Estimating changes and trends in ecosystem extent with dense time‐series satellite remote sensing

• DOI: 10.1111/cobi.13520
• Publication Date: 2020-04-23T02:05:31Z
• Authors: Calvin K.F. Lee; Emily Nicholson; Clare Duncan; Nicholas J. Murray
• Keywords: Deforestation

Abstract Quantifying trends in ecosystem extent is essential to understanding the status of ecosystems. Estimates loss are widely used track progress toward conservation targets, monitor deforestation, and identify ecosystems undergoing rapid change. Satellite remote sensing has become an important source information for estimating these variables. Despite regular acquisition satellite data, many studies change use only static snapshots, which ignores considerable amounts data. This approach limits ability explicitly estimate trend uncertainty significance. Assessing accuracy multiple snapshots also requires time‐series reference data often very costly sometimes impossible obtain. We devised a method that uses all available Landsat imagery. dense time series classified maps accounted covariates affect estimates (e.g., time, cloud cover, seasonality). applied this Hukaung Valley Wildlife Sanctuary, Myanmar, where deforestation greatly affecting lowland rainforest. generalized additive mixed model forest from more than 650 image classifications (1999–2018). Forest declined significantly at rate 0.274%/year (SE = 0.078). 91.70% 0.02) study area 1999 86.52% 2018. Compared with snapshot method, our improved estimated by allowing significance testing confidence intervals incorporation nonlinear relationships. Our can be significant over reduces need extensive through provides quantitative uncertainty.