Monitoring of satellite data used in operational NWP is a critical activity. The NWP SAF monitoring covers the following data type categories:
- Level-1 products (radiances)
- Level-2 derived products (atmospheric motion vectors, scatterometer surface winds and ozone products).
and is divided into 2 complementary activities:
- Near-real time (NRT) monitoring and anomaly alert focuses primarily on quality checks of the available systems. This helps to ensure that forecast quality is not compromised by degraded performance of the data, which can occur due to instrument degradation or errors within the product delivery chain. There are several elements to this:
NRT availability Currently data coverage, but plan to extend to provide more complete information on timeliness and availability of the data under CDOP-3 (2017 onwards) NRT quality statistics Focus on instantaneous quality (noise, bias, stability) and time evolution of the NRT data Data alerts Automatic interrogation of the global monitoring to identify potential anomalies and communicate as alerts to data providers and wider user community DBNet monitoring Checks availability and consistency of data from the direct broadcast network compared with global datasets
- Data assessment, impact and investigation focuses on more in-depth assessment of data quality. This is very important as successful data usage requires a good understanding of data quality and long-term error characteristics including biases. This work should help to accelerate the identification of problems and exploitation of new datasets. The work here is divided into 2 categories:
Winds quality evaluation Monthly monitoring of AMVs and scatterometer winds to drive in-depth analysis of the data Data assimilation and Ensemble (DAE) diagnostics Provision of diagnostics, forecast sensitivity to observations (FSO) and others, providing additional information on the value of observing systems.
Further information and links to the monitoring charts can be accessed using the Monitoring drop-down tabs above.