Section 9.2 Screen Level Temperatures and Moisture

Derivation of 2m Temperatures

2m temperature is a diagnostic, not a direct output from the atmospheric model.  Forecast screen level variables are not directly evaluated but are determined via knowledge of the the lowest model level temperature (at 10m) and the "skin" temperature of the underlying surface.  In turn the skin temperature is determined via the land surface scheme HTESSEL, or the lake surface scheme FLake, or the sea-surface temperature (through NEMO).  Temperatures at 2m are then derived by interpolation between the forecast 10m temperature and the forecast surface skin temperature.

Fig9.2.1:  Values at the 2m level (e.g. temperature) are not taken from a model level as such but are computed via interpolation between skin temperature and the level of the lowest atmospheric model level (level 137). 2m dew point is derived from the humidity interpolated in a similar way.  The nature of the interpolation profile used depends on other factors, such as wind speed.

FLake is currently not used directly in re-forecasts and consequently M-climate temperatures can be error prone over and adjacent to inland lakes, particularly if these are large (e.g. Lake Superior).  This can affect Extreme Forecast Index(EFI) and Shift of Tails (SOT) values.