Weather forecasting over the Nordic region spans a wide range of phenomena and scales and includes continental, maritime, and polar conditions. During summer, convective systems are common (in particular in the southeast), while polar processes, including severe polar lows, are frequently observed during winter (in the northern parts). Finland, Norway and Sweden have varying topography, complex coastlines, and gradients in land use, which imply local variations in weather. Thus, many aspects of weather forecasting in the Nordic region benefit from a local and probabilistic description of small-scale phenomena and forcing. The focus of this presentation is the predictability of precipitation and its scale‐dependency.
The MetCoOp ensemble prediction system (MEPS) has 9 members and 1 control. Each ensemble member is perturbed with respect to initial and boundary (lateral and surface) conditions. ECMWF forecasts are used at the boundaries and 3DVAR provides the initial conditions for the control. We show that in MEPS predictability of precipitation for scales smaller than ∼60 km is lost rapidly within the first 6 hours of the forecast with the smallest predictable scale growing more slowly to ∼100 km over the following 18–24 hours. However, there is large case‐to‐case variability and the ensemble perturbations fail to become fully saturated, especially in winter, suggesting a current weakness in the design of the ensemble. The added value of MEPS over deterministic forecasts and coarser resolution EPSs (i.e. ECMWF ENS) is discussed with summary statistics and case‐studies. It is shown that the added value varies between seasons and lead times. For precipitation there is an added value for both severe precipitation events and for precipitation/no precipitation decisions. The added value is higher in summer compared to winter and for shorter lead times compared to longer lead times.
Probabilistic weather forecasts serve different user requirements. To optimize the added value of MEPS, MET Norway is currently rethinking how weather warnings are best issued and distributed to end‐users depending on the weather situation. For example, EPS data from both MEPS and ECMWF will be used to make an automatic monitoring system for dangerous weather. The monitoring will frequently and quickly scan the numerical forecast data depending on given thresholds for, e.g., different warning levels, and present the meteorologist with potential warnings by weather phenomenon as a function of forecast range and geographical region. This will give the meteorologist a quick overview for further elaboration before issuing official warnings. We start with precipitation at different scales, intensities and durations, and if successful, will add warnings of dangerous weather for wind, forest fire, storm surge, icing and polar lows.
MEPS forecasts are at the time of writing issued every 6 hours, and 10 members is on the low side when it comes to a realistic sampling of uncertainties. In addition we would like to use the computing resources more optimally. We present first results of a new MEPS configuration that produces ensemble members continuously and every hour. The new system gives around 30 members per every 6 hours with improved representation of the uncertainties. Since members will be available at different times (during a 6 hour interval), the new system introduces both challenges and possibilities with respect to lagging and postprocessing of the model output. Many end-users expect or believe the forecast are continuously updated, and during severe and extreme weather events rapid updates corrected with recent observations are of importance for, e.g., the meteorologists. Reduced jumpiness between subsequent forecasts is another potential advantage. A challenge is to present the large and complex data volumes to the general public for better decisions. We will present some recent examples from MET Norway’s web and app service Yr.
Future challenges include getting the practical predictability closer to the intrinsic predictability. Ongoing work includes to design better perturbations representing uncertainties in the initial conditions, model formulations and lateral boundary conditions. A scale‐dependent predictability diagnosis will highlight what scales are affected by the proposed perturbations. The approach presented here is also beneficial for both prioritizing and evaluating the introduction of new observation types and data assimilation algorithms for increased predictability on smaller scales. Another future challenge is model systems that make land-surface atmosphere feedbacks more relevant to weather predictions, and whether interactions between soil moisture, convection and precipitation increase small-scale predictability and forecast quality.
MEPS a branch of the shared ALADIN-HIRLAM NWP system, which is jointly developed within the framework of the two major European modelling consortia, a research cooperation of 26 meteorological institutes. The ALADIN-HIRLAM consortia collaborate on long-term research, as well as its code and system development. MetCoOp delivers an operational NWP system through short-term research focusing on aspects of the model system which need some extra push to become operational and/or are unique to the Nordics. A close two-way coupling between research and operations is a main pillar in the organization of MetCoOp: It efficiently transfers results from research to operations, creating a lasting legacy in our weather prediction capacities; and the operational weather forecasting pushes the model development.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner