We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Intercomparison on Four Irrigated Cropland Maps in Mainland China.
- Authors
Yizhu Liu; Wenbin Wu; Hailan Li; Imtiaz, Muhammad; Zhaoliang Li; Qingbo Zhou
- Abstract
Wide-coverage spatial information on irrigated croplands is a vital foundation for foodsecurity and water resources studies at the regional level. Several global irrigated-cropland mapshave been released to the public over the past decade due to the efforts of the remote sensingcommunity. However, the consistency and discrepancy between these maps is largely unknownbecause of a lack of comparative studies, limiting their use and improvement. To close this knowledgegap, we compared the latest four irrigated-cropland datasets (GMIA, GRIPC, GlobCover, andGFSAD) in mainland China. First, the four maps were compared quantitatively and neutral regional-and provincial-level statistics of the relative proportions of irrigated land were obtained throughregression analysis. Second, we compared the similarities and discrepancies of the datasets on spatialgrids. Furthermore, the contributions of mosaic cropland pixels in GlobCover and GFSAD were alsoanalyzed because of their extensive distribution and ambiguous content. Results showed that GMIAhas the lowest dispersion and best statistical correlation followed by GRIPC, while the correspondingfeatures of GlobCover and GFSAD are approximately equal. Spatial agreement of the four mapsis higher in eastern than western China, and disagreement is contributed mostly by GlobCoverand GFSAD. However, divergence exists in the ratios of the different agreement levels, as well astheir sources, on a regional scale. Mosaic pixels provide more than half of the irrigated areas forGlobCover and GFSAD, and they include both correct and incorrect information. Our results indicatea need for a uniform quantitative classification system and for greater focus on heterogeneous regions.Furthermore, the results demonstrate the advantage of numerical restriction in the calculations.Therefore, special attention should be paid to integrating databases and to exploring remote sensingfeatures and methods for spatial reconstruction and identification of untypical irrigation areas.
- Publication
Sensors (14248220), 2018, Vol 18, Issue 4, p1
- ISSN
1424-8220
- Publication type
Article
- DOI
10.3390/s18041197