temperature extremity was increasing in Boulder in the previous study, but now we can see that other external weather variables are getting MUCH more extreme in variation along with the temperature.
After analyzing the data, volatilities, and coming to some brief conclusions, I wanted to see how exactly boulder temperature tends to stack up with wind speed to further explain my findings, which you can see in Figure 3 below. Although the correlation between the two variables is insignificant, you can generally see the spread of how the two align, with colder and hotter temperatures meaning generally lower wind speed, and the temperature range of 25–65 degrees featuring the most extreme wind speed values. However, you can also use this visualization to explain how as temperatures vary at a more extreme level. For example, in the wintertime instead of staying at around 30 degrees, if the temperature was varying from 10–45, there would be more variation in wind speed, thus accounting for its extremer rate of volatility.
Overall, I found that in Boulder, Colorado, between the years 2010 and 2022 specifically, the volatility of temperature, visibility, and wind speed has gotten much more extreme, with these variables likely influencing each other as they shift. Similarly to my last analysis, these findings really did surprise me, and it makes me nervous about how apparent the change in volatility of these weather variables is. I genuinely believe that a very similar analysis must be performed using weather data across the nation, not just in Boulder. These are findings that I feel really show how weather, temperature, and our planet are getting more extreme/variant over time and I feel that a large-scale study would prove a variety of factors regarding global warming.
