Data for: The pace of global river meandering influenced by fluvial sediment supply

# Data for "The Pace of Global River Meandering Set by Fluvial Sediment Supply" --- This dataset includes all underlying data used in the manuscript. This includes surface water mask files (.tif), derived channel centerline files (.csv, .pkl), bar-averaged migration, and aggregated tabular data. To work with any of the derived data, we recommend using a Python-based workflow. ## Description of the data and file structure The contents of the dataset is organized into 3 directories: ### 1) Dammed\_Rivers: This includes the underlying data for the upstream-to-downstream comparisons of river mobility across dams. It includes files for the Flint, Iowa, and Red Rivers. Files are organized by: ├── Dammed_Rivers/   ├── FlintRiver/    ├── compare.py    ├── FlintRiver_WBMdata.csv    ├── gpkg_shapes/    ├── masks/     ├── FlintDownstream/      ├── bar_migration/      ├── centerline/      ├── centerline_csv/      ├── mask/     ├── FlintUpstream/    ├── WBM_columns.txt   ├── IowaRiver/    ├── compare.py    ├── IowaRiver_WBMdata.csv    ├── gpkg_shapes/    ├── masks/     ├── IowaDownstream/      ├── bar_migration/      ├── centerline/      ├── centerline_csv/      ├── mask/     ├── IowaUpstream/    ├── WBM_columns.txt   ├── RedRiver/    ├── codes/    ├── data/     ├── 1995/      ├── mask/      ├── width/     ├── 2015/     ├── migration/      ├── bar/      ├── combine.py      ├── DownstreamMigration.csv      ├── points/      ├── UpstreamMigration.csv The structures of the Iowa and Flint River directories are roughly the same: * **compare.py**: A Python script that provides the analysis for the two rivers comparing the upstream and downstream (of the dam) portions of the reaches. * The derived **WBMsed data** along the river path. e.g. FlintRiver_WBMdata.csv. I've included a description of columns as an additional text file (WBM_columns.txt). * **gpkg_shapes/**: A directory that holds .gpkg files of Polygon shape files that cover the analyzed reaches. * **masks/**: Holds all of the geospatial and derived migration data. * **bar_migration/**: Holds the aggregated bar-scale migration data. The naming convention includes the year1 and year2 over which the migration is measured. e.g. 1990_2021 indicates the migration comparing the 1990 and 2021 year centerlines. The column descriptions are given in the directory (bar_migration_csv_column_desc.txt). * **centerline/**: Holds .pkl objects of the centerlines derived from the channel masks. The method to open and work with these pickle files is provided in the github repository: 10.5281/zenodo.8341894. * **centerline_csv/**: Holds .csv files for the channel masks. The file naming convention includes the channel mask year the centerline is derived from. e.g. FlintDownstream_1990_centerline.csv is the centerline from 1990. The column descriptions are given as a separate file in the directory (centerline_csv_column_desc.txt). * **mask/**: Contains the raster data for the channel masks used to generate the centerlines. These are provided as single band binary .tif files. The structure for the Red River directory is slightly different because this analysis was completed earlier than the other two rivers. Descriptions follow: * **codes/:** Holds a number of codes used to merge all derived centerline files, calculate the migration rates, and compare upstream and downstream portions of the reach. * **combine_widths.py**: Combines all the width csv files into a single dataframe. * **compare.py**: Statistically compares the upstream and downstream portions of the reach. * **get_migration.py**: Calculates the migration rates from the width dataframes. * **get_sinuosity.py**: Calculates the sinuosity from the width dataframes. * **Data/**: Holds all the used data for this analysis. * 1995 and 2015 are the two years compared to get the migration rate. * **mask/**: Holds all of the .tif raster files for channel water. The entire measured reach is broken down into 65 segments. * **width/**: Holds the centerline .csv files. The column descriptions are given in a separate file (red_river_width_column_desc.csv). * **migration/**: contains the migration data comparing the two