Abstract:

The purpose of this study is to accurately extract spatial information of lotus root cultivation in the basin area of Sichuan Province and to grasp the hotspot areas for this crop. First of all, we used the remote sensing images of GF-6 and ground survey data as data sources, analyzed spectral characteristics of lotus root fields in order to determine the best period of remote sensing, and extracted spatial information of lotus root cultivation in the area. Then, the spatial distribution characteristics and hotspot areas of lotus root cultivation in Sichuan were analyzed by Moran's I index and Getis-Ord G_i^* index. The results showed that: (1) the total accuracy of extracting spatial information of lotus root based on the maximum likelihood classification method was 92.35%, and the Kappa coefficient was 0.9045; (2) the area of lotus root cultivation in Sichuan basin based on remote sensing was about 128.57 km², which was distributed in 1374 townships, and some areas had aggregation phenomenon; the area of lotus root cultivation in 73 hotspot townships was 54.31 km², accounting for 42.24% of the lotus root cultivation area in the Sichuan basin, with obvious spatial polarization, mainly in Anju of Suining, Lezhi of Ziyang, Jianyang of Chengdu, etc. The results of the study can provide reference for the development of specialized lotus root cultivation in townships and the optimization of lotus root industry in Sichuan Province.

Key words: lotus root, spectral characteristics, Moran index, hot spot analysis, spatial distribution