基于星地遥感耦合的南疆枣树氮素监测机理与模型研究

Using the remote sensing technology to obtain the data of jujube tree quickly and accurately to guide for the application of nitrogen (N) fertilizer efficiently, this has played an important role in arising the utilization rate of N fertilizer and sustainable development of red date industry. As an decisive factor of cash crop for regional economic development, this work mainly focus on the red date in southern Xinjiang where jujube tree cultivated widely, and the content of N in jujube tree. Base on the field experiments, collecting data of canopy spectra and N content over different period of duration by ground sensor, this work aims to access their response relationships and identify the specific wavelengths. Variety of spectral index models developed using simulated multi-band remote sensing data from proximity sensor, in order to choose the optimal model for decision-making. The remote sensing multispectral data from synchronization satellite calibrated by canopy spectra for reference obtained from ground sensors, in order that developing the best model to predict their N contents in large scale. And then using data of field survey for model validation and optimization; thereby, building a remote sensing model for monitoring jujube tree cultivated widely in southern Xinjiang, and accessing the N contents quickly and reliably. Providing reliable basis quantitatively could be benefit for the rationale application of N fertilizer in southern Xinjiang where jujube tree cultivated widely, and would be expected to ease the pressure of environmental degradation causing by the application of N fertilizer blindly.

利用遥感技术快速准确获取枣树氮素数据并用于指导氮肥的科学合理施用，对提高氮肥利用率及红枣产业的可持续发展具有重要意义。项目以南疆大面积种植且对区域经济发展具有决定性作用的经济作物红枣为研究对象，以枣树氮素为研究因子。通过大田氮肥试验及地面传感器获取枣树不同生育期的冠层光谱数据和氮素含量数据，揭示枣树冠层光谱对氮素的响应机制，明确枣树氮素的敏感波段。利用近地传感器数据模拟的不同卫星传感器的多光谱数据构建多种光谱指数模型，优选最佳模型。以近地面传感器获取的枣树冠层高光谱数据为基准，对同步的卫星遥感多光谱数据进行定标，将定标后的多光谱数据应用于优选出的最佳模型，进行大面积枣树氮素的定量反演，并利用地面调查数据对模型进行验证与优化，建立一种能够快速获取南疆大面积枣树氮素含量数据的遥感监测模型。研究结果可为南疆大面积枣树氮肥的科学合理施用提供可靠的数字依据，有望缓解氮肥盲目施用造成的环境恶化问题。

氮素对作物生长发育和农产品品质与产量具有关键作用，利用遥感技术快速监测作物冠层氮含量数据，对农田氮素管理具有重要作用。本项目以南疆大面积种植且对地方经济有重要贡献的枣树为研究对象，通过3年枣树大田氮肥试验与不同生育期的9次采样数据及2次研究区的采样数据，采用地面高光谱数据和卫星遥感数据耦合的方法，研究了枣树冠层氮素遥感监测的机理并构建了模型。本研究发现，枣树冠层在不同生育期具有不同的高光谱特征，其氮素特征波段随生育期变化而变化，应针对不同生育期分别构建监测模型方能取得满意的监测效果；利用地面高光谱数据对卫星遥感数据进行校正可有效提高模型的监测精度；多植被指数模型较单一植被指数模型具有更高的监测精度；RGRI、VARI和BAI等植被指数与枣树冠层氮含量相关性较好；线性模型和非线性模型的预测能力没有本质区别，均可有效监测枣树冠层氮含量；遥感影像的空间分辨率对模型监测能力有明显影响作用，提高空间分辨率可明显提高模型的监测精度；利用多源卫星遥感数据可有效提高监测的时间分辨率，解决了阴、雨天气干扰卫星遥感监测的问题。研究结果为南疆枣园氮素的宏观监测提供了科学有效的方法，可为枣园氮肥科学管理提供高精度的数据支持。