In-situ observations are an important ingredient of numerical weather prediction (NWP). Their availability, timeliness, and quality determine their usefulness in data assimilation and thereby affect the skill of numerical weather forecasts. In Europe, EUMETNET, a network of national meteorological services, operates and develops the EUMETNET Composite Observing System (EUCOS), which in turn makes a major regional contribution to the World Meteorological Organization (WMO) Integrated Global Observing System (WIGOS). ECMWF helps EUCOS to achieve its aims by providing quality monitoring results of in-situ observations that are available in the Integrated Forecasting System (IFS).
Originally created as an operational network in 2002, EUCOS established collaboration across much of Europe with the aim of optimising surfacebased observation activities to improve the quality and cost-effectiveness of NWP at the European scale. Achieving this goal requires day-to-day monitoring of observation availability and quality as well as rapid feedback to data providers in case of missing or erroneous observations. ECMWF’s role is to provide comprehensive observation monitoring statistics to EUCOS on a daily basis, such that action can be taken close to real time if an issue is identified. ECMWF is in a good position to do so because of its advanced data assimilation and observation quality control systems.
Observation types for which ECMWF provides monitoring information to EUCOS are SYNOP (weather observations from ground stations), SHIP (ship-based surface meteorological observations), PROFILER (vertical profiles of wind), BUOY (surface meteorological observations provided by moored and drifting buoys), TEMP (radiosonde reports)/ASAP (Automated Shipboard Aerological Programme), and AMDAR (Aircraft Meteorological Data Relay). Metrics typically include background departures, analysis departures, and usage status. In the case of TEMPs, for example, these are provided for the parameters of temperature, relative humidity, and vector wind. While a single large departure may be due to either an observational or model problem, a step-change increase in departures over several assimilation cycles often indicates an observational issue. In this way observation providers benefit from NWP, which in turn benefits from using their observations. This feedback loop is illustrated by the figure.
Because of its successful operation, EUCOS observation monitoring has served as a model for the WIGOS Data Quality Management System (WDQMS), which aims to expand the EUCOS concept to the whole globe, again with strong involvement of ECMWF in the monitoring component. In this case the monitoring information is received by the WIGOS Regional Centres, which then contact data providers so they can initiate corrective action. WDQMS is still in the process of being fully established, and it is expected that its operation will lead to another major improvement in the quality and reliability of meteorological in-situ observations worldwide.