Calibration and validation and Fiducial Reference Measurements

All EUMETSAT missions are operational. This implies that Cal/Val is driven by constraints on data dissemination, data continuity, and data quality, as requested by EUMETSAT users. These three key needs are interdependent and can be visualised as the three pillars of a temple (Figure 1), all essential for supporting its roof.

1. Data dissemination

Fast data dissemination, known as near real time (NRT) dissemination, is acquired from the satellite data acquisition. This is essential for fulfilling meteorological nowcasting and forecasting needs. For low Earth orbit (LEO) satellites, data are disseminated in a couple of hours (due to the need for data downloads at reception stations), and for geostationary (GEO) satellites, data are disseminated within 20 minutes.

2. Data continuity

Ensuring consistent data over the mission/program lifetime and between different systems, is crucial for producing long-term data records.

3. Data quality

Data quality, assessed through uncertainties, must meet the standards outlined in the user requirements and Cal/Val plans.

Fiducial Reference Measurements (FRM) are ground-based and airborne measurements and they can be considered a subset of in-situ datasets. These in-situ datasets play a vital role in the Cal/Val process by helping to assess the quality of space-borne Earth observation data for both operational and scientific missions throughout their lifetimes.

Cal/Val often relies on existing ground-based or in-situ data obtained through the World Meterological Organization’s Global Telecommunication System (GTS), the World Meterological Organization’s Global Atmosphere Watch (GAW) Programme, NASA’s Network for the Detection of Atmospheric Composition Change (NDACC), and the European Environment Agency (EEA) as well as through scientific contributions. As beneficial as they are, these existing datasets may not always fully meet the specific or evolving needs of new missions, which is why Fiducial Reference Measurements are so crucial.

Unlike general in-situ data, FRM are specifically defined as a suite of independent, tailored, and fully characterized measurements. These measurements serve as a benchmark for calibrating and validating other data. They maximise the return on investment (ROI) for satellite missions by delivering the necessary confidence required in data products. This is achieved through independent validation results and satellite measurement uncertainty estimates maintained throughout the lifecycle of a satellite mission.

Conducted according to strict protocol and often linked to international standards, FRM are crucial in Earth observation and remote sensing, ensuring accuracy, reliability and consistency in satellite data over time. These measurements are considered the gold standard for comparing against other datasets to assess their accuracy and quality, which is especially important for operational meteorological missions.

Fiducial Reference Measurements (FRM) must address three key requirements directly related to EUMETSAT’s Cal/Val needs, as described in Section 2, whilst also acknowledging the dependencies between them.

1. Data access & format of the ground-based products
   • Easy access to data with a supported product list and a central facility for handling networks.
   • Well-documented products, including detailed descriptions of measurements, retrievals/processing methods, uncertainties and limitations.
   • Timeliness of data availability – ideally less than a month, depending on the satellite products.
   • Data format should be readable and self-explanatory, such as the widely used NetCDF.
   • Clear description of data access and usage policies.
2. Gapless data products
   • Long-term availability of ground-based measurements from stations and networks to ensure coverage for the entire mission lifetime.
3. Accurate ground-based products

• Quality comparable or better than satellite mission accuracy.
• Measurements should be traceable to recognised standards and/or community best practices, following standard operating procedures.
• Consistent and traceable processing, including quality checks and transparent uncertainty assessments.
• Algorithms should be tailored to align better with satellite overpass times.

EUMETSAT’s Cal/Val planning focuses on achieving two key objectives:

1. Providing data products with documented and traceable uncertainty estimates, aligned with the mission user requirements and associated Cal/Val plans.
2. Understanding sensor performance and algorithm characteristics to improve their quality and reliability.

To support these objectives, EUMETSAT has developed a Scientific roadmap for Fiducial Reference Measurements, outlining its approach to supporting FRMs. The roadmap identifies the needs, scientific processes, and activities for Fiducial Reference Measurements, and provides an outlook for the next five to ten years.

The goal of this roadmap is to pinpoint FRM needs for EUMETSAT’s Cal/Val activities and propose a strategy to address future requirements. FRM needs are centralised through a product-wide gap analysis, recognising that many missions provide similar geophysical products. This also highlights commonalities and synergies between the instruments.

The gaps analysis covers:

• Sources of key ground-based measurements.
• Gaps in parameters based on the EUMETSAT mission requirements and Cal/Val plans.
• Gaps in algorithms for ground-based measurements (e.g., how the accuracy of a Radiative Transfer model, a model of atmospheric behaviour, could be improved).
• Gaps in auxiliary data.
• Gaps in station locations.

Current assessments of EUMETSAT needs have identified the necessity of consolidating or developing a clear and user-friendly interface for accessing ground-based and satellite validation data via EUMETSAT tools (Figure 2). Furthermore, a centralised access point would simplify storage management, prevent duplication of data versions from ground-based stations, and streamline EUMETSAT’s Cal/Val tools.