Canopy structure regulates autumn phenology by mediating microclimate in temperate forests

Data Descriptions
Here, we provide the data and code used for the paper “Canopy structure regulates autumn phenology by mediating microclimate in temperate forests” by Wu et al.
Code and Data folders includes all the code and data for analysis in the main text, organized by:
Fig1_DataAnalysis.R and Fig1.ipynb contain code for analyzing the spatial variations of autumn phenology metrics and relationships between canopy structure metrics and autumn phenology. Related datasets include Data1_Cano_Micro_Phe.csv, Data1_DataSlope.csv, Data1_DataYear.csv, Data1_Mean_Phenology.csv, and Data1_variance_explained.csv.
Fig2_DataAnalysis.R and Fig2.ipynb contain code for analyzing the relationships between canopy structure and microclimate factors and relative contributions of microclimate factors to autumn phenology metrics. Related datasets include Data2_all_PAI_GLI.csv, Data2_all_pai_Tbuffer.csv, Data2_DataOneYear.csv, and Data2_Var_importance.csv.
Fig3_DataAnalysis.R and Fig3.ipynb contain code for analyzing the relationships between microclimate factors and autumn phenology metrics. Related datasets include Data3_DataSlope.csv and Data3_DataYear.csv.
Fig5_DataAnalysis.R and Fig5.ipynb contain code for comparing the accuracy of the phenology model and projecting future autumn phenology under the shared socio-economic pathway (SSP) 1-2.6, SSP3-7.0, and SSP5-8.5 scenarios. Related datasets include Data5_all_Preds.csv, Data5_AllData_Micro_Macro_Cano.csv, Data5_AllFutureFallDOY.csv, Data5_DelayDOY.csv, Data5_FutureFallDOY_roll.csv, Data5_Model_Pvalue.csv, Data5_ModelAccuracy.csv, and Data5_Ttest_model_accuracy.csv.
SupplementaryAnalysis.R and SupplementaryFigs.ipynb contain code for analyzing
1. Spatial variations of autumn phenology metrics in the Changbai Mountain (CBS) site. Related datasets include SupplementaryData.pickle and ED1_CBS_all_data_sp_scale.csv.
2. Relationships between canopy structure, microclimate and autumn phenology metrics within the CBS site. Related datasets include SupplementaryData.pickle, ED2_CBS_Canopy.csv, ED2_CBS_Microclimate.csv, and ED2_CBSData.csv
3. Spatial variations of autumn phenology metrics categorized by tree species within the five NEON study sites. Related datasets include SupplementaryData.pickle and Datas2_DataSpecies.csv.
4. Distributions of canopy structure and simulated microclimate metrics within the five NEON study sites. Related datasets include SupplementaryData.pickle and Datas3_DataYear.csv
5. Accuracy assessment of simulated microclimate factors. Related datasets include SupplementaryData.pickle, Datas4_allData_Meas_Simu.csv, Datas4_GLI_accuracy.csv and Datas4_Tbuffer_accuracy.csv.
6. Partial dependence plots demonstrating the response of autumn phenology metrics to microclimate factors. Related datasets include SupplementaryData.pickle, Datas5_PDP_DataYear.csv, Datas5_partial_dof.csv, Datas5_partial_sof.csv.
7. Pathways linking canopy structure and autumn phenology through mediating microclimate conditions within the five NEON study sites. Related datasets include Datas6_SEMData.csv.
8. Relationships between autumn phenology metrics and microclimate factors simulated at different heights within the five NEON study. Related datasets include SupplementaryData.pickle, Datas7_AllData_heights.csv and Datas7_DiffHeights.csv.
9. Correlation between start of autumn (SOA) and duration of autumn (DOA) in 2019 within the five NEON study sites. Related datasets include SupplementaryData.pickle and Datas8_SOFDOF_DataYear.csv.
10. Relationships between microclimate factors and end of autumn (EOA) within the five NEON study sites in 2019. Related datasets include SupplementaryData.pickle, Datas9_MiroEOF_DataYear.csv and Datas9_Micro_Lmm.csv.
11. Temporal changes of microclimate conditions with the progression of autumn phenological transition. Related datasets include SupplementaryData.pickle and Datas10_all_micro_time.csv.
12. Selection of areas with similar elevations for all statistical analyses in the current study. Related datasets include SupplementaryData.pickle and Datas12_DTMData.csv.
Maps folder includes all autumn phenology data (SOA, DOA, and EOA) from 2018 to 2022 within the five National Ecological Observatory Network (NEON) study sites and the CBS site for 2020.
Note: SOF (start of fall), DOF (duration of fall), and EOF (end of fall) used in the code share the same meaning of SOA, DOA, and EOA.

