Brockmann Consult GmbH NetCDF Format Conversion User Guide 31.08.2020
6
2 Motivation
This SMOS Earth Explorer to NetCDF converter software shall enable a broader range of tools to make use of the
SMOS data. Therefore, the widely supported NetCDF 4 file format has been chosen as target format.
The Earth Explorer format as being distributed by ESA is well suited for the SMOS data and for certain
architectures of processing, especially cell-by-cell Level 3 operations. Nevertheless, in other situations it is more
convenient to access the measurement variables directly than to be forced to iterate over a sequence of structures.
The converter tool performs this re-mapping of the data by flattening the structures and mapping variables to data
arrays ordered by grid-point or by snapshot.
3 Output file format
The converter output file format is NetCDF 4 with the option of writing the data in different compression levels.
3.1 Data Format
The data structure as present in the original Earth Explorer binary data files is not suited to be directly transformed
to NetCDF. The essential structure in the SMOS EEF product format is the grid point, which contains all
measurements that were acquired for that grid point by different snapshots. This original data structure can roughly
be described as a “list of structures that contains lists of structures”. Although, the format is perfectly suitable to
represent the SMOS data, it needs to be modified to match the requirements of users.
Therefore, the NetCDF file contains a serialised version of the structured data. Each grid point or snapshot data
variable is transformed into a NetCDF variable with an appropriate dimension. In the case of e.g. L1C Brightness
Temperature (BT) measurements, the structure member is translated into a two-dimensional array, one dimension
of this array is the number of grid-point measurements in the EE file, the other is the maximal number of snapshot
measurements in all grid points of the product.
All size reference variables translate into NetCDF dimensions; all structures are flattened. Array data with a
variable dimension (like e.g. Brightness Temperature data for grid points) translates into NetCDF arrays with a
fixed dimension (either set to the maximum value allowed by the data type or to the maximum value occurring in
a file).
Variable attributes in NetCDF files like scaling, units, fill values, valid ranges, flag masks, and flag meanings are
defined according to the product specifications.
3.2 Metadata
All metadata contained in the Earth Explorer file is transferred to the NetCDF file. In contrast to the XML-based
metadata in the original file, NetCDF does not allow for structured global metadata elements. Therefore, the
inherent structure is mapped to the metadata attribute names. Any metadata attribute originally contained in a
structure will be converted to a NetCDF attribute whose name is preceded by the structure name, separated by a
colon (“:”), nested structures are treated recursively, according to this rule.
Example:
The “Validity_Start” metadata-element contained in the “Validity_Period” structure nested within the
“Fixed_Header” structure is stored in the NetCDF file as a global attribute, which is denoted
"Fixed_Header:Validity_Period:Validity_Start".
3.3 Dimensions
A NetCDF file requires all dimensions being used for variables to be stored as global meta-information. The
dimension names chosen for the output file reflect the entities stored from the Earth Explorer file and are self-
explaining.
剩余28页未读，继续阅读
徴羽Axon
- 粉丝: 1
- 资源: 16
我的内容管理
展开
最新资源
- zlib-1.2.12压缩包解析与技术要点
- 微信小程序滑动选项卡源码模版发布
- Unity虚拟人物唇同步插件Oculus Lipsync介绍
- Nginx 1.18.0版本WinSW自动安装与管理指南
- Java Swing和JDBC实现的ATM系统源码解析
- 掌握Spark Streaming与Maven集成的分布式大数据处理
- 深入学习推荐系统：教程、案例与项目实践
- Web开发者必备的取色工具软件介绍
- C语言实现李春葆数据结构实验程序
- 超市管理系统开发：asp+SQL Server 2005实战
- Redis伪集群搭建教程与实践
- 掌握网络活动细节：Wireshark v3.6.3网络嗅探工具详解
- 全面掌握美赛：建模、分析与编程实现教程
- Java图书馆系统完整项目源码及SQL文件解析
- PCtoLCD2002软件：高效图片和字符取模转换
- Java开发的体育赛事在线购票系统源码分析
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
