Chapter 8 Introduction

This project is to produce a warming stripes visualization, first released by Ed Hawkins, a climate scientist in the National Centre for Atmospheric Science (NCAS) at the University of Reading in 2018.

For this project, we use the data relesed by the UK Met Office. We selected the Tropics (30 deg S to 30 deg N) Annual dataset, which contains a time series presented as temperature anomalies (deg C) that are relative to 1961-1990 for the years 1850 to 2019.

The dataset consists of 170 records with 12 columns as described below:

  1. Column 1 is the date.
  2. Column 2 is the median of the 100 ensemble member time series.
  3. Columns 3 and 4 are the lower and upper bounds of the 95% confidence interval of bias uncertainty computed from the 100 member ensemble.
  4. Columns 5 and 6 are the lower and upper bounds of the 95% confidence interval of measurement and sampling uncertainties around the ensemble median. These are the combination of fully uncorrelated measurement and sampling uncertainties and partially correlated uncertainties described by the HadCRUT4 error covariance matrices.
  5. Columns 7 and 8 are the lower and upper bounds of the 95% confidence interval of coverage uncertainties around the ensemble median.
  6. Columns 9 and 10 are the lower and upper bounds of the 95% confidence interval of the combination of measurement and sampling and bias uncertainties.
  7. Columns 11 and 12 are the lower and upper bounds of the 95% confidence interval of the combined effects of all the uncertainties described in the HadCRUT4 error model (measurement and sampling, bias and coverage uncertainties).

The dataset can be downloaded from here.
Information on the dataset format can be found here.

More details are given in the paper introducing the dataset.