Hello, and welcome to our posts! We are Team Vāyuh.ai: a group largely consisting of scientists and engineers from the UC Berkeley community with deep expertise in AI, Climate, and all things in between.
This year we have seen, heard of, or experienced a variety of temperature anomalies. For example, in the summer (July and August), a heat wave occurred over the Pacific Northwest and other regions in the Continental United States (CONUS). Temperatures soared, with Seattle hitting a historical high of 95°F. In this article, we will take a closer look at temperature anomalies over time, and the consequences going forward.
First, a few necessities. Climatology - an important but somewhat nebulous term in climate science - is a meteorological variable’s (e.g., temperature, precipitation, etc.) historical average for a specific location and a specific time period. For example, the climatology for Oakland in December can be seen in the plot below Figure 1 (pretty, isn’t it?). In layperson’s terms, climatology can be thought to imply what was true for my grandpa, is true for me.
For all our analysis in this article, we will use data from ERA5 (European Center for Medium Weather Forecast (ECMWF)) which is updated frequently and provides data into the past.
Figure 1: Climatology for Oakland. (Source: Weatherspark.com) . The horizontal axis shows time. The vertical axis shows the temperature. The red (dark) plot is the average high temperature, and the blue (dark) plot is the low. Lighter colors show standard deviations (or deviations from the mean).
Now, let us examine the average anomalies (i.e. deviations from historical averages) for temperature across the Continental United States (CONUS) across the years in Figure 2 below.
Figure 2: The anomalies here were calculated by averaging across all locations (in CONUS) and dates (in a year). On the horizontal axis, we see time (years). On the vertical axis, we see the temperature anomalies at 2m height. That is, surface temperature. We see an increasing trend of anomalies here over the years.
While the previous plot shows that there is an increase in anomalies, this is an incomplete story. For example, in a month, we could have 15 days of positive (heating) anomalies and 15 days of negative (cooling) anomalies. If we averaged the month, we would see no change in the average temperature.
So, let us break this apart further and look at heating and cooling anomalies separately. May they be increasing or decreasing? In Figures 3 and 4, we see that the heating anomalies are decreasing over the years while the cooling anomalies are increasing. That is, the number of days where we have 5 or 10 degrees Celsius ( ~15 to 30 Fahrenheit) anomalies are changing.
Figure 3: The anomalies here were calculated by averaging across all locations (in CONUS) and dates (in a year) but broken into heating and cooling anomalies. An anomaly here is defined as over 5’ C in deviation.
Figure 4: The anomalies here were calculated by averaging across all locations (in CONUS) and dates (in a year) but broken into heating and cooling anomalies. An anomaly here is defined as over 10’ C in deviation.
These trends reveal that, in 2020, CONUS experienced 100 days a year where temperatures were warmer by at least 5 degrees Celsius. Furthermore, at least 50 of these days are deviating over 10 degrees Celsius. Days with extreme heat are more common now than they were just 50 years ago.
Heat waves are expected to cost hundreds of millions of dollars in loss in the next few decades. What do future scenarios look like? How might this affect other perils like wildfires, extreme precipitation, or severe storms? In these articles, we will explore a variety of topics related to climate change, forecasts, AI (as pertaining to climate), and their impact on society.
Коментарі