Supporting data for “Monitoring and understanding fine-scale phenological variability in temperate forests”

For my PhD thesis, I have utilized a range of data from various sources.1). I utilized time-series PlanetScope data with a 3 m spatial resolution from the Planet Labs Inc. to generate phenological maps for six temperate forest sites within the National Ecological Observatory Network (NEON). The derived phenological maps can be used to monitor plant phenology at fine scales.2). I used ground records of plant phenology for six temperate forest sites to validate PlanetScope-derived phenological metrics. These data are publicly available in the data repository of NEON. Detailed phenophase status and intensity are recorded for a number of individual trees, which can be used to derive the timing of key phenological events.3). Maps of seven foliar traits and five structural traits for four NEON temperate forest sites were also used. Foliar traits include leaf mass per area (LMA), carbon content, nitrogen content, area-based chlorophyll concentration, phenolics concentration, starch concentration, and ẟ¹³C. Structural traits include canopy height, plant area index (PAI), height skew, entropy, and rugosity. These traits are derived from remote sensing data, provide detailed characterizations of plant attributes.4). I utilized time-series satellite imagery from harmonized Landsat and Sentinel-2 data to generate maps of interannual phenological variability for four NEON temperate forests sites. The derived maps offer a visualization of the patterns of interannual phenological variability at the site level. By integrating these maps with functional trait data, further investigations can be conducted to explore the underlying biotic drivers of the variability.Overall, the combination of these data provided a comprehensive dataset for my thesis, allowing for a thorough investigation of the patterns of phenological variability and its underlying drivers. <br>

本博士论文研究中，本研究采用了多来源的多组数据集：
1) 采用行星实验室公司（Planet Labs Inc.）提供的空间分辨率为3米的时序PlanetScope数据，为美国国家生态观测站网络（National Ecological Observatory Network, NEON）下辖的6个温带森林样地生成了物候图。所得到的物候图可用于精细尺度下的植物物候监测。
2) 采用6个温带森林样地的植物物候地面观测记录，以验证由PlanetScope数据反演得到的物候指标。此类数据可在NEON的数据仓储库中公开获取。数据集记录了多株单株树木的详细物候期状态与强度，可用于推导关键物候事件的发生时间。
3) 同时采用了NEON下辖4个温带森林样地的7项叶性状与5项结构性状数据集。叶性状包括比叶重（leaf mass per area, LMA）、碳含量、氮含量、面积基准叶绿素浓度、酚类物质浓度、淀粉浓度以及δ¹³C；结构性状包括冠层高度、植物面积指数（plant area index, PAI）、高度偏度、熵以及粗糙度。此类性状均由遥感数据反演得到，可对植物属性进行精细化表征。
4) 采用融合后的Landsat与Sentinel-2时序卫星影像数据，为NEON下辖的4个温带森林样地生成了年际物候变异性分布图。所得到的分布图可实现样地尺度下年际物候变异性格局的可视化。通过将此类分布图与功能性状数据相结合，可开展进一步研究以探索该变异性背后的潜在生物驱动因子。

综上，上述多源数据集的整合为本论文构建了一套完整的研究数据集，可支撑对物候变异性格局及其潜在驱动因子的系统性探究。