土壤养分影响着土壤的质量,也影响着植被、农作物等的生长。为快速准确地估测艾比湖流域土壤养分状况,选择艾比湖流域精河县作为研究区,以精河县内不同地表覆盖类型土壤为研究对象,基于实地采集的75个土壤样品的室内ASD Field Spec3实...土壤养分影响着土壤的质量,也影响着植被、农作物等的生长。为快速准确地估测艾比湖流域土壤养分状况,选择艾比湖流域精河县作为研究区,以精河县内不同地表覆盖类型土壤为研究对象,基于实地采集的75个土壤样品的室内ASD Field Spec3实测光谱数据和3种光谱变换形式,利用10 nm间隔重采样进行去噪处理,再结合多元逐步回归法(SMLR)、偏最小二乘法回归法(PLSR)、人工神经网络法(ANN)分别建立土壤养分预测模型,以探索最优模型。结果表明:土壤实测光谱的一阶微分、二阶微分变换形式能显著提高光谱与土壤养分之间的相关性,尤其是一阶微分变换与土壤有机质和全氮的相关性最高分别达0.87和0.91,光谱变换技术能显著增强土壤养分与高光谱之间的敏感度,达到更好的建模效果;SMLR、PLSR和ANN这3种模型都具有良好的预测能力,其中,ANN建立的模型预测效果最好,二阶微分变换的ANN模型对有机质、全氮的预测决定系数(R2)分别为0.886和0.984,均方根误差(RMSE)分别为2.614和0.147,PLSR次之;全氮的预测效果明显优于有机质的预测效果,说明高光谱和全氮之间的敏感性更高。总体来说,光谱二阶微分变换形式的人工神经网络模型可以最精确稳定地完成土壤养分含量的快速预测,能够实现艾比湖流域的土壤养分空间分布状况和动态变化特征的动态监测。展开更多
China Marine Surveillance Force was equipped with modern aerial equipments for marine lawexecute with the advantage of functioning agilely at a large scale of surveillance coverage, providing powerful all-round safegu...China Marine Surveillance Force was equipped with modern aerial equipments for marine lawexecute with the advantage of functioning agilely at a large scale of surveillance coverage, providing powerful all-round safeguard, which is of benefit to the harmonious and sustainable development of coastal economy. Onboard the planes, three kinds of remote sensing sensors have been installed, including a marine airborne multi-spectrum scanner （MAMS）, an optical-electronic platform, and an airborne hyper-spectral system AISA＋. The specifications of remote sensing platforms were introduced briefly first, then examples of water quality monitoring by airborne remote sensing were presented, including the monitoring in coastal suspended material, oil-spill and abnormal warm water, etc.展开更多
文摘土壤养分影响着土壤的质量,也影响着植被、农作物等的生长。为快速准确地估测艾比湖流域土壤养分状况,选择艾比湖流域精河县作为研究区,以精河县内不同地表覆盖类型土壤为研究对象,基于实地采集的75个土壤样品的室内ASD Field Spec3实测光谱数据和3种光谱变换形式,利用10 nm间隔重采样进行去噪处理,再结合多元逐步回归法(SMLR)、偏最小二乘法回归法(PLSR)、人工神经网络法(ANN)分别建立土壤养分预测模型,以探索最优模型。结果表明:土壤实测光谱的一阶微分、二阶微分变换形式能显著提高光谱与土壤养分之间的相关性,尤其是一阶微分变换与土壤有机质和全氮的相关性最高分别达0.87和0.91,光谱变换技术能显著增强土壤养分与高光谱之间的敏感度,达到更好的建模效果;SMLR、PLSR和ANN这3种模型都具有良好的预测能力,其中,ANN建立的模型预测效果最好,二阶微分变换的ANN模型对有机质、全氮的预测决定系数(R2)分别为0.886和0.984,均方根误差(RMSE)分别为2.614和0.147,PLSR次之;全氮的预测效果明显优于有机质的预测效果,说明高光谱和全氮之间的敏感性更高。总体来说,光谱二阶微分变换形式的人工神经网络模型可以最精确稳定地完成土壤养分含量的快速预测,能够实现艾比湖流域的土壤养分空间分布状况和动态变化特征的动态监测。
基金supported by the NO. 2007402 Science Foundation of SOAthe scientific research fund NO.JG0719 of the Second Institute of Oceanography, SOA+1 种基金special funds for scientific research on public cause (NO. 200805028)China "908" Project under contract No.908-03-02-08.
文摘China Marine Surveillance Force was equipped with modern aerial equipments for marine lawexecute with the advantage of functioning agilely at a large scale of surveillance coverage, providing powerful all-round safeguard, which is of benefit to the harmonious and sustainable development of coastal economy. Onboard the planes, three kinds of remote sensing sensors have been installed, including a marine airborne multi-spectrum scanner （MAMS）, an optical-electronic platform, and an airborne hyper-spectral system AISA＋. The specifications of remote sensing platforms were introduced briefly first, then examples of water quality monitoring by airborne remote sensing were presented, including the monitoring in coastal suspended material, oil-spill and abnormal warm water, etc.
