Abstract

Climate-induced forest mortality is being widely observed across the globe, impacting water and carbon cycles at multiple scales. Given that physiological mechanisms causing mortality are not yet fully understood and empirical relations between climatology and mortality are subject to change with species competition and adaptation, challenges in predicting mortality still remain. This study uses the temporal loss of resilience, a phenomenon detected in a wide range of systems while approaching a tipping point, as an early warning signal (EWS) to predict forest mortality directly from remotely sensed vegetation dynamics. An application of the proposed approach in California forests indicates that EWS can often be detected between 6 to 19 months prior to mortality. The EWS shows species-specific relation with mortality, which captures mortality variations both throughout the years and across the space. The expressed associations highlight the potential of EWS for operational monitoring and near-term prediction of forest mortality.

Citation

Kumar, Mukesh, Yanlan Liu, Gabriel George Katul, Amilcare M Porporato (2019): Detecting Climate-Stress Induced Forest Mortality Before the Canonical Symptoms Appear. American Geophysical Union Fall Meeting, San Francisco, CA, December 9-13, 2019.

This Paper/Book acknowledges NSF CZO grant support.