If you have been following the news lately, you might have noticed that many regions around the world have experienced extreme heat waves and record-breaking temperatures. A report from the United Nations World Meteorological Organization (WMO) found that Earth just experienced its hottest series of three months in a row from June to August 2023. This has not only caused discomfort and health problems for millions of people, but also triggered wildfires, droughts, and power outages.
But how do we know that these events are part of a larger trend of global warming? How do we measure the temperature of the entire planet? And how reliable and accurate are these measurements?
In this post, I will try to answer these questions by explaining how the global average temperature is calculated and what are the challenges and uncertainties involved in this process.
What is the global average temperature and why does it matter?
The global average temperature, also known as the global mean surface temperature (GMST), is a single number that represents the average temperature of the Earth’s surface over a given period of time. It is one of the most widely used indicators of climate change, as it reflects how much heat energy is stored in the Earth’s system.
The global average temperature is important because it affects many aspects of life on Earth, such as weather patterns, sea level rise, ice melt, biodiversity, agriculture, health, and economy. A small change in the global average temperature can have significant impacts on the environment and society.
For example, according to the Intergovernmental Panel on Climate Change (IPCC), the global average temperature has increased by about 1.1°C since the pre-industrial era (1850-1900), and this increase has already caused more frequent and intense heat waves, droughts, floods, storms, and wildfires. The IPCC also warns that if the global average temperature rises by more than 1.5°C or 2°C above the pre-industrial level, the risks and consequences will be much more severe and irreversible.
Therefore, monitoring and reporting the global average temperature is essential for understanding the current state and trends of the climate system, as well as for informing policy decisions and actions to mitigate and adapt to climate change.
How are temperature measurements collected?
The first step in calculating the global average temperature is to collect temperature measurements from various sources around the world. These sources include land-based stations, such as weather stations, airports, or research stations; ocean-based platforms, such as ships, buoys, or floats; and space-based instruments, such as satellites or radiosondes.
These measurements are then transmitted to different organizations that collect and publish temperature data, such as the National Oceanic and Atmospheric Administration (NOAA), the National Aeronautics and Space Administration (NASA), the World Meteorological Organization (WMO), or the Met Office Hadley Centre. These organizations use different methods and procedures to standardize, quality-control, and adjust these data for biases and gaps.
Some of the adjustments that are made include correcting for changes in measurement methods or instruments over time; accounting for differences in elevation, location, or exposure of measurement sites; filling in missing data or removing outliers; and homogenizing data from different sources or regions.
These adjustments are necessary to ensure that the data are consistent and comparable over time and space. However, they also introduce some uncertainty and error in the data, which need to be quantified and reported.
How are temperature anomalies calculated?
The second step in calculating the global average temperature is to compute temperature anomalies for each measurement. A temperature anomaly is the difference between a measured temperature and a reference value (such as a long-term average) for a given location and time. For example, if the measured temperature at a certain place on a certain day is 25°C, and the reference value for that place and day is 20°C, then the anomaly is +5°C.
Temperature anomalies are preferred over absolute temperatures for calculating the global average temperature for several reasons. First, anomalies are more representative of climate change than absolute temperatures, as they reflect how much warmer or cooler a place is compared to its normal or expected condition. Second, anomalies are more homogeneous and comparable than absolute temperatures, as they reduce the effects of factors that vary spatially and temporally, such as elevation, latitude, season, or time of day. Third, anomalies are more robust and reliable than absolute temperatures, as they are less affected by gaps or errors in the data.
However, calculating anomalies also involves some choices and assumptions that can affect the results. For example, one of the choices is how to define the reference value for each location and time. Different reference values can result in different anomalies, depending on the baseline period and the spatial and temporal resolution used. NOAA, for instance, uses a 20th century average (1901-2000) as the reference value for its global temperature dataset, while NASA uses a more recent average (1951-1980) for its dataset. Similarly, some datasets use monthly averages as the reference values, while others use daily or annual averages.
Another choice is how to deal with missing or incomplete data. For example, some regions of the world, such as the polar areas or the oceans, have fewer or no measurements available, especially in the past. This means that there are gaps in the data that need to be filled or estimated somehow. Different methods can be used to fill these gaps, such as interpolation, extrapolation, or statistical modeling. However, these methods also introduce some uncertainty and error in the anomalies, which need to be quantified and reported.
How are temperature anomalies averaged?
The third and final step in calculating the global average temperature is to average the temperature anomalies over the entire globe. This step involves weighting and combining the anomalies from different sources and regions to obtain a single mean value that represents the global temperature for a given period of time.
However, averaging temperature anomalies is not a straightforward or simple task, as it involves some challenges and uncertainties. For example, one of the challenges is how to account for the uneven spatial coverage and distribution of measurements. Some regions of the world have more measurements than others, which means that they have more influence on the global average than they should. To avoid this problem, some techniques are used to weight or adjust the anomalies according to their spatial representativeness or area coverage. For instance, some datasets use a grid-based approach, where they divide the globe into equal-sized cells and calculate an average anomaly for each cell. Then they weight each cell by its area and sum them up to get the global average. Other datasets use a station-based approach, where they group nearby stations into clusters and calculate an average anomaly for each cluster. Then they weight each cluster by its spatial variability and sum them up to get the global average.
Another challenge is how to deal with the variations and discrepancies among different datasets and products. Different datasets and products may use different sources of data, methods of adjustment, definitions of reference values, techniques of averaging, or periods of reporting. These differences can result in different estimates of the global average temperature, which can sometimes be contradictory or confusing. For example, according to NOAA’s dataset, 2016 was the warmest year on record globally, with an average temperature anomaly of +0.94°C above the 20th century average. However, according to NASA’s dataset, 2020 was the warmest year on record globally (tied with 2016), with an average temperature anomaly of +1.02°C above the 1951-1980 average.
Therefore, it is important to compare and evaluate different datasets and products carefully and critically, and to understand their strengths and limitations. It is also important to report and communicate the uncertainty and error associated with each estimate of the global average temperature, as well as the confidence intervals and significance levels.
To learn more you can check out these references:
How temperature measurements are collected and adjusted:
- GISS Surface Temperature Analysis (GISTEMP)
- The Raw Truth on Global Temperature Records
- HadCRUT5: a new global surface temperature dataset
How temperature anomalies are calculated and compared: