Waves to Weather (W2W)
Description:
This special collection comprises the results of the Collaborative Research Center “Waves to Weather” (W2W), which is funded by the Deutsche Forschungsgemeinschaft (German Research Foundation) for a period of 4 years with possible extensions up to 12 years. The main topic of W2W is predictability and prediction of weather. The current scientific themes of W2W are "Upscale error growth", "Cloud-scale uncertainties", and "Predictability of local weather". It includes theoretical studies, numerical modeling, and process studies based in part on cutting edge statistical methods and visualization tools, NWP models and data collected during the field campaign NAWDEX.
The aim of W2W is to identify the limits of predictability of weather and to produce the best forecasts that are physically possible. The focus of W2W is on the most important causes of remaining uncertainties in weather prediction, which include:
- the quick upscale growth of forecast errors from insufficiently resolved or represented processes like convection or boundary layer mixing, which modify synoptic-scale waves,
- our limited understanding of processes in clouds, and
- the influence of local factors on weather that influence the predictability associated with larger-scale wave disturbances.
W2W addresses these three areas in a concerted effort involving contributions from the disciplines of atmospheric dynamics, cloud physics, statistics, inverse methods and visualization.
W2W uses, and further develops a broad range of tools, including numerical models with detailed treatment of cloud processes and aerosols, and ensemble forecasts with sophisticated statistical post-processing to describe uncertainty. Improved insight has already been gained through the development of new interactive visualization methods, that enable rapid exploration of forecast ensembles to identify the sources and evolution of uncertainty in meteorologically significant features, as well as through the unprecedented dataset collected during the international field campaign NAWDEX.
W2W currently consist of eighteen individual scientific projects located in Germany (Ludwig-Maximilians University of Munich, Karlsruhe Institute of Technology, Johannes Gutenberg University in Mainz, German Aerospace Center (DLR) Oberpfaffenhofen, and University of Heidelberg).
Collection organizers:
Audine Laurian and George C. Craig, Meteorological Institute, Ludwig-Maximilians University, Munich, Germany
Waves to Weather (W2W)
Abstract
Weather predictions 2–4 weeks in advance, called the subseasonal time scale, are highly relevant for socioeconomic decision-makers. Unfortunately, the skill of numerical weather prediction models at this time scale is generally low. Here, we use probabilistic random forest (RF)-based machine learning models to postprocess the subseasonal to seasonal (S2S) reforecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF). We show that these models are able to improve the forecasts slightly in a 20-winter mean at lead times of 14, 21, and 28 days for wintertime central European mean 2-m temperatures compared to the lead-time-dependent mean bias-corrected ECMWF’s S2S reforecasts and RF-based models using only reanalysis data as input. Predictions of the occurrence of cold wave days are improved at lead times of 21 and 28 days. Thereby, forecasts of continuous temperatures show a better skill than forecasts of binary occurrences of cold wave days. Furthermore, we analyze if the skill depends on the large-scale flow configuration of the atmosphere at initialization, as represented by weather regimes (WRs). We find that the WR at the start of the forecast influences the skill and its evolution across lead times. These results can be used to assess the conditional improvement of forecasts initialized during one WR in comparison to forecasts initialized during another WR.
Significance Statement
Forecasts of winter temperatures and cold waves 2–4 weeks in advance done by numerical weather prediction (NWP) models are often unsatisfactory due to the chaotic characteristics of the atmosphere and limited predictive skill at this time range. Here, we use statistical methods, belonging to the so-called machine learning (ML) models, to improve forecast quality by postprocessing predictions of a state-of-the-art NWP model. We compare the forecasts of the NWP and ML models considering different weather regimes (WRs), which represent the large-scale atmospheric circulation such as the typical westerly winds in Europe. We find that the ML models generally yield better temperature forecasts for 14, 21, and 28 days in advance and better forecasts of cold wave days 21 and 28 days in advance. The quality of forecasts depends on the WR present at the forecast start. This information can be used to assess the conditional improvement of forecasts.
Abstract
Weather predictions 2–4 weeks in advance, called the subseasonal time scale, are highly relevant for socioeconomic decision-makers. Unfortunately, the skill of numerical weather prediction models at this time scale is generally low. Here, we use probabilistic random forest (RF)-based machine learning models to postprocess the subseasonal to seasonal (S2S) reforecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF). We show that these models are able to improve the forecasts slightly in a 20-winter mean at lead times of 14, 21, and 28 days for wintertime central European mean 2-m temperatures compared to the lead-time-dependent mean bias-corrected ECMWF’s S2S reforecasts and RF-based models using only reanalysis data as input. Predictions of the occurrence of cold wave days are improved at lead times of 21 and 28 days. Thereby, forecasts of continuous temperatures show a better skill than forecasts of binary occurrences of cold wave days. Furthermore, we analyze if the skill depends on the large-scale flow configuration of the atmosphere at initialization, as represented by weather regimes (WRs). We find that the WR at the start of the forecast influences the skill and its evolution across lead times. These results can be used to assess the conditional improvement of forecasts initialized during one WR in comparison to forecasts initialized during another WR.
Significance Statement
Forecasts of winter temperatures and cold waves 2–4 weeks in advance done by numerical weather prediction (NWP) models are often unsatisfactory due to the chaotic characteristics of the atmosphere and limited predictive skill at this time range. Here, we use statistical methods, belonging to the so-called machine learning (ML) models, to improve forecast quality by postprocessing predictions of a state-of-the-art NWP model. We compare the forecasts of the NWP and ML models considering different weather regimes (WRs), which represent the large-scale atmospheric circulation such as the typical westerly winds in Europe. We find that the ML models generally yield better temperature forecasts for 14, 21, and 28 days in advance and better forecasts of cold wave days 21 and 28 days in advance. The quality of forecasts depends on the WR present at the forecast start. This information can be used to assess the conditional improvement of forecasts.
