CAJ | 학술논문

以盐城湿地主要分布的3个县市:射阳、大丰和东台作为研究区,通过1988年至2006年每隔6 a的4期TM遥感影像,开展盐城湿地类型的动态变化分析,根据遥感影像解译结果,利用基于BP神经元网络的湿地演化CA模型进行湿地类型的变化预测分析.结果表明:研究区从1988年到2006年,耕地、居民用地、养殖场、米草、碱蓬面积总体呈增长趋势,滩地、芦苇和盐田呈急剧减少的趋势.大面积的滩涂开发是造成变化最主要的原因.预测结果为:未来的20 a内居民用地和养殖场将一直保持增加的趋势,米草逐渐占据优势物种的演化趋势.
이염성습지주요분포적3개현시:사양、대봉화동태작위연구구,통과1988년지2006년매격6 a적4기TM요감영상,개전염성습지류형적동태변화분석,근거요감영상해역결과,이용기우BP신경원망락적습지연화CA모형진행습지류형적변화예측분석.결과표명:연구구종1988년도2006년,경지、거민용지、양식장、미초、감봉면적총체정증장추세,탄지、호위화염전정급극감소적추세.대면적적탄도개발시조성변화최주요적원인.예측결과위:미래적20 a내거민용지화양식장장일직보지증가적추세,미초축점점거우세물충적연화추세.
This paper took three main wetland distributing counties in Yancheng: Dongtai, Dafeng and Sheyang as study areas. With the processing of four-period (every six years from 1988 to 2006) remote sensing (RS) images, a dynamic change analysis of the Yancheng wetland types was illuminated. According to the results of the RS images interpretation, the change prediction of the wetland types was analyzed by using cellular automata(CA) model based on BP. The results showed that from 1988 to 2006, the area of cropland, residential land, squaculture farm, Spartina patens, seepweed had a tendency to increase in general, at the same time, the area of mudflat, reed, brine pan tended to decrease violently. The most important reason causing these changes was the current policy of large-scale coast exploitation. The prediction results indicated that the area of residential land and aquaculture farm would maintain their increasing tendencies in the next 20 years and Spartina patens will become the dominant species gradually as its evolution trend.