资源环境科技发展态势分析平台

Improved seasonal forecasting can help to mitigate the impacts of prolonged extreme heat, a prominent climate hazard in the south-central United States. Temperatures in this region exhibit significant temporal autocorrelation on monthly to seasonal timescales, and the magnitude of temperature persistence can provide useful information for seasonal climate forecasts. This study examines spatiotemporal variations in temperature persistence in the south-central US using high-resolution temperature data from 1900-2015. Consistent with previous studies, temperature persistence varies substantially across the region and is strongest during the summer months. Statistically significant temporal autocorrelations are primarily observed on a monthly timescale; however, at some locations, temperature persistence is also statistically significant at 3 mo timescales. This research builds upon previous findings by using the skill of monthly persistence forecasts to examine how temperature persistence varies with the magnitude of temperature anomalies. Results show that forecast skill regularly increases as the magnitude of the temperature anomalies increase. For example, anomalously warm temperatures during the spring can serve as an early warning for a warmer than normal summer. Our results suggest that persistence forecasts can aid in the subseasonal-to-seasonal prediction of temperatures, particularly when anomalous conditions occur.