Estimating damages from climate-related natural disasters for the Caribbean at 1.5 °C and 2 °C global warming above preindustrial levels

Burgess, Christopher and Taylor, Michael and Spencer, Nekeisha and Jones, Jhordanne and Stephenson, Tannecia

Nekeisha Omellia Spencer

Regional Environmental Change

2018

8

18

Hurricanes ; Natural Disasters ; Sea Surface Temperature ; Rainfall ; Modulators ; Rainfall ; Artificial Neural Networks ; Distribution Functions ; Neural Networks ; Global Warming ; Natural Disasters ; Climate Change ; Global Climate Models ; Natural Disasters ; Hurricanes ; Neural Networks ; Rainfall ; Caribbean ; Damages ; Climate Change

2297–2312

This paper examines historical and future changes in normalised damages resulting from climate-related natural disasters for the Caribbean. Annualised damages of USD824 million are shown to be non-stationary over the historical period 1964 to 2013. Perturbations of (i) sea surface temperatures (SST) in the tropical North Atlantic (TNA) and (ii) the Atlantic multi-decadal oscillation (AMO) appear to be associated with historical damages. Both the TNA and AMO are known modulators of hurricane activity and rainfall amounts in the Caribbean. Indicative future damages are determined using (i) cumulative distribution functions (CDFs) of perturbed climate states and (ii) an artificial neural network (ANN) model of damages using projected TNA values and the state of the AMO derived from an ensemble of five coupled model intercomparison project phase 5 (CMIP5) global climate models (GCMs) run under the RCP 4.5 scenario. Estimates of future damages are determined when global mean surface temperatures...

Springer Nature B.V.