timesteps. * **bar/**: Bar aggregated migration distances for each of the 65 segments. Column descriptions are given in separate file (bar_column_desc.txt). * **points/**: Point comparisons pinned to the 1995 centerline showing the migration distances. Column descriptions are given in separate file (point_column_desc.txt). * **combine.py**: Python script combining the 65 segment data into single data tables. * **DownstreamMigration.csv**: Bar-scale migration data downstream of Lake Texoma. Column descriptions are found in a separate file (migration_csv_column_desc.txt). * **UpstreamMigration.csv**: Bar-scale migration data upstream of Lake Texoma. Column descriptions are found in a separate file (migration_csv_column_desc.txt). * **RedRiver_WBMdata.csv**: Contains the WBMsed data for the Red River portion. Column descriptions are given as a separate file (red_river_wbm_column_desc.txt). ### 2) Single\_Rivers: This includes the underlying data for the individual rivers for which I estimated my own migration rates. The file structure is the same for each river. I give one example below, which follows: ├── Single_Rivers/   ├── Algeria/    ├── Algeria/     ├── bar_migration/     ├── centerline/     ├── centerline_csv/     ├── mask/    ├── Algeria.gpkg   ├── ...   ├── Column_Desc/    ├── bar_migration_column_desc.txt    ├── centerline_csv_column_desc.txt   ├── ... Descriptions of what each of the subfolders contains: * **bar_migration/**: .csv file containing bar-scale migration rates for the compared timesteps. The file naming convention contains the compared years. e.g. Algeria_1991_2021_bar_migration.csv is the migration data between 1991 and 2021. The column descriptions are provided in a separate file (Column_Desc/bar_migration_column_desc.txt). * **centerline/**: .pkl objects containing the centerline data. This data format is used by the software I use to generate the centerline data. You can find more information on this in the github repository: **TODO** * **centerline_csv/**: .csv files for the centerlines generated from the channel water masks. Column descriptions are given in a separate file (Column_Desc/centerline_csv_column_desc.txt). * **mask/**: Binary raster .tif files containing channel water. I used these to generate centerlines. ### 3) Tabular\_Data This contains all of the aggregated tabular data used in the analysis. I have here the collected primary data, collated published data, and averaged WBMsed data. Note that N/A values populate empty cells. These are missing values that are not available in the published sources or not present in the WBMsed model. ├── Single_Rivers/   ├── Column_Desc/   ├── combine_data.py   ├── FullCombinedAvgData_050423.csv   ├── FullCombinedData_050423.csv   ├── FullWBM_data.csv   ├── Primary_Data_050423.csv   ├── Published_Data/    ├── bend_data/     ├── "river".csv    ├── citations.txt    ├── Column_Desc    ├── Published_Data_050423.csv    ├── PublishedBendData_050423.csv For .csv files, column descriptions are given as separate files in the Column_Desc/ directory following the pattern of "*_file_name_column_desc.txt*." There is overlap between column names. I've included enough to understand all columns in the files provided. * **combine_data.py**: A Python script used to aggregate the individual bend-scale river migrationi data. * **FullCombinedAvgData_050423.csv**: All reach-averaged data for the rivers within the dataset. * **FullCombinedData_050423.csv**: All bend-scale data for the rivers within the dataset. * **FullWBM_data.csv**: All WBMsed data for the measured rivers. * **Primary_Data_050423.csv**: Just the primary data. * **bend_data/**: Contains the published bend-scale data for each river it exists for. Note, the meander wavelength field was measured by me for this study. * **citations.txt**: Sources used for published migration rates. * **Published_Data_050423.csv**: Aggregated reach-average published data. * **PublishedBendData_050423.csv**: Aggregated bend-scale published data. ## Sharing/Access information We leverage Google Earth Engine (GEE) Landast data for the natural data. Links to the relevant datasets are: [Landsat catalog on GEE](https://developers.google.com/earth-engine/datasets/catalog/landsat)

