NEON 植物叶面化学观测数据（冠层氮浓度）

简介使用 NEON 植物叶面化学观测数据校准并使用 L1 地表方向反射率源自 NEON 成像光谱仪 (NIS)预测的模拟冠层氮浓度。波段包括1) 冠层氮百分比2) 冠层氮模型不确定性3) 针叶与非针叶模型的分类结果4) 基于 NDVI 阈值的有效像素掩码。冠层氮数据产品可提供每个像素的冠层氮百分比估计值。模型结果是通过训练随机森林模型生成的该模型利用定向反射光谱作为预测变量并利用实地采集的冠层叶片化学测量数据进行模型校准和验证。由于针叶植物和非针叶植物的最终模型形参存在显著差异因此同时使用针叶植物和非针叶植物的氮模型。模型使用 NEON 观测站历史上收集的样本进行训练并针对数据质量问题进行了适当的过滤。这些值仅基于植被样本进行训练因此模型在其他景观类型中可能不准确。虽然提供了所有像素但我们得出了 NDVI 阈值掩码以隔离包含植被的像素。GEE 中的可用性可能并不代表 NEON 数据门户链接如下中的完整可用性。如需向 GEE 添加其他地点和年份请发送电子邮件至 listaopgeebattelleecology.org 提出申请。数据集说明空间信息目录所有者美国国家生态观测网络像素大小1 米所有波段数据集可用时间2013-01-01T00:00:00Z–2021-07-18T16:36:45Z数据集生产者NEONEarth Engine 代码段ee.ImageCollection(projects/neon-prod-earthengine/assets/CNC/002)变量波段表格NameDescriptionMinMaxResolutionUnitsNitrogen_PercentCanopy Nitrogen Percent051 meters%Nitrogen_UncertaintyUncertainty associated with the %N predictions; calculated by taking the standard deviation of the %N predictions from each decision tree in the random forest model.051 meters%Needle_Non-needle_ClassificationA binary map created using Support Vector Machine (SVM) classification. The non-needle class includes all vegetation types that are not needle leaf, such as broadleaf trees, shrubs, herbaceous cover, and so forth. Separate random forest regression models were developed to predict foliar nitrogen values for needle and non-needle vegetation types.1 metersValid_Pixel_ClassificationA binary map based on NDVI threshold of 0.2, where pixels with NDVI 0.2 are set to 0 (invalid). Intended to enable exclusion of non-vegetated areas, such as roads, water bodies, built-up areas, and so forth.1 metersClass Table: Needle_Non-needle_ClassificationValueColorColor ValueDescription0#ffffffneedle leaf1#a9a9a9non-needle leafClass Table: Valid_Pixel_ClassificationValueColorColor ValueDescription0#ffffffinvalid1#a9a9a9valid影像属性NameTypeDescriptionAOP_VISIT_NUMBERIntUnique visit number to the NEON site.CITATIONStringData citation. See NEON Data Policies and Citation Guidelines.DOIStringDigital Object Identifier. NEON data that have been released are assigned a DOI.FLIGHT_YEARIntYear the data were collected.NEON_DOMAINStringNEON eco-climatic domain code, D01 to D20. See NEON Field Sites and Domains.NEON_SITEStringNEON four-digit site code. See NEON Field Sites.NEON_SITE_NAMEStringFull name of the NEON site. See NEON Field Sites.NEON_DATA_PROD_URLStringNEON data product url. Always set to: https://data.neonscience.org/data-products/DP3.30018.002.SENSOR_IDStringID of NEON Imaging Spectrometer (NIS), or Global Airborne Observatory (GAO): NIS1, NIS2, NIS3, GAO.PROVISIONAL_RELEASEDStringWhether the data are Provisional or Released. See https://www.neonscience.org/data-samples/data-management/data-revisions-releases.RELEASE_YEARIntIf data are released, the year of the NEON Release Tag.代码// Read in the NEON AOP CNC Image Collection var cnc ee.ImageCollection( projects/neon-prod-earthengine/assets/CNC/002); // Display available images in the CNC Image Collection print(NEON Canopy Nitrogen Images) print(cnc.aggregate_array(system:index)) // Filter by specific image index var cnc_harv2019 cnc.filter(ee.Filter.eq(system:index, 2019_HARV_6)).first(); // Mask nitrogen band using valid pixel band var cnc_valid cnc_harv2019.select(Valid_Pixel_Classification); var cnc_masked cnc_harv2019.updateMask(cnc_valid); // Read in the uncertainty band and mask using valid pixel band var cnc_unc cnc_harv2019.select(Nitrogen_Uncertainty); var cnc_unc_masked cnc_unc.updateMask(cnc_valid); // Define the visualization parameters // The min/max values below use a 2 sigma (~95%) stretch var visParamsNit {bands: [Nitrogen_Percent], min: 1.6, max: 2.14, palette: [440154, 3b528b, 21908c, 5dc963, fde725]}; var visParamsUnc {min: 0.4, max: 0.7, palette: [0d0887, 7e03a8, cc4778, f89540, f0f921]}; // Add the masked nitrogen and uncertainty layers to the map and center on the site Map.addLayer(cnc_unc_masked, visParamsUnc, HARV 2019 Nitrogen Uncertainty (%)); Map.addLayer(cnc_masked, visParamsNit, HARV 2019 Nitrogen (%)); Map.centerObject(cnc_harv2019, 11);代码链接https://code.earthengine.google.com/b000c906ba224b63fe5d4462da7b5741结果引用许可CC0-1.0除与稀有、受威胁或濒危RTE物种相关的数据外NEON收集并作为数据产品提供的所有数据均以Creative Commons CC0 1.0“无保留权利”发布到公共领域。NEON数据未应用任何版权任何人均可出于商业或非商业目的复制、修改或分发数据无需征求许可。NEON数据可能仍受其他法律或权利例如隐私权的约束NEON不对数据做出任何保证并否认所有责任。使用或引用NEON数据时不应暗示NEON对此表示认可。在大多数国家/地区数据和事实不受版权保护。通过将NEON数据放入公共领域我们鼓励广泛使用尤其是在科学分析和数据聚合中。但是请注意以下学术规范NEON数据的使用方式应考虑到数据的局限性并以与数据包相关的文档为指导。有关如何正确使用和引用NEON数据以及发布使用NEON数据的研究的最佳实践的详细信息请参阅NEON数据指南和政策。机器学习https://www.cbedai.net/xg