SPATIAL CORRELATION ANALYSIS OF CROP YIELD IN THE MIDDLE AND WEST OF JILIN PROVINCE

In this paper,spatial correlation of crop yield in the middle and west of Jilin province is analyzed by us-ing the method of geostatistics swmivarivogram,taking the NDVI of NOAA/AVHRR spectrum data as the regionalized vari-able,aiming to provide theory and practical basis for field sampling of crop yield estimation using remote sensing.The ratio of nugget variance and sill of semivariograms are 21.1% and9.7% in the west and middle regions in Jilin Province respectively.This shows that the crop yields are spatially correlated.The degree and range of correlation are far different in the different situations.In the west test region,the range is 49.9km and the sill is 0.00019 .In the middle test re-gion,the range is 16.5km and the sill is 0.00453.The dissimilarity in the western test region is larger than that in the middle one.The range in which the correlation existed of the former is far larger than the later.Different characteristics of spatial correlation of crop yield are decided by the environmental factors.Samples for crop yield estimation should be extracted according to the characteristic of spatial distribution of crop yield to promote the efficiency of sampling.

Spatial correlation analysis of crop yield in the middle and west of Jilin Province

In this paper, spatial correlation of crop yield in the middle and west of Jilin Province is analyzed by using the method of geostatistics semivariogram, taking the NDVI of NOAA/AVHRR spectrum data as the regionalized variable, aiming to provide theory and practical basis for field sampling of crop yield estimation using remote sensing. The ratio of nugget variance and sill of semivariograms are 21.1% and 9.7% in the west and middle regions in Jilin Province respectively. This shows that the crop yields are spatially correlated. The degree and range of correlation are far different in the different situations. In the west test region, the range is 49.9km and the sill is 0.00019. In the middle test region, the range is 16.5km and the sill is 0.00453. The dissimilarity in the western test region is larger than that in the middle one. The range in which the correlation existed of the former is far larger than the later. Different characteristics of spatial correlation of crop yield are decided by the environmental factors. Samples for crop yield estimation should be extracted according to the characteristic of spatial distribution of crop yield to promote the efficiency of sampling. Foundation item: Under the auspices of Major Program of the Chinese Academy of Sciences (KZ95T-03-01). Biography: LI Lin-yi (1968 –), female, a native of Jilin Province, doctor candidate. Her research interest includes land use change monitoring and crop yield estimation using ‘3S’ techniques.