Remotely-Sensed Early Warning Signals on a Critical Transition of Tree Mortality

Kurzfassung

Land degradation and desertification have severe negative effects on land-use, water resources, soil stability, agriculture and biodiversity. Drylands cover 33.8% of northern Mediterranean countries: foremost Spain and Cyprus with more than 65%. The European Environment Agency (EEA) indicated that 8% of the territory of the European Union (mostly in Bulgaria, Cyprus, Greece, Italy, Romania, Spain and Portugal) experience a ‘very high’ or ‘high’ sensitivity to desertification. For the island of Cyprus, 9.68% of the land area was found to be susceptible to land degradation. Sensitive ecosystems such as lakes, wetlands and forests may experience irreversible changes, called ‘regime shifts’ or ‘critical transitions’, with extreme changes of dynamic environmental variables. In case a certain threshold / tipping point in the resilience capability of an ecosystem is exceeded, the state of the ecosystem might become irretrievable. Early warning signals can be derived from statistical metrics that have been proposed to identify impeding regime shifts. Yet, regular monitoring of key ecosystem variables is limited by the lack of appropriate in-situ measurements and data and therefore the identification of these transitions is ambiguous. Correlating the available data with the available Earth Observation (EO) data, can enable the assessment of ecosystem resilience against global change. The proposed study presents a holistic view on remotely-sensed, meteorological (drought) data and metrics-based indicators to identify critical transitions in a forest ecosystem. Mediterranean forests experience regime shifts frequently. The potential of using detailed EO-based time-series for monitoring Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI) and Land Surface Temperature (LST), in combination with statistical metrics can provide a comprehensive analysis to determine and characterise the timing of regime shifts and specifically of tree mortality of Cyprus forests in Akamas, Machairas and Stavrovouni. The analysis of these three study areas that faced tree mortality in three different periods allows us to analyse and understand under which conditions critical transitions appear. Statistical metrics of climate variables, as well as EO-based time-series can describe the conditions leading to tree mortality in forest ecosystems. Eventually, a detailed EO-based time-series monitoring and analysis of statistical metrics could propose a new composite indicator set that may serve as early warning signal indicating the risk of increased tree mortality, which might further lead to land degradation or even desertification.