莱州湾海岸带土壤光谱分析与有机质反演研究  

Spectral Analysis and Retrieval of Soil Organic Matter in Coastal Zone of Laizhou Bay

作　　者：李勇志[1,2] 唐家奎[1,2] 王德强[1] 王春磊[3] 张成雯[1,2] 

机构地区：[1]中国科学院海岸带环境过程重点实验室/山东省海岸带环境过程重点实验室/中国科学院烟台海岸带研究所,山东烟台264003 [2]中国科学院研究生院,北京100049 [3]河北联合大学,河北唐山063009

出　　处：《地理与地理信息科学》2012年第4期79-82,共4页Geography and Geo-Information Science

基　　金：国家自然科学基金项目(40801124);山东省中青年科学家科研奖励基金(2010BSA06013);中科院创新团队国际合作伙伴计划"海岸带典型环境过程和资源效应";中科院数字地球重点实验室开放基金(2011LDE015);中科院研究生院院长基金

摘　　要：利用土壤光谱反射率预测海岸带典型土壤有机质含量。对莱州湾海岸带典型地区97个土壤样本的光谱反射率特性进行分析,把光谱曲线划分为4个区域,提取每个区域的代表性特征参数,与土壤有机质含量进行相关性分析,最终选用458～587.1nm区间的挠度(相关系数达0.87)作为自变量进行模型回归,并利用均方根误差(RMSE)和预测残差(RPD)进行模型检验与评价。结果表明,以458～587.1nm区间的挠度作为自变量建立的对数函数预测模型具有较高的精度和稳定性,经验证计算出其RMSE为0.39,RPD为2.5,该模型应用效果较好。Organic matter content of typical soil in coastal zone was retrieved by using soil spectral reflectance.Spectral reflectance characteristics of 97 coastal soil samples of Laizhou Bay were analyzed.The spectral curves were divided into four regions.In each region,typical characteristic parameters were picked up to make a correlation with soil organic matter content.Finally,regression analysis model was made with the deflection in the range of 458～587.1 nm as independent variable.The model was evaluated by Root-Mean-Square Error（RMSE） and RPD（the ratio between the stand deviation of the reference method against that of the RMSE）.RMSE and RPD of the logarithm function model were 0.39 and 2.5 respectively,which demonstrated that the regression analysis prediction model can be applied to retrieve coastal organic matter content with good accuracy and stability.

关 键 词：海岸带 土壤有机质 高光谱 挠度 

分 类 号：S153.2[农业科学—土壤学] TP79[农业科学—农业基础科学]