More About Global Warming Levels

What is the main difference between GWLs and the emissions scenario approach?

The primary distinction between a GWL approach and the more common emissions scenarios approach is in the articulation of regional climate projections. In the GWL approach, regional climate projections are expressed in relation to a specified increase in global average temperature above pre-industrial levels (or other baseline period), with no reference to when this increase in global average temperature could be reached. The emissions scenario approach presents climate projections by emissions scenario and, thus, indicates the trajectory of climate change against time for a given scenario. In this approach, the main uncertainty in regional climate projections is the magnitude of climate change for a specified emissions scenario. For the GWL approach, the main uncertainty in regional climate projections shifts to the time when specific GWLs will be reached. For some applications, e.g., for economic analyses, the lack of a continuous time series between now and a future time horizon may limit the utility of the GWL approach.

GWLs and baseline periods

It is important to note that GWLs are expressed as changes (or anomalies) calculated with respect to a baseline reference period. While the Paris Agreement² refers to the pre-industrial period and IPCC uses specific definitions of (near) pre-industrial periods, most commonly 1850-1900, different baseline periods may be used to define GWLs. For example, the Climate Resilient Buildings and Core Public Infrastructure Report (CRBCPI)³, which provides future building design values, uses 1986-2016 as its baseline. As such, care needs to be taken when comparing projections based on GWLs to ensure that baselines are understood. 

It is possible, however, to reconcile differences in baselines to standardize the definition of GWLs. For example, Figure 1 illustrates the timing of GWLs using two different baseline periods: the IPCC pre-industrial period (1850-1900; right axis) and the 1986-2016 baseline (left axis) used by CRBCPI. About 0.8°C of global warming has occurred between the IPCC and CRBCPI baselines.

Noting that the Paris Agreement target is to limit global warming to “well below 2°C above pre-industrial levels”, in this example we compare 1.5°C and 2°C above 1850-1900 with the corresponding GWLs using the CRBCPI baseline period. The 2°C GWL above the 1850-1900 level corresponds to about 1.2°C of warming if using the 1986-2016 baseline.  The  1.2°C of warming is projected to occur around 2040. Similarly, the more stringent 1.5°C target corresponds to a GWL of about 0.7°C using the CRBCPI baseline of 1986-2016 and is projected to occur in the mid-2020s.

The CRBCPI report and the Pacific Climate Impacts Consortium’s Design Value Explorer (DVE) tool provide information about future design values for a range of GWLs to supplement the information in the National Building Code of Canada. The Future Building Design Value summaries available on ClimateData.ca provide information for two GWLs, 1.5°C and 3.0°C, relevant for the design of short- to medium-term design service life components (10 to 30 years) and medium- to long-term design service life components (50+ years), respectively. These GWLs are relative to the 1986-2016 baseline and correspond to GWLs of 2.3°C and 3.8°C using the pre-industrial baseline.

Another baseline period that you may encounter is 1971-2000. About 0.5°C of global warming has occurred between the IPCC’s pre-industrial baseline reference period of 1850-1900 and 1971-2000. This means that a GWL of 2°C relative to 1971-2000 is roughly equivalent to a GWL of 2.5°C from the pre-industrial 1850-1900 baseline. Finally, while most of the time a GWL from pre-industrial means 1850-1900, there are different definitions of “pre-industrial”, so it is always important to check the actual time period that is being used to define the baseline.

What do you need to be aware of?

The GWL approach is based on robust relationships between regional climate changes and global average temperature shifts. This approach results from research into how regional temperature and precipitation extremes correlate with changes in global average temperature¹. However, it is important to note that there are certain limitations that may affect the applicability of this approach across different climate variables, regions, and emissions pathways:

1. The GWL approach is generally more successful at identifying robust relationships between global and regional scales for temperature-based indices, since they are directly related to changes in radiative forcing, which in turn are related to GHG emissions. Many temperature-based indices have been shown to scale linearly with global temperature change, e.g., number of frost days (minimum temperature < 0°C) and summer days (maximum temperature > 25°C).  In contrast, it is not always possible to identify robust relationships between the regional response of precipitation-based and other climate indices and global temperature change, and the strength of these relationships may vary by region⁴. In addition, the GWL approach does not tend to work well with variables that exhibit a substantial delay in their response to global temperature increase, e.g., sea level rise and glacial melt.  However, plotting global- versus regional-scale changes is a straightforward method for identifying robust relationships.
2. For heavy precipitation events, although the ensemble-average response does scale robustly with global average temperature change, individual model projections may diverge strongly from this average response8. For precipitation, internal ‘natural’ climate variability plays a larger role at regional-to-local scales and precipitation extremes are much more spatially variable than temperature extremes.  Consequently, it can be more difficult to identify strong regional- to global-scale relationships for variables describing precipitation extremes. That said, however, for Canada as a whole, precipitation indices, including extremes, have been shown to scale robustly with GWLs.⁶
3. Regional-scale relationships may also change over time, particularly if the response of a climate variable is strongly dependent on local forcing such as aerosols, land-use, and land cover change. For example, local aerosol forcing influences local precipitation amount, but becomes less important for larger regions. Also, projections of regional extreme precipitation may be different in emissions pathways which have the same, or similar, radiative forcing (e.g., SSP2-4.5 and RCP4.5) because the underlying assumptions for these emissions pathways may result in differences in local aerosol forcing.
4. While robust relationships exist between the climate response of large regions when compared to the global scale, these relationships are expected to become less reliable for smaller regions, meaning that relationships between local and global scales are not always sufficiently strong to support the use of the GWL approach. For smaller regions, internal climate variability and local climate processes play a more significant role, increasing the ‘noise’ in the data and making it more difficult to see a distinct relationship between a climate variable and global average temperature change.
5. Strong, quantitative (e.g., linear) relationships between climate variables and GWLs are less likely to be apparent under lower emissions scenarios, such as RCP2.6 or SSP1-2.6. This is due to a less strong climate signal when compared with the background noise of natural climate variability in projections using these emissions scenarios.

In short, expressing climate projections using the GWL approach makes it easier to relate global warming targets to regional impacts and is a useful complement to the emissions scenario approach for presenting climate change projections.