The FIDUCEO project aims to generate Fundamental Climate Data Records (FCDRs) with rigorous uncertainties developed through the use of metrological standards and practices.
Metrology is the science of measurement which became a recognised discipline in 1875, when representatives of 17 nations signed the Convention du Mètre. This convention founded the Bureau International des Poids et Mesures (BIPM) as a scientific body, researching physical standards and the determination of physical constants. Such research continues both at the BIPM itself and at the National Metrology Institutes (NMIs) of member nations.
It is this metrology community that is responsible for maintaining the International System of Units (SI) in a way that ensures the units are stable over time, uniform worldwide, insensitive to the conditions of measurement, independent of the method used to realise the unit and able to be improved as technological advances enable and demand higher accuracy. The SI achieves this by using definitions of the units that are linked to a stable reference, and are independent of the means of realising that reference. When unit definitions are improved, this is always done consistently with the historical definition.
The metrology community therefore has significant experience in achieving long term stability of measurements, which is what is needed to provide a benchmark for long term climate records. This is achieved within metrology through three key concepts: Traceability, Uncertainty Analysis and Comparison. Traceability, in a metrological sense, is the existence of an unbroken chain of calibrations, which links a measurement of interest to the SI. Uncertainty analysis propagates uncertainties through this chain, with each step taking in the uncertainties of the previous step and combining these with the uncertainties inherent in the new step. This should be done in a way that also considers the correlations between quantities being combined. Uncertainty and traceability is validated in the metrology community through both formal peer review and regular intercomparisons where the measured values and associated uncertainties of one NMI are formally compared with those of other NMIs.
The FIDUCEO project is aiming to see how these metrological concepts can help with obtaining long term FCDRs that likewise are stable, uniform, insensitive to the conditions of measurement and independent of the methods used. It is clear that the lab-based world of the NMIs does not immediately translate to the challenges of Earth Observation sensors, particularly without the availability of an SI-reference in space and especially for the older sensors which were designed and calibrated for purposes that did not have the same requirements as climate research and therefore do not have rigorous pre-flight SI-traceable characterisation.
However, the approach and world-view of metrology can help with establishing the stability of a multi-decadal dataset even where no traceable absolute calibration is available. This will involve understanding, as far as possible, the physical processes involved in the detection optics, sensor and electronic systems of all the different sensors in order to describe, and quantify, the different sources of uncertainty and to propagate them through these processes. It will also involve harmonising the series of sensors – performing cross-comparisons to an independent reference (where available) and to other sensors in the series where both the differences, and uncertainties, are compared and tested. This will give us confidence in the uncertainty and stability estimated for even the earliest sensors, as long as there is an unbroken chain of such comparisons up to modern sensors.
The FIDUCEO project has an audacious aim – to introduce this metrological rigour into the historical record. This will rely on a strong collaboration between metrologists, instrument experts, producers and users of FCDRs. This collaboration needs to and will be, based on a continuous dialogue of mutual understanding and learning to speak each other’s language.
Add new comment