Creating FCDRs

What is an FCDR?

A Fundamental Climate Data Record (FCDR) is a long-term record of a geophysical quantity measured by a satellite with all the necessary information to interpret that record in a quantitative manner.

An FCDR consists of a long, stabilised record of uncertainty-quantified sensor observations that are calibrated to physical units and located in time and space, together with all ancillary and lower-level instrument data used to calibrate and locate the observations and to estimate uncertainty.

FCDRs are produced as an initial step in the data processing chain. They are converted into Climate Data Records (CDRs) for higher level products, a process that can combine FCDR data values from different spectral channels and combine FCDR data values from different observational pixels.

An FCDR should be stabilised, which means it should be harmonised or otherwise improved so as to maximise observational stability.  Harmonisation is the recalibration of the sensor observations, using a stated reference and/or overlaps with other sensors in the series.  The purpose of the recalibration is to bring consistency between sensors given the known (best estimate) differences in instrument characteristics (e.g., spectral response functions) between sensors.  A harmonised (or otherwise stabilised) FCDR should enable more stable CDRs to be derived.

What is the challenge?

To ensure that CDRs generated from FCDRs have meaning and can be relied upon, it is important to have a full understanding of the uncertainties.  It is vital to understand the uncertainty associated with the FCDR at the pixel level, and also the error covariance of the FCDR from pixel-to-pixel and from spectral channel to spectral channel.  This can often be a complex task.

How can FIDUCEO help?

The FIDUCEO method provides you with a systematic way to present error covariance information so that it can be used for CDR generation.  This systematic approach also acts as a way of documenting the origin of uncertainties in the FCDR.  The FCDR uncertainty analysis can be broken down into the following steps:

STEP 1:  Determining the measurement function.  This expresses the relationship between the measurand (usually pixel-level Earth radiance, reflectance or brightness temperature) and the input raw quantities used to calculate this (measured counts, calibration system radiances, gains and nonlinearity terms, correction factors).

STEP 2:  Defining uncertainties effects.  For each term in the measurement function, its origin and the different uncertainty effects that influence the value obtained must be defined and recorded.

STEP 3: Determining uncertainties.   For each uncertainty effect, an associated uncertainty should be evaluated.  The correlation of an error in that effect between observational pixels, across time and across spectral bands should also be considered.  All available information should be used to provide the most robust estimate of this uncertainty and correlation and that evidence should be documented.

STEP 4:  Completing the effects table.  For each uncertainty effect the FIDUCEO effects table should be filled in.  This provides a common approach to documenting the necessary information and the details in the table can then more readily be coded into a NetCDF-compliant full-FCDR.

STEP 5: Generating an FCDR.  Choose between producing a full FCDR or an “easy” version.  A full FCDR requires all uncertainties to be fully evaluated and considered.  The Easy FCDR incorporates some simplification and summary, where possible, to allow for a more accessible process, whilst still maintaining the minimum amount of rigour.