作者: François Waldner , Steffen Fritz , Antonio Di Gregorio , Pierre Defourny
DOI: 10.3390/RS70607959
关键词:
摘要: Timely and accurate information on the global cropland extent is critical for applications in fields of food security, agricultural monitoring, water management, land-use change modeling Earth system modeling. On one hand, it gives detailed location where to analyze satellite image time series assess crop condition. other isolates agriculture component focus security monitoring potential impacts climate lands. The class often poorly captured land cover products due its dynamic nature large variety agro-systems. overall objective was evaluate current availability datasets order propose a strategic planning effort distribution future mapping activities and, therefore, maximize their impact. Following very comprehensive identification collection national maps, multi-criteria analysis designed at country level identify priority areas mapping. As result, highlighted regions, such as Western Africa, Ethiopia, Madagascar Southeast Asia, remote sensing community efforts. A Unified Cropland Layer 250 m year 2014 produced combining fittest products. It assessed using validation yields an accuracy ranging from 82%–94%. Masking with forest map reduced commission errors 46% down 26%. Compared GLC-Share International Institute Applied Systems Analysis-International Food Policy Research (IIASA-IFPRI) significant spatial disagreements were found, which might be attributed discrepancies definition. This advocates shared definition cropland, well relevant systematically existing maps.