How to measure ecosystem stability? An evaluation of the reliability of stability metrics based on remote sensing time series across the major global ecosystems

Abstract

Increasing frequency of extreme climate events is likely to impose increased stress on ecosystems and to jeopardize the services that ecosystems provide. Therefore, it is of major importance to assess the effects of extreme climate events on the temporal stability (i.e., the resistance, the resilience, and the variance) of ecosystem properties. Most time series of ecosystem properties are, however, affected by varying data characteristics, uncertainties, and noise, which complicate the comparison of ecosystem stability metrics ( ESM s) between locations. Therefore, there is a strong need for a more comprehensive understanding regarding the reliability of stability metrics and how they can be used to compare ecosystem stability globally. The objective of this study was to evaluate the performance of temporal ESM s based on time series of the Moderate Resolution Imaging Spectroradiometer derived Normalized Difference Vegetation Index of 15 global land‐cover types. We provide a framework (i) to assess the reliability of ESM s in function of data characteristics, uncertainties and noise and (ii) to integrate reliability estimates in future global ecosystem stability studies against climate disturbances. The performance of our framework was tested through (i) a global ecosystem comparison and (ii) an comparison of ecosystem stability in response to the 2003 drought. The results show the influence of data quality on the accuracy of ecosystem stability. White noise, biased noise, and trends have a stronger effect on the accuracy of stability metrics than the length of the time series, temporal resolution, or amount of missing values. Moreover, we demonstrate the importance of integrating reliability estimates to interpret stability metrics within confidence limits. Based on these confidence limits, other studies dealing with specific ecosystem types or locations can be put into context, and a more reliable assessment of ecosystem stability against environmental disturbances can be obtained.

Bibliographic Information