OB.DAAC Logo
NASA Logo
Ocean Level-2 Data Format Specification

Ocean Level-2 Data Format Specification

Table of Contents

  1. Introduction
  2. Naming Convention
  3. Global Attributes
  4. Data Objects
  5. Sensor-Specific Examples

1. Introduction

This document describes the specifications of Ocean Level-2 archive products that are produced by the NASA Goddard Space Flight Center's Ocean Data Processing System (ODPS) and distributed by the Ocean Biology Distributed Active Archive Center (OB.DAAC). The products are implemented in the Network Common Data Form 4 (NetCDF4), and NetCDF terminology is used in this document. These specifications are given in terms of the logical implementation of the products in NetCDF and are not a physical description of file contents. NetCDF4 software must be used to create or read these products. NetCDF4 is also compatible with the Hierarchical Data Format 5 (HDF5).

A Level-2 data product is generated from either a Level-1A or Level-1B product. The main data contents of the product are the geophysical values for each pixel, derived from the Level-1 radiance by applying the sensor calibration (for Level-1A), atmospheric corrections, and geophysical parameter algorithms. Each Level-2 product corresponds exactly in geographical coverage (scan-line and pixel extent) to that of its parent Level-1 product and is stored in one physical NetCDF file.

2. Naming Convention

The form of our older Level-2 file name is iyyyydddhhmmss.L2_ttt_ppp.nc, where i is the instrument identifier (S for SeaWiFS, A for Aqua MODIS, T for Terra MODIS, O for OCTS, C for CZCS, M for MERIS, V for VIIRS, H for HICO), yyyydddhhmmss are the UTC year, day of the year, hours, minutes, and seconds of the first scan line, ttt is an optional data type identifier, and ppp is the product identifier (OC for ocean color, SST for sea surface temperature or IOP for inherent optical properties). The data type can be:

We are in the process of converting our data over to a new file-naming convention, so be aware that the previously described format will change to: MMMM_IIII_TTT.YYYYMMDDTHHMMSS.LLLL.PPPP.SSSS.pppp.RRRR.NRT.nc. Until all of our data is reprocessed, a mix of both conventions will distributed. Get more details of the updated file-naming convention and view examples.

3. Global Attributes

3.1 Mission and Documentation

This section includes attributes that are common to all sensors, followed by sensor-specific attributes.

3.2 Data Time

These attributes specify the temporal coverage of the data product, and the equator crossing time and direction

3.3 Data Structure

These attributes describe the structure of data objects in the file, in addition to the information provided in the named dimensions.

3.4 Scene Coordinates

These attributes describe the geospatial extent of the data product.

4. Data Objects

Of the following data object groups, three (Scan-Line Attributes, Geophysical Data, and Navigation) contain data that are functions of scan lines. That is, each data object within these groups have data for each scan line and is therefore dimensioned number_of_lines. Thus, to get all the data corresponding to a specific scan line, n, the nth values of all data objects in these four groups would need to be read. All of the data objects have standard attributes long_name, units, valid_min, valid_max and _FillValue.

4.1 Scan-Line Attributes

The data objects in the group scan_line_attributes contain information that applies to an entire line.

4.2 Geophysical Data

The data objects in the group geophysical_data are derived from the sensor science data.  This group also includes the data quality flags. All of the data objects have dimensions  number_of_lines x pixels_per_line. Most parameters are stored as integers, which are scaled according to the attributes scale_factor and add_offset attached to the data object.

Parameters that are wavelength-specific (e.g., Rrs) have separate data objects for each band used to derive the parameter; a list of wavelengths for each sensor provided in the Mission and Sensors entry. The SST and SST4 are stored in separate archive products with their respective quality fields.  The specific fields are listed in the Level-2 File Format for each sensor.  A complete list of parameters that can be output by the software is given in the L2gen User Guide.

The Level-2 data flags and quality fields are stored in the l2_flags data object in each Level-2 file. This object has attributes flag_masks and flag_meanings that provide the names of the algorithms used in determining the setting of the corresponding bits in l2_flags (the least significant bit being the first bit). The algorithms associated with these names, and the use of the corresponding bits as masks or as flags, are described in volumes of the SeaWiFS TM Series. Note that for SST, only the LAND, SSTWARN and SSTFAIL flags are meaningful, as the quality indicators are stored in sst_qual and sst4_qual, respectively.

4.3 Navigation

The data objects in the group navigation_data contain the geolocation information for the data product. This includes the latitudes and longitudes of the observed locations, the control point indices (needed if the latitudes and longitudes are subsampled and not full-resolution), and the G-ring coordinates for the swath.  For tilting sensors, the tilt angle is also included.

4.5 Sensor Band Parameters

The data objects in the group sensor_band_parameters contain information about all of the bands used in the Level-2 processing for each sensor, as well as the band-specific parameters used in the geophysical retrieval algorithms.

4.6 Processing Control

The group processing_control contains two sub-groups: input_parameters, which contains attributes listing all of the parameters used in the l2gen processing run for generating the data product; and flag_percentages, which contains attributes listing the percent of data for which each of the l2_flags bits is set.

5. Sensor-specific Examples

The links below provide detailed information on the attributes and content of sensor-specific files. These listing were automatically generated from the header information of randomly-selected standard Level-2 files produced and distributed by NASA.