The supply of affordable crop insurance is hampered by the existence of systemic weather risk which results in large risk premiums. In this article, we assess the systemic nature of weather risk for 17 agricultural production regions in China and explore the possibility of spatial diversification of this risk. We simulate the buffer load of hypothetical temperature-based insurance and investigate the relation between the size of the buffer load and the size of the trading area of the insurance. The analysis makes use of a hierarchical Archimedean copula approach (HAC) which allows flexible modeling of the joint loss distribution and reveals the dependence structure of losses in different insured regions. Our results show a significant decrease of the required risk loading when the insured area expands. Nevertheless, a considerable part of undiversifiable risk remains with the insurer. We find that the spatial diversification effect depends on the type of the weather index and the strike level of the insurance. Our findings are relevant for insurers and insurance regulators as they shed light on the viability of private crop insurance in China.