# 论文《由河流沉积物供给 (fluvial sediment supply) 决定的全球河流蜿蜒 (meandering) 速率》配套数据集 --- 本数据集包含论文手稿中使用的全部基础数据，涵盖地表水掩膜 (surface water mask) 文件（.tif）、衍生河道中心线 (channel centerline) 文件（.csv、.pkl）、沙坝 (bar) 平均迁移数据以及聚合表格数据 (aggregated tabular data)。建议使用基于Python的工作流 (Python-based workflow) 处理所有衍生数据。

## 数据与文件结构说明
本数据集内容分为3个目录：

### 1) 筑坝河流 (Dammed_Rivers)
该目录包含跨大坝的河流流动性上下游对比分析所需的基础数据，涉及弗林特河（Flint River）、艾奥瓦河（Iowa River）与雷德河（Red River）。文件组织方式如下：
├── Dammed_Rivers/
├── FlintRiver/
├── compare.py
├── FlintRiver_WBMdata.csv
├── gpkg_shapes/
├── masks/
├── FlintDownstream/
├── bar_migration/
├── centerline/
├── centerline_csv/
├── mask/
├── FlintUpstream/
├── WBM_columns.txt
├── IowaRiver/
├── compare.py
├── IowaRiver_WBMdata.csv
├── gpkg_shapes/
├── masks/
├── IowaDownstream/
├── bar_migration/
├── centerline/
├── centerline_csv/
├── mask/
├── IowaUpstream/
├── WBM_columns.txt
├── RedRiver/
├── codes/
├── data/
├── 1995/
├── mask/
├── width/
├── 2015/
├── migration/
├── bar/
├── combine.py
├── DownstreamMigration.csv
├── points/
├── UpstreamMigration.csv

艾奥瓦河与弗林特河目录的结构大致相同：
* **compare.py**：用于对比研究河段大坝上下游区域的Python分析脚本。
* 沿河道分布的衍生WBMsed数据 (WBMsed data)，例如FlintRiver_WBMdata.csv。本数据集附带列说明文本文件WBM_columns.txt。
* **gpkg_shapes/**：存储覆盖研究河段的多边形矢量文件（.gpkg）的目录。
* **masks/**：存储所有地理空间数据与衍生迁移数据的目录。
* **bar_migration/**：存储聚合后的沙坝尺度迁移数据。文件名包含迁移测量的两个年份，例如1990_2021代表对比1990年与2021年中心线的迁移数据。该目录下附带列说明文件bar_migration_csv_column_desc.txt。
* **centerline/**：存储由河道掩膜生成的中心线.pkl对象文件。打开与处理这些Pickle文件的方法可在GitHub仓库10.5281/zenodo.8341894中获取。
* **centerline_csv/**：存储河道掩膜对应的.csv格式中心线文件。文件名包含生成中心线所用的河道掩膜年份，例如FlintDownstream_1990_centerline.csv代表1990年生成的中心线文件。该目录下附带列说明文件centerline_csv_column_desc.txt。
* **mask/**：包含用于生成中心线的河道掩膜栅格数据，以单波段二进制.tif格式提供。

雷德河目录的结构略有不同，因为该分析早于另外两条河流，具体说明如下：
* **codes/**：存储用于合并所有衍生中心线文件、计算迁移速率以及对比研究河段上下游区域的代码文件。
* **combine_widths.py**：将所有宽度.csv文件合并为单个数据帧的Python脚本。
* **compare.py**：用于统计对比研究河段上下游区域的脚本。
* **get_migration.py**：从宽度数据帧中计算迁移速率的脚本。
* **get_sinuosity.py**：从宽度数据帧中计算蜿蜒度 (sinuosity) 的脚本。
* **Data/**：存储本次分析所用的全部数据。
* 1995与2015为用于计算迁移速率的两个对比年份。
* **mask/**：存储所有河道水体.tif栅格文件。整个研究河段被划分为65个分段。
* **width/**：存储中心线.csv文件。该目录下附带列说明文件red_river_width_column_desc.csv。
* **migration/**：包含两个时间步长的迁移对比数据。
* **bar/**：65个分段各自的沙坝聚合迁移距离数据。该目录下附带列说明文件bar_column_desc.txt。
* **points/**：锚定1995年中心线的点位对比数据，展示迁移距离。该目录下附带列说明文件point_column_desc.txt。
* **combine.py**：将65个分段的数据合并为单个数据表的Python脚本。
* **DownstreamMigration.csv**：特克索马湖 (Lake Texoma) 下游的沙坝尺度迁移数据。列说明可在配套文件migration_csv_column_desc.txt中查看。
* **UpstreamMigration.csv**：特克索马湖 (Lake Texoma) 上游的沙坝尺度迁移数据。列说明可在配套文件migration_csv_column_desc.txt中查看。
* **RedRiver_WBMdata.csv**：雷德河段的WBMsed数据。该目录下附带列说明文件red_river_wbm_column_desc.txt。

