鄱阳湖区粮食供给功能的空间格局分析  

Spatial Pattern of Food Provision Sevice in Poyang Lake Region,China

作　　者：李鹏[1,2] 姜鲁光[1] 封志明[1] 张景华[1,2] 闫慧敏[1] 赵慧霞[3] 

机构地区：[1]中国科学院地理科学与资源研究所,北京100101 [2]中国科学院研究生院,北京100049 [3]国家气象中心,北京100081

出　　处：《自然资源学报》2011年第2期190-200,共11页Journal of Natural Resources

基　　金：国家自然科学基金项目"鄱阳湖区洪水调蓄与粮食供给服务的多情景互竞研究"(40901285);国家重点基础研究发展计划(973计划)"中国主要陆地生态系统服务功能与生态安全"(2009CB421106)

摘　　要：在种植制度复杂、地块破碎及多阴雨天气的南方地区,应用卫星遥感植被指数时间序列数据提取复种指数受到中低空间分辨率的限制。使用较高空间分辨率影像是提取该区域复种信息有效的数据源。论文以鄱阳湖区为研究区,通过遥感解译提取水田空间信息;在界定单/双季稻生长期物候历的前提下,根据水稻不同生长期内归一化植被指数(NDVI)的明显差异,选择合理时间窗口的TM影像获取水田NDVI数据,采用非监督分类法提取单/双季稻的空间分布信息;结合湖区乡镇不同熟制水稻单产数据估算出基于栅格的水稻产量。研究表明,4月下旬到6月下旬是判别双季早稻与单季稻空间分布的合理时间窗口;7月上旬到8月上旬及9月中旬到10上旬是判别单季稻与双季晚稻空间分布的合理时间窗口。2005年,单季稻播种面积为3 081.58 km2,晚稻/早稻播种面积为3 602.97 km2,水稻复种指数为153.9%。单季稻普遍分布在市县建成区周边,双季稻主要分布在河口三角洲等地。全年水稻总产量约1 650×104t,单季稻占30.5%,双季稻占69.5%。赣江下游地区两种熟制水稻产量均较高,而湖汊及湖区外围丘陵地区产量较低。Generally, in southern China with complex growing system, fragmentized and dispersed paddy field, and long-term overcast and rainy weather, the performance of using vegetation in- dex （VI） time-series datasets derived from remote sensing imageries to extract multiple-cropping index was seriously constrained by the lower spatial resolution. Currently, the application of high- er spatial resolution images can be the exclusive and effective way to extract the spatial pattern of different rice cropping systems annually in these regions. In this paper, firstly, the spatial distri- bution of paddy field in Poyang Lake Region （PLR） was obtained through one TM imagery inter- pretation. Secondly, the annual phenological calendar of various systems of paddy rice was de- fined with the agro-meteorological data. According to the significant characteristics that Normalized Difference Vegetation Index （NDVI） fluctuates sharply along with the growth process of paddy rice, map of NDVI for paddy field was derived from another TM image within the applicable time window. Then, different cropping systems of paddy rice were classified by means of Unsuper- vised Classification in Erdas Imagine 9.2. Finally, yield of each raster （100 m） was calculated with unit yield from local statistical department. The results showed that, late April to late June can be the time window to differentiate early rice and single-season rice, while early July till early August and middle September to early October could be the time window for the differentiation between single-season rice and late rice. Specifically, the planting areas of single-season and early/ late rice are 3081.58 km2 and 3602.97 km2 in 2005, respectively, indicating that the nmltiple- cropping index is 153.9%. Single-season rice is generally distributed around the periphery of the built-up area, while double-season rice expanded along the delta. The total yield of paddy rice reached to nearly 16.5 million tons with a proportion of single-season to double-season approx

关 键 词：粮食供给 水稻熟制 时间窗口 NDVI 鄱阳湖区 

分 类 号：F326.11[经济管理—产业经济]