中国长序列地表冻融数据集——决策树算法（1987-2009）

中国长序列地表冻融数据集——决策树算法（1987-2009）是利用被动微波遥感 SSM/I亮度温度资料通过决策树分类提取得到。
该数据集采用EASE-Grid投影方式（等积割圆柱投影，标准纬线为±30°），空间分辨率25.067525km，提供逐日的中国大陆主体部分的地表冻融状态分类结果。数据集按年份存放，共由23个文件夹组成，从1987到2009。每个文件夹里包含当年的逐日地表冻融分类结果，为ASCII码文件，命名规则为：SSMI-frozenYYYY***.txt，其中YYYY代表年，***代表儒略日（001~365/366）。冻融分类结果txt文件可直接用文本程序打开察看，还可用ArcView + Spatial Analyst扩展模块或者Arcinfo的Asciigrid命令打开。
提取地表冻融的原始数据来源于由美国国家雪冰数据中心（NSIDC）处理的1987 年以来的逐日被动微波数据，这一数据集采用EASE-Grid（等面积可扩充地球网格）作为标准格式。
中国地表冻融长时间序列数据集——决策树算法（1987-2009）属性由该数据集的时空分辨率、投影信息、数据格式组成。
时空分辨率：时间分辨率为逐日，空间分辨率为25.067525km，经度范围为60°～140°E，纬度为15°～55°N。
投影信息：全球等积圆柱EASE-Grid投影，关于EASE-Grid投影的详细信息见数据准备中关于这种投影的说明。
数据格式：数据集由1987到2009共23个文件夹组成，每个文件夹里包括当年的逐日地表冻融分类结果，按日存储为txt文件。文件命名规则：例如SMI-frozen1994001.txt代表1994年第1天的地表冻融分类结果。该数据集的ASCII码文件是由头文件和主体内容构成。头文件包括行数、列数、x-轴左下点坐标、y-轴左下点坐标、栅格大小、无数据区标值等6行描述信息组成；主体内容为根据行数列数组成的二维数组，以列为优先进行排列，数值为整数型，从1到4，1代表冻结，2代表融化，3代表沙漠，4代表降水。因为该数据集中的所有ASCII码文件所描述的空间为我国全国范围，所以这些文件的头文件是不变的，现将头文件摘录如下（其中xllcenter， yllcenter和cellsize单位为m）：
ncols 308
nrows 166
xllcorner 5778060
yllcorner 1880060
cellsize 25067.525
nodata_value 0
该数据集中的所有ASCII码文件可以直接用文本程序（如记事本）打开。除了头文件，主体内容为数值表征地表冻融的状态：1代表冻结，2代表融化，3代表沙漠，4代表降水。如果要用图示来显示的话，我们推荐用ArcView + 3D 或 Spatial Analyst 扩展模块来读取，在读取过程中会生成grid格式的文件，所显示的grid文件就是该ASCII码文件的图形表达。读取方法：
[1] 在ArcView软件中添加3D或Spatial Analyst扩展模块，然后新建一个View；
[2] 将View激活，点击File菜单，选择Import Data Source选项，弹出Import Data Source选择框，在此框中的Select import file type:中选择ASCII Raster，自动弹出选择源ASCII文件的对话框，点击寻找该数据集中的任一个ASCII文件，，然后按OK键；
[3] 在Output Grid对话框中键入的Grid文件名字（建议使用有意义的文件名，以便以后自己查看）和点击存放此Grid文件的路径，再次按Ok键，然后按Yes（要选择整型数据），Yes（把生成grid文件调入到当前的view中）。生成的文件可以按照Grid文件标准进行属性编辑。这样就完成了显示将ASCII文件显示成Grid文件的过程。
[4] 批处理时，可以使用ARCINFO的ASCIIGRID命令，编写成AML文件，再用Run命令在Grid模块中完成：
Usage: ASCIIGRID <in_ascii_file> <out_grid> {INT | FLOAT}

China Long-Time Series Surface Freeze-Thaw Dataset (Decision Tree Algorithm, 1987–2009) was derived via decision tree classification using passive microwave remote sensing SSM/I brightness temperature data.