### 2) 单一河流 (Single_Rivers)
该目录包含我自行估算迁移速率的单条河流的基础数据，每条河流的文件结构一致。以下以阿尔及利亚河段为例：
├── Single_Rivers/
├── Algeria/
├── Algeria/
├── bar_migration/
├── centerline/
├── centerline_csv/
├── mask/
├── Algeria.gpkg
├── ...
├── Column_Desc/
├── bar_migration_column_desc.txt
├── centerline_csv_column_desc.txt
├── ...

各子文件夹的内容说明：
* **bar_migration/**：存储对比时间步长下沙坝尺度迁移速率的.csv文件。文件名包含对比年份，例如Algeria_1991_2021_bar_migration.csv代表1991年至2021年的迁移数据。列说明可在配套文件Column_Desc/bar_migration_column_desc.txt中查看。
* **centerline/**：存储中心线数据的.pkl对象文件，该格式为我生成中心线数据所用软件的标准格式。更多信息可在GitHub仓库**TODO**中查询。
* **centerline_csv/**：存储由河道水体掩膜生成的中心线.csv文件。列说明可在配套文件Column_Desc/centerline_csv_column_desc.txt中查看。
* **mask/**：存储河道水体的二进制栅格.tif文件，用于生成中心线数据。

### 3) 表格数据 (Tabular_Data)
该目录包含本次分析所用的全部聚合表格数据，涵盖收集的原始数据、整理的已发表数据以及平均WBMsed数据。请注意，空单元格以N/A值填充，代表公开来源中无对应数据或WBMsed模型未输出相关结果。
├── Tabular_Data/
├── Column_Desc/
├── combine_data.py
├── FullCombinedAvgData_050423.csv
├── FullCombinedData_050423.csv
├── FullWBM_data.csv
├── Primary_Data_050423.csv
├── Published_Data/
├── bend_data/
├── "river".csv
├── citations.txt
├── Column_Desc
├── Published_Data_050423.csv
├── PublishedBendData_050423.csv

对于.csv文件，列说明均以`*_file_name_column_desc.txt`的格式存储在Column_Desc/目录下。部分列名存在重叠，本数据集已提供足够信息以帮助理解所有文件中的列含义。
* **combine_data.py**：用于聚合单个弯道尺度 (bend-scale) 河流迁移数据的Python脚本。
* **FullCombinedAvgData_050423.csv**：本数据集内所有河流的河段平均 (reach-averaged) 数据。
* **FullCombinedData_050423.csv**：本数据集内所有河流的弯道尺度数据。
* **FullWBM_data.csv**：所有测量河流的WBMsed数据。
* **Primary_Data_050423.csv**：仅包含原始收集数据的文件。
* **bend_data/**：存储各河流已发表的弯道尺度数据（若有）。请注意，本研究中弯道波长字段为我自行测量所得。
* **citations.txt**：已发表迁移速率数据的来源列表。
* **Published_Data_050423.csv**：聚合后的河段平均已发表数据。
* **PublishedBendData_050423.csv**：聚合后的弯道尺度已发表数据。

## 共享与访问信息
本研究使用谷歌地球引擎 (Google Earth Engine, GEE) 的陆地卫星 (Landsat) 数据作为自然数据源。相关数据集的链接如下：[GEE上的Landsat目录](https://developers.google.com/earth-engine/datasets/catalog/landsat)