Temperature forecasting and derivatives pricing in the Yangtze River economic belt of China

In the Yangtze River Economic Belt, climate variability significantly affects sectors such as agriculture, energy, and related industries, making accurate temperature forecasting essential. This study develops and compares Seasonal Autoregressive Integrated Moving Average (SARIMA) and Ornstein-Uhlenbeck (O-U) models using temperature data from 11 provincial-level regions spanning 2004–2023. After model evaluation, the more accurate forecasting model was selected to support the design of temperature-indexed weather derivatives via option pricing theory. Our findings reveal: (1) temperature series become stationary after first-order differencing, validating their suitability for time series modeling; (2) both SARIMA and O-U models produce predictions closely aligned with observed data; (3) SARIMA exhibits lower forecasting errors for cumulative cooling degree days (CDDs), confirmed through Monte Carlo simulations; and (4) option pricing results show that increased climate volatility raises derivative premiums, reflecting heightened climate risk.This research demonstrates the potential of weather derivatives as a risk mitigation tool in the Yangtze River Economic Belt, contributing to climate-resilient economic development.