生态用地比例遥感估算及应用分析——以宁夏沿黄城市带为例

Remote Sensing Estimation of Ecological Land Proportion and Its Practical Analysis:Taking Urban Belt Along Yellow River in Ningxia,China for An Example

针对生态用地比例遥感估算缺乏典型土地覆盖类型遥感产品的一致性分析以及缺乏针对宁夏沿黄城市带的实践分析问题,从估算生态用地角度比较分析了四种典型全球土地覆盖类型遥感产品(MODIS-MCD12Q1、GLOBCOVER、GLOBELAND30、FROM-GLC)的一致性,并基于时序性较好的MODIS-MCD12Q1产品,运用Mann-Kendall非参数检验、Sen’s斜率估计以及Hurst指数,分析研究区生态用地比例的演变过程和趋势。结果表明:1)从估算生态用地角度,MODIS-MCD12Q1产品与GLOBELAND30和FROM-GLC产品一致性较高,与GLOBCOVER一致性较差;研究区草地与裸地间的混淆是典型土地覆盖类型产品存在不一致的主要原因。2)2001—2017年,宁夏除金凤区和西夏区生态用地比例呈下降趋势外,其余区县均呈增长或波动增长趋势,研究区总体呈现显著增长趋势,且具有很强持续性;该生态用地比例的增长主要来源于自然植被和耕地类型的增加以及裸地类型的减少。

In view of the lack of consistency analysis of remote sensing estimation on products of typical land cover types and the lack of practical analysis of the urban belt along the Yellow River in Ningxia,this paper takes the urban belt along the Yellow River in Ningxia as the study area,and explores the monitoring and trend analysis of ELP based on typical remotely sensed LCT data.Firstly,evaluate the consistency among four global LCT products,which are MODIS-MCD12Q1,GLOBCOVER,GLOBELAND30,and FROM-GLC,from the perspective of estimating ELP.Secondly,select the longtime serial MODIS-MCD12Q1 to explore the evolution trend of ELP in this area.Mann-Kendall non-parametric test,Sen’s slope estimation and Hurst index methods are employed in the study.The results show that:1)MODIS-MCD12Q1 has a high consistency with GLOBELAND30 and FROM-GLC from the perspective of estimating ELP,but it has a low consistency with GLOBCOVER.The confusion between grasslands and barren is the chief reason of the inconsistencies among various global LCT products in the study area.2)From 2001 to 2017,Jinfeng and Xixia districts of Ningxia have shown a decreasing trend for the ELP,while other areas showing an increasing trend.The whole study area has presented a significant increasing trend with strong sustainability.The increasing ELP mainly comes from the growth of natural vegetation and cultivated land,as well as the decrease of bare land.