Detecting and correcting logically inconsistent crop rotations and other land-use sequences

摘要: ABSTRACTConverting multi-year land-use data into crop rotation history is relatively simple in the absence of classification errors, but severely compromised their presence. Several errors can theory be detected with a matrix logically forbidden (or extremely unlikely) year-to-year transitions. We categorized 730 3249 potential transitions among 57 classes western Oregon as being permissible, remaining 77.5% forbidden. Applying these restrictions to eight consecutive years revealed that an average 26.7% apparent agricultural were illogical, contrast only 2.5% and 0.6% urban development forest The most useful correction applied involved replacement original majority-rule values (generated during prior pixel- object-based conversion) second-place categories for fields inconsistent land-uses iden...