Climate adaptation for politicians: shocking a macroeconomic model with stochastic natural disaster impacts

Historically, much of natural disaster impact modelling has focussed on the damage to private assets. But to a government decision-maker it is not always clear how impacts to individual assets translate into a cost to the national economy. Understanding this important for adaptation decision-making: which communities are resilient enough to withstand and recover from disasters? When is a disaster large enough to have regional or national knock-on effects? What is the long-term, compounding economic cost of inaction?

In collaboration with the Thai and Egyptian governments, we have prototyped a coupling between an open-source probabilistic disaster impact model (CLIMADA) and an open-source macroeconomic model (DGE-CRED). We present a modelling framework and codebase designed for more data-scarce environments, where data and modelling can be collected and iterated on in the space of weeks or months.

The coupled model starts with publicly available, open-source data (with their known limitations). Data and insights from local partners are then critical to calibrate and enhance the data. The model creates thousands of plausible future timelines of shocks from natural disasters (fluvial flood, heatwave and drought), models their impacts on the economic sectors of most interest to our partners (agriculture, manufacturing, energy, tourism/services), and simulates their short- and long-term macroeconomic impacts (on e.g. GDP, employment rates, prices, well-being indicators) and, guided by local knowledge, the benefits of different adaptation measures.