基于CA_Markov模型的伊犁河三角洲景观格局预测

在遥感和水文数据的支持下,结合伊犁河三角洲的区域特点,确定了伊犁河三角洲的景观分类系统.通过景观生态学的定量分析和CA_Markov模型模拟方法,系统分析了1975—2007年伊犁河三角洲景观格局的时空演变特征,并对2020年景观格局的情景进行了模拟与预测.结果表明:在人为间接干扰和自然环境的双重影响下,1975—2007年伊犁河三角洲的景观格局变化具有一定的周期性和不稳定性;对CA_Markov模型的可利用性进行分析与检验,1990—2007年伊犁河三角洲景观变化符合CA_Markov模型应用的条件;2020年,研究区景观的斑块破碎化严重,连通性趋于弱化,与水密切相关的湖泊、沼泽和漫滩的时空变化存在很大的不确定性;分别基于1990—2000年和2000—2007年概率转移矩阵预测的2020年研究区景观格局存在显著的差异性,不同景观类型的差异程度和表现形式各异,原因在于这两个时段自然条件的非一致性和不稳定性以及各景观类型的转移方式和方向不同.

Prediction of landscape patterns in Ili River Delta based on CA_Markov model

Based on the remote sensing and hydrologic data and in combining with the characteristics of arid area, a landscape classification system for Ili River Delta was established. In the meantime, by the methods of quantitative analysis in landscape ecology and the simulation method with CA_Markov model, the landscape change pattern in the Delta from 1975 to 2007 and the scenarios simulation landscape pattern in 2020 were analyzed. Under the effects of anthropogenic activities and environmental factors, the landscape change pattern in 1975-2007 presented definite periodicity and instability. In 1990-2007, the landscape change accorded with the applied conditions of CA_Markov model; in 2020, the predicated landscape pattern showed severe patch fragmentation, the landscape connectivity tended to be weakened, and the spatiotemporal change patterns of the landscapes closely related to water, such as lake, marsh, and floodplain, were of great uncertainty. There was a significant differentiation between the predicated landscape patterns in 2020 based on the 1990-2000 and 2000-2007 probabilistic transfer matrixes. The variation degrees and expression forms of different landscapes also differed, due to the discrepancy and instability of natural conditions, as well as the different change ways and directions of landscape types in 1900-2000 and 2000-2007.