基于GIS技术的福建省烟区烤烟产量估测与分区

Estimation and Division of Flue-cured Tobacco Output in Fujian Tobacco-producing Area Based on GIS Technology

采用烟区代表性样点的烤烟产量和相关影响因素调查资料,在筛选出烤烟K326、云烟87(85)品种产量最佳估测模型的基础上,借助GIS与烤烟产量最佳估测模型集成技术开展福建烟区不同烤烟品种产量的估测与空间分区,揭示福建省烟区耕地不同烤烟品种产量的空间差异。结果表明,人工神经网络模型对K326和云烟87(85)产量的估测精度最高,分别高达99.48%和99.51%,可作为福建烟区烤烟产量预测的最佳模型。福建烟区K326、云烟87(85)高产量区面积分别为96 695.67和208 683.58 hm^2,集中分布于浦城、建阳、建瓯等县市;中产量区面积分别为121 879.70 hm^2和91 803.92 hm^2;低产量区面积分别为296 096.70 hm^2和214 184.60 hm^2,中产区主要分布于永定、宁化和上杭等县(市),表明福建省烟区种植云烟系列烤烟品种可获得更高的产量。

Based on the flue-cured tobacco output and its relevant influence factors data of representative samples in the tobacco area,this study was conducted to reveal spatial variation of flue-cured tobacco output of different flue-cured tobacco varieties in Fujian tobacco production areas. Based on the ideal estimation model of flue-cured tobacco yield for tobacco K326,cloud and mist 85 and 87, and the aid of GIS the yield and spatial distribution of different lue-cured tobacco variaties were estimated.The results showed that the estimation accuracy of the artificial neural network towards tobacco K326,cloud and mist85 and 87 were 99.48% and 99.51%,respectively. So the neural network model was the best model to estimate the flue-cured tobacco production in Fujian province. The high yield regions of tobacco K326,Yunyan 87（85） were concentrated in Pucheng,Jianyang,Jianou,with the area of 96 695.67 hm^2 and 208 683.58 hm^2,respectively;The medium yield regions of tobacco K326,Yunyan 87（85） were mainly distributed in Yongding, Ninghua and Shanghang,with the area of 121 879.70 hm^2 and 91 803.92hm^2,respectively;the low yield regions of tobacco K326,Yunyan 87（85） were 296 096.70 hm^2 and 214 184.60 hm^2,respectively.The research demonstrates that planting mist series in Fujian tobacco-growing areas can lead to higher yields.