Consider the global land and ocean temperature anomalies from January 1880 to May 2017. The monthly data are available from the National Oceanic and Atmospheric Administration (NOAA) at www.ncdc.noaa.gov and are in the file globaltemp18802017.

csv. The data are in degrees Celsius and the base period is 1901–2000. The temperature anomalies show an upward trend so consider the change series, i.e. the first difference. Is the differenced series nonlinear? Perform some tests and draw a conclusion. Does the series show threshold nonlinearity? Why?