Nowcasting gas demand based on timely available data sources

Abstract

Natural gas consumption is a critical indicator in both economic analysis and environmental monitoring. Despite its significance for economic policy, energy security, and climate goals, official consumption data are typically published with a considerable time lag. In this work, we describe a method for nowcasting the monthly natural gas consumption of EU member states. Our approach is based on linear regression models and utilizes an external source of gas flow data. In addition, we propose a complementary modeling strategy for countries lacking sufficient flow data, by leveraging correlations across countries. The methodology has achieved promising results in the European Statistics Awards Energy Nowcasting Challenge – GAS.