Harbingers are early warnings of imminent ecosystem collapse and thus are aids to preventing land degradation. Dynamic systems are hypothesized to exhibit dampening or inflation of critical attributes at or near a threshold, which is a decreasing or increasing spatial or temporal trend, respectively. This behavior is diagnostic of a state change and can be operationalized as an early detection system. Consequently, we used a time series from 1972 to 1997 of seasonal soil-adjusted vegetation index (SAVI) data, a proxy for canopy cover that was derived from Landsat imagery of the Marine Corps Air Ground Combat Center. We used dynamical, trend, and autocorrelation function (ACF) time series analysis to find that the time series had an increasing linear trend that correlated with wet periods of the Pacific Decadal Oscillation (PDO) and the El Niño-Southern Oscillation (ENSO). High and low SAVI values were in wet and dry basins of attraction with a rapid shift from dry to wet period from 1981 to 1982. Mean SAVI dampening appears to have occurred 2–3 years prior to the shift. Consequently, this study suggests that this dampening trend of the mean SAVI can be used as a harbinger of land degradation.

Full citation

Washington-Allen R.A., Ramsey R.D., Van Niel T.G., West N.E. (2014). Detection of Harbingers of Catastrophic Regime Shifts in Drylands. Ch. 13 in: Guntenspergen G.R., ed. Application of Threshold Concepts in Natural Resource Decision Making. New York, NY: Springer.