Cumulonimbus clouds, which are associated with many hazards for aviation, can be automatically detected from weather radar data.
Observations and forecasts of Cumulonimbus (CB) clouds are an important part of aviation weather service, because they are associated with strong turbulence, hail, graupel and icing which may be fatal during the flight. Traditionally, CB clouds are observed by human eyes. When a human observer is not available, radar based methods can be used. As part of this development FMI has made quality assessment of the radar-based CB warnings.
A human observer can recognize CB clouds from their anvil/mushroom-like appearance. However, the clouds cannot be seen by human eyes in darkness or behind heavy rain or intervening lower cloud layers. It is also difficult to estimate distances to clouds with bare eyes, and the highest CB clouds can be seen from 200 km distance.
FMI aviation weather stations use human observers when they are available. When the human observer is not in duty, automatic weather stations perform the observations of temperature, wind and other meteorological variables. These automating weather reports are complemented with CB-analysis based on weather radar measurements.
In this project we have assessed the quality of the radar-based CB cloud observations by comparing reports by human, satellite, lightning location sensors and radar. The verification of this automatic CB detection is challenging, as good ground truth data are not often available. Therefore, statistical estimation of the relevant verification measures from imperfect observations using Latent Class Analysis (LCA) was explored. LCA has been used in similar studies in social and medical sciences, but not for meteorological data. Results suggest that LCA gives reasonable estimates of verification measures and, based on these estimates, the CB detection system at FMI gives results comparable to human observations.
Researcher Otto Hyvärinen, tel. +358 50 337 7289, firstname.lastname@example.org
Otto Hyvärinen, Elena Saltikoff, and Harri Hohti, 2015: Validation of Automatic Cb Observations for METAR Messages without Ground Truth. J. Appl. Meteor. Climatol., 54, 2063–2075.
The Finnish Meteorological Institute is a leading expert in meteorology, air quality, climate change, earth observation, marine and arctic research areas. FMI is in a unique position to study various themes of climate change in the Northern context.
High-quality observational data and research is utilized to develop services to benefit our everyday life. Visible examples are improvement of weather forecasts, development of new expert and warning services as well as applications of the newest research results.