Long-term, medium spatial resolution annual land surface phenology with a Bayesian hierarchical model

• DOI: 10.1016/j.rse.2021.112484
• Publication Date: 2021-05-06T21:44:13Z
• Authors: Xiaojie Gao; Josh M. Gray; Brian J. Reich
• Keywords: Temporal resolution

Land surface phenology (LSP) is a consistent and sensitive indicator of climate change effects on Earth's vegetation. Existing methods estimating LSP require time series densities that, until recently, have only been available from coarse spatial resolution imagery such as MODIS (500 m) AVHRR (1 km). products these datasets improved our understanding phenological at the global scale, especially over era (2001-present). Nevertheless, may obscure important finer scale patterns longer-term changes. Therefore, we developed Bayesian hierarchical model to retrieve complete annual sequences Landsat (1984-present), which has medium (30 but relatively sparse temporal frequency. Our approach uses Markov Chain Monte Carlo (MCMC) sampling quantify individual phenometric uncertainty, when considering long with variable observation quality density, rarely demonstrated. The estimated spring had strong agreement ground records Harvard Forest (R2 = 0.87) Hubbard Brook Experimental 0.67). were recently released 30 m product, MSLSP30NA, in its period 2016 2018 0.86 0.73 for autumn phenology, respectively). an step forward extending back it accomplishes both critical goals retrieving accurately uncertainty.