数据集说明
本文提供了Wu等发表于论文《温带森林冠层结构通过调控微气候影响秋季物候》的相关数据与代码。
代码与数据集文件夹包含正文中所有分析所用的代码与数据，按以下结构组织：
Fig1_DataAnalysis.R 与 Fig1.ipynb 包含用于分析秋季物候指标空间变异、冠层结构指标与秋季物候间关联的代码，相关数据集包括 Data1_Cano_Micro_Phe.csv、Data1_DataSlope.csv、Data1_DataYear.csv、Data1_Mean_Phenology.csv 及 Data1_variance_explained.csv。
Fig2_DataAnalysis.R 与 Fig2.ipynb 包含用于分析冠层结构与微气候因子间关联、微气候因子对秋季物候指标相对贡献的代码，相关数据集包括 Data2_all_PAI_GLI.csv、Data2_all_pai_Tbuffer.csv、Data2_DataOneYear.csv 及 Data2_Var_importance.csv。
Fig3_DataAnalysis.R 与 Fig3.ipynb 包含用于分析微气候因子与秋季物候指标间关联的代码，相关数据集包括 Data3_DataSlope.csv 与 Data3_DataYear.csv。
Fig5_DataAnalysis.R 与 Fig5.ipynb 包含用于对比物候模型精度、在共享社会经济路径（Shared Socio-economic Pathway，SSP）1-2.6、SSP3-7.0及SSP5-8.5情景下预测未来秋季物候的代码，相关数据集包括 Data5_all_Preds.csv、Data5_AllData_Micro_Macro_Cano.csv、Data5_AllFutureFallDOY.csv、Data5_DelayDOY.csv、Data5_FutureFallDOY_roll.csv、Data5_Model_Pvalue.csv、Data5_ModelAccuracy.csv 及 Data5_Ttest_model_accuracy.csv。
SupplementaryAnalysis.R 与 SupplementaryFigs.ipynb 包含用于以下分析的代码：
1. 长白山（Changbai Mountain，CBS）样地秋季物候指标的空间变异，相关数据集包括 SupplementaryData.pickle 与 ED1_CBS_all_data_sp_scale.csv。
2. 长白山样地内冠层结构、微气候与秋季物候指标间的关联，相关数据集包括 SupplementaryData.pickle、ED2_CBS_Canopy.csv、ED2_CBS_Microclimate.csv 及 ED2_CBSData.csv。
3. 5个国家生态观测站网络（National Ecological Observatory Network，NEON）研究样地内按树种分类的秋季物候指标空间变异，相关数据集包括 SupplementaryData.pickle 与 Datas2_DataSpecies.csv。
4. 5个NEON研究样地内冠层结构与模拟微气候指标的分布，相关数据集包括 SupplementaryData.pickle 与 Datas3_DataYear.csv。
5. 模拟微气候因子的精度评估，相关数据集包括 SupplementaryData.pickle、Datas4_allData_Meas_Simu.csv、Datas4_GLI_accuracy.csv 及 Datas4_Tbuffer_accuracy.csv。
6. 展示秋季物候指标对微气候因子响应的偏依赖图，相关数据集包括 SupplementaryData.pickle、Datas5_PDP_DataYear.csv、Datas5_partial_dof.csv、Datas5_partial_sof.csv。
7. 5个NEON研究样地内通过调控微气候条件连接冠层结构与秋季物候的路径，相关数据集包括 Datas6_SEMData.csv。
8. 5个NEON研究样地内不同高度模拟的微气候因子与秋季物候指标间的关联，相关数据集包括 SupplementaryData.pickle、Datas7_AllData_heights.csv 及 Datas7_DiffHeights.csv。
9. 2019年5个NEON研究样地内秋季始期（SOA）与秋季持续时长（DOA）的相关性，相关数据集包括 SupplementaryData.pickle 与 Datas8_SOFDOF_DataYear.csv。
10. 2019年5个NEON研究样地内微气候因子与秋季末期（EOA）间的关联，相关数据集包括 SupplementaryData.pickle、Datas9_MiroEOF_DataYear.csv 及 Datas9_Micro_Lmm.csv。
11. 随秋季物候转变进程的微气候条件时间变化，相关数据集包括 SupplementaryData.pickle 与 Datas10_all_micro_time.csv。
12. 本研究所有统计分析所用的相似海拔区域选择，相关数据集包括 SupplementaryData.pickle 与 Datas12_DTMData.csv。
Maps文件夹包含5个NEON研究样地2018-2022年的全部秋季物候数据（SOA、DOA及EOA），以及2020年CBS样地的相关数据。
注：代码中使用的SOF（start of fall，秋季始期）、DOF（duration of fall，秋季持续时长）与EOF（end of fall，秋季末期），与SOA、DOA及EOA的含义完全一致。