Part 3 - Annual maximum temperature (TXx)

Prepared by Mathias Hauser.

Until now we have mostly looked at the global-mean annual-mean temperature. In Part 3 we will expand our view to local changes of the annual maximum temperature (abbreviated as TXx).

TXx is computed as the annual maximum of daily maximum temperature. It is the highest temperature measured in a year. So for mid-latiude regions it usually occurs in summer. We will compare the spatial pattern and warming of the model that warmed the most and the least.

Learning goals

  • programming goals

    • create a map plot

  • scientific and data analyis goals

    • learn how differences in the global mean temperature translate to differenes in other variables

    • to show spatial data we often aggregate over time (in contrast to Part 2 where we showed temporal data by averaging over the spatial dimensions)

Preparation

Create a new notebook and select ipp_analysis as kernel. Rename the notebook to pp_txx_name.ipynb where you replace name with your ETH username. Don’t forget to add a title, your name and the imports!

Load TXx data

As always we start by loading and exploring the relevant data.

Model with least warming

  1. Open the gridded txx data file of the model that warmed the least (over 20 years): INM-CM5-0.

    • The data is available at "../data/cmip6/tasmax/txx/txx_tasmax_day_INM-CM5-0_historical-ssp585_r1i1p1f1_g025.nc"

    • Here the historical data and the projection is already concatenated.

  2. Familiarise yourself with the data

    • What variables are on the dataset?

    • Create an exploratory plot of the data.

Model with most warming

  1. Which model warmed the most? Check Part 2.

  2. Read the data as well. Choose a good name for the variable - so that you can distinguish it from the one above.

Data preparation

The goal is to create a map of TXx anomalies for the end of the 21st century. Before we can create the plot we need to prepare the data for both models:

  1. Calculate anomalies over 1850-1900 using computation.calc_anomaly(...).

  2. Select the last 20 years (2081-2100) and calculate the mean over time.

Compare spatial pattern of warming - plot

To effectively compare the data we will now plot both 20-year means on a map.

  1. Prepare the figure - check the intro how to do this. The figure must have two subplots with a projection of your liking.

  2. Add coastlines to both subplots.

  3. Add the 20 year (2081-2100) mean temperature change of the two models.

  4. Make sure the two models show the same temperature range or levels.

  5. Remove the two colorbars (if you use xarray to create the plots).

  6. Add a single colorbar (plt.colorbar(h, ax=axs, orientation="horizontal", extend="max", aspect=30, pad=0.05) should give a reasonable result).

  7. Add a the name of the model as title.

  8. Use f.suptitle to describe what variable is shown and the used time period.

Interpretation of the figure

  1. How do ocean and land areas contrast?

  2. How do the temperature anomalies of the two models compare?

  3. Reflecting on what you learned in Part 2 - can you explain why CanESM5 shows so much more warning than INM-CM5-0?

What have you learned so far?

Reflect back on Part 3. Note what you learned at the end of the notebook.

  • Which part of creating the map plots did you find most challenging?

What have you learned so far?

You now have two maps that show temperature projections at the end of the 20th century. These again highlight the differences between the two models with the lowest/ highest warming and raises the question which of those is more realistic. In the next part we will take another look at this data and see that the differences in TXx displayed here are mostly due to the large differences in the global mean temperature.

This concludes Part 3.