Publication Details

MATLAB TRAINING COURSE: INTRODUCTION

This presentation introduces the NetCDF data format, which was developed by Unidata at UCAR to store multidimensional scientific data. A key feature of NetCDF is that it is self-describing, meaning it has embedded information, and it is also machine-independent. Because NetCDF is a binary data format, it cannot be read correctly with standard programs like Notepad or MS Excel ; it requires specialized software such as MATLAB, R, Panoply, GrADS, or Climate Data Operators (CDO) to process the information. The presentation uses the analogy of a "sealed box" to describe a NetCDF file, where the binary data is the box, the information inside is the "treasure," and specific software is the "pair of scissors" needed to open it. Before processing the data, it is crucial to read the file's metadata, which acts like an "information sticker" on the box. This metadata describes everything about the data, including its variables, dimensions, and other attributes. You can retrieve this metadata using commands like ncdump -h in a terminal or functions such as ncinfo in MATLAB. As a practical example, the slides show metadata from a Malaysia Department of Environment (DOE) observation station in Pasir Gudang, Johor, with the station ID CA0001. From this metadata, we learn that the file contains hourly ambient temperature data in degrees Celsius. The data represents a single point because the latitude and longitude dimensions are both equal to one.