引用本文
  •    [点击复制]
  •    [点击复制]
【打印本页】 【下载PDF全文】 查看/发表评论下载PDF阅读器关闭

←前一篇|后一篇→

过刊浏览    高级检索

本文已被:浏览 918次   下载 595 本文二维码信息
码上扫一扫!
基于最大熵模型和ArcGIS的陕西省单季稻种植分布预测及适宜性
高蓓, 高茂盛, 董金芳
0
(陕西省农业遥感信息中心, 西安 710015)
摘要:
依据陕西省4个定点单季稻农气观测站长期观测数据、87个单季稻地理分布数据和14个与单季稻生长有关的环境指标(8个气候指标、4个土壤指标和3个地形指标)数据,运用ArcGIS空间分析模块与最大熵模型(MaxEnt),建立陕西省单季稻种植适宜性分布模型,对单季稻在陕西省的种植分布和适宜性进行预测。研究表明:单季稻最适宜种植区(适宜系数P≥0.69)面积为8 949.24 km2;适宜种植区(0.38≤P<0.69)面积为21 637.68 km2;边缘适宜区(0.07≤P<0.38)面积为73 351.66 km2,不适宜区(P<0.07)面积101 804.22 km2,陕西省单季稻生长区域狭窄,对生态环境要求苛刻,其中适宜种植区域主要集中在陕南的河谷川道、关中盆地的渭河冲积平原区域,适宜区内土层肥沃、水土资源好,土壤pH为弱酸性到中性,地形平坦、海拔高度为230~800 m。ROC曲线的AUC=0.916,表明预测结果可信度高。影响陕西省单季稻种植分布的主要环境指标有≥10 ℃积温(贡献率28.97%)、年降水量(贡献率25.99%)、年平均气温(贡献率16.87%)、海拔(8.27%)、年日照时数(5.66%)和年湿润指数(4.12%),陕西省单季稻最适宜区环境参数为≥10 ℃积温4 865 ℃·d、年降水量762 mm、年平均气温15 ℃、海拔436 m、年日照时数1 573 h和年湿润指数1.6。结果显示基于最大熵模型与ArcGIS构建的种植适宜性分布模型可以准确地对陕西单季稻的种植适宜性做出科学综合评价。
关键词:  单季稻  种植分布  MaxEnt模型  ArcGIS
DOI:10.11841/j.issn.1007-4333.2016.03.02
投稿时间:2015-06-11
基金项目:国家自然科学基金资助项目(31070293); 国家"863"计划(2002AA2121493)
Planting distribution and suitability assessment of single-season rice based on MaxEnt and ArcGIS in Shaanxi Province
GAO Bei, GAO Mao-sheng, DONG Jin-fang
(Remote Sensing Information Center for Agriculture of Shaanxi Privince, Xi'an 710015, China)
Abstract:
By adopting ArcGIS spatial analyst module and MaxEnt, long term observation data of four-single-season rice fixed-point agricultural meteorological stations, 87 occurrence records, 14 environmental factors (8 climatic, 4 soil and 3 topographic factors) were used to predict plant distribution and assess suitability.The results showed that highly suitable, moderately suitable, marginally suitable and unsuitable planting areas for single-season are 8 949.24 km2, 21 637.68 km2, 73 351.66 km2 and 101 804.22 km2 respectively.The marginally planting suitable area and highly planting suitable area are mainly located in valley and the bottom of some gullies of south of Shaanxi, alluvial plain of Guanzhong basin.These areas have good fertile soil, rich climate resources, soil texture ranged from weak acid to neuter.The land is flat and its sea level is from 230 to 800 meters.The AUC of ROC curve was 0.916, indicating that the predictive result with MaxEnt model was highly precise.The main environmental factors determining the suitable planting distribution were accumulate temperature above 10 ℃ (28.97%), annual precipitation (25.99%), annual average temperature (16.87%), altitude elevation (8.27%), annual sunshine duration (5.66%) and annual humid index (4.12%).The environmental variables in the highly suitable planting area were determined as accumulate temperature above 10 ℃ around 4 865 ℃·d, annual precipitation around 762 mm, annual average temperature around 15 ℃, altitude elevation around 436 m, annual sunshine duration around 1 573 h and annual humid index around 1.6.In conclusion, the planting suitability assessment model based on ArcGIS and Maxent could accurately evaluate the planting suitability of Rice, quantify the area of suitable habitat, and analyze the spatial distribution.
Key words:  single-season rice  planting distribution  MaxEnt  ArcGIS