Fourier Transform Infrared Spectroscopy (FTIR) template description

Document history

Authors

Overview

Template description

• Instrument Naming
Although FTIR spectrometer systems have the capability to measure multiple atmospheric entities simultaneously, we adopt the convention that each file reports data for a single target gas. Therefore, the different files are distinguished by the instrument name that appears in the filename and that is a concatenation of the instrument name (FTIR) + primary measured target gas [GAS] reported in the file, as shown in Table .


• Variable Reporting
Each measured primary entity [GAS] requires a mandatory set of variables to be reported within a FTIR.[GAS] file. A description of the variables is given in Table . Notice that all variables are mandatory for the submission of vertical profile data. If only total column data are reported, then some variables need not be reported.

Individual data will be reported, each with its own random and systematic error estimates and averaging kernels. As such the same data files that are archived in NDACC can be used for satellite validation purposes and any other applications.


• Additional Notes
To accommodate variable altitude retrieval grids from one reported data point to another, however always with the same number of retrieval levels (PROFFIT retrieval code) or layers (SFIT2 retrieval code), the ALTITUDE.LEVEL.INDEX and ALTITUDE.LAYER.INDEX vectors have been introduced, respectively. The retrieval altitude grid (variable ALTITUDE) is then referenced to this variable. In the case of SFIT2, ALTITUDE contains the output altitude vector (i.e., the center altitudes of the retrieval layers, N in number). The target gas mixing ratio profile and the pressure and temperature profiles are referenced to this ALTITUDE vector. The ALTITUDE.BOUNDARIES vector provides the additional information about the layer boundary altitudes; it is a variable dependent on ALTITUDE (its first dimension must have size N; its second dimension has size 2). It provides the boundaries of the layers for which the vertical partial columns of the target gas are reported ([GAS].COLUMN.VERTICAL.PARTIAL_ABSORPTION.SOLAR (LUNAR)). In the case of PROFFIT, ALTITUDE contains the N altitude levels upon which the target gas mixing ratio profile is retrieved and reported; also the pressure and temperature profiles are reported on this altitude grid. The ALTITUDE.BOUNDARIES vector has again size (N,2); the boundary altitudes are equal to the successive level altitudes. As the number of partial columns between the levels is one less than the number of levels, the highest partial column is set to zero. To avoid end-user confusion, it is essential to explain the interpretation of the dependencies of the geophysical parameters (mixing ratio, pressure, temperature, and partial columns) on the altitudes in the VAR_DESCRIPTION variable.


• Variable Fill Values
The variable fill value is a number inserted as a substitute data element if a data element of a variable is missing or erroneous.


• File Granularity
The choice of the file granularity for archiving FTIR measurements in NDACC is left to the data provider, but it is forbidden that the file covers more than 1 year of data. It is possible now to store in the same file data retrieved with different a-priori profiles, and therefore different averaging kernels and uncertainty estimates, as all individual data are reported. The most common granularities will be monthly, seasonal (3 months) or yearly.


• Metadata - Global Attributes
Each FTIR.[GAS] file requires one set of Global Attributes. These have been grouped into three categories describing the file contents, namely Originator Attributes, Dataset Attributes and File Attributes. An example of global attributes for an FTIR methane measurement at Reunion (PI and data submitter: M. De Mazière, BIRA.IASB; Data analysis using SFIT2 (layer based retrieval): Corinne Vigouroux, BIRA.IASB) is given in Table .


• Metadata - Variable Attributes
Each variable reported in a FTIR.[GAS] file requires one set of Variable Attributes. These have been grouped into two categories describing the variable, namely the Variable Description Attributes and the Variable Visualization Attributes. An example of an attribute set is given in Table .

Templates and examples