This dataset adopts the EASE-Grid projection (Equivalent Cylindrical Projection with standard parallels at ±30°), with a spatial resolution of 25.067525 km, and provides daily surface freeze-thaw state classification results for the main mainland China. The dataset is stored by year, consisting of 23 folders spanning from 1987 to 2009. Each folder contains the daily surface freeze-thaw classification results of that year, saved as ASCII files. The naming convention is: "SSMI-frozenYYYY***.txt", where YYYY represents the 4-digit year, and *** denotes the Julian day (001–365/366). The freeze-thaw classification result TXT files can be directly opened and viewed with text editors, and can also be opened using the ArcView + Spatial Analyst extension module or ArcInfo's Asciigrid command.

The raw data for surface freeze-thaw extraction originates from daily passive microwave data since 1987 processed by the US National Snow and Ice Data Center (NSIDC), which adopts EASE-Grid (Equal Area Scalable Earth Grid) as its standard format.

The attributes of China Long-Time Series Surface Freeze-Thaw Dataset (Decision Tree Algorithm, 1987–2009) consist of its spatiotemporal resolution, projection information, and data format.

Spatiotemporal Resolution: The temporal resolution is daily, with a spatial resolution of 25.067525 km. The spatial coverage ranges from 60°E to 140°E in longitude and 15°N to 55°N in latitude.

Projection Information: Global Equivalent Cylindrical EASE-Grid projection. For detailed information about the EASE-Grid projection, please refer to the projection description in the Data Preparation section.

Data Format: The dataset consists of 23 folders from 1987 to 2009. Each folder contains the daily surface freeze-thaw classification results of the corresponding year, stored as daily TXT files. The naming convention example: "SSMI-frozen1994001.txt" represents the surface freeze-thaw classification result for the 1st day of 1994.
Each ASCII file in this dataset is composed of a header section and a main content section. The header includes 6 lines of descriptive information: number of columns, number of rows, lower-left x-coordinate, lower-left y-coordinate, grid cell size, and no-data value. The main content is a 2D array arranged in column-major order, with dimensions determined by the header's row and column counts. The values are integers ranging from 1 to 4, where 1 represents frozen, 2 represents melted, 3 represents desert, and 4 represents precipitation. Since all ASCII files in this dataset cover the entire territory of China, their header sections are identical. The standard header is excerpted below (note that the units of xllcenter, yllcenter, and cellsize are meters):
ncols 308
nrows 166
xllcorner 5778060
yllcorner 1880060
cellsize 25067.525
nodata_value 0

All ASCII files in this dataset can be directly opened with text programs such as Notepad. Apart from the header section, the main content uses numerical values to indicate surface freeze-thaw states: 1 for frozen, 2 for melted, 3 for desert, and 4 for precipitation. For visualization, we recommend using ArcView + 3D or Spatial Analyst extension module to read the files, which will generate a grid format file as the graphical representation of the ASCII file. The reading steps are as follows:
[1] Add the 3D or Spatial Analyst extension module in ArcView, then create a new View;
[2] Activate the View, click the File menu, select the Import Data Source option, which will pop up the Import Data Source dialog box. Select ASCII Raster in the Select import file type: dropdown menu, then a dialog box for selecting the source ASCII file will automatically appear. Locate any ASCII file in this dataset and click OK;
[3] Enter the name of the Grid file (it is recommended to use a meaningful filename for easy future reference) and select the storage path for the Grid file in the Output Grid dialog box, then click OK again. Select Yes when prompted to use integer data, then select Yes to load the generated Grid file into the current View. The generated file can be edited for attributes according to Grid file standards, completing the process of displaying the ASCII file as a Grid file.
[4] For batch processing, you can use the ASCIIGRID command of ArcINFO by writing an AML file, then execute it with the Run command in the Grid module:
Usage: ASCIIGRID <in_ascii_file> <out_grid> {INT | FLOAT}

搜集汇总

背景与挑战

背景概述

该数据集是一个基于决策树算法和被动微波遥感SSM/I亮度温度资料生成的中国长序列地表冻融数据集，覆盖时间范围为1987年至2009年，提供逐日的中国大陆主体部分地表冻融状态分类结果，包括冻结、融化、沙漠和降水四种状态。数据采用EASE-Grid投影，空间分辨率为25.067525km，以ASCII码文件格式存储，便于文本程序或GIS软件处理，适用于冰冻圈研究和环境监测应用。