An Investigation into Land Surface Temperature-Based Climate Indices

Monitoring climate change is crucial to society, industry, and the environment to assess their vulnerability to such extreme temperature events and plan adaptation and mitigation strategies. Temperature-based climate indices, which are used for climate monitoring, are usually calculated using near-surface air temperature (NSAT) from weather stations. A limitation of NSAT is the inhomogeneity of station distribution across the globe, with some regions under-sampled. This study evaluates the potential of using land surface temperature (LST) retrieved via satellite to calculate temperature indices, because low Earth-orbiting satellites can observe almost all of the Earth. This work compares indices calculated from NSAT to those produced from LST, and the indices are aggregated by Intergovernmental Panel on Climate Change AR6 regions. Results show that there is often moderate to high correlation between indices produced using NSAT and LST. In addition, it is found that monthly to seasonal changes in indices are similar for both variables, particularly in percentile-based indices. Threshold-based indices could not be calculated for some regions due to low data availability, and therefore this work also applies “relative” indices, where indices are normalised relative to the amount of data available. These generally also demonstrate high correlations between NSAT and LST and can better show when events, such as heatwaves, occur. Lastly, this work presents the potential of an LST-based temperature indices dataset and how uncertainties could be evaluated.