Melting layer outliers

[tukiains]

Melting layer method produces sometimes clear outliers. Outliers affect the subzero-region identification because it is a combination of the model wet-bulb temperature and radar-detected melting layer.

Possible solutions:

• Screen the melting layer bit to remove outliers
• Replace gdas model data with ecmwf. Gdas temperature can't be trusted very much! Thus we have to seek the melting layer from wide model temperature range, allowing possible outliers to emerge. Has anyone compared ecmwf / gdas temperatures to the radar-detected melting layer?!

[tukiains]

Kind of solved taking the model type into account when setting the temperature range.