Canadian Forest Service Publications

Probabilistic Tracking of Annual Cropland Changes over Large, Complex Agricultural Landscapes Using Google Earth Engine. Xiong, S.; Baltezar, P.; Crowley, M.A.; Cecil, M.; Crema, S.C.; Baldwin, E.; Cardille, J.A.; Estes, L.

Year: 2022

Issued by: Great Lakes Forestry Centre

Catalog ID: 41086

Language: English

Availability: Not available through the CFS (click for more information).

Available from the Journal's Web site.

† This site may require a fee

Mark record

Plain Language Summary

The objective of this article are to present a Bayesian method for mapping land features using multi-source Earth observations including optical and synthetic-aperture-radar sources. The geographic scope of the study focused on Zambia, the sub-Saharan country, and spanned across the 2000 to 2015 time period. Methodological advancements included using an unsupervised Bayesian classification technique with shapelet and slope thresholding of the cropland classifications. The key findings are that Bayesian data fusion and shapelet/slope-based thresholding are useful methods for synthesizing optical and SAR data. The scientific impact and importance are that this method can be used in classification scenarios when training data are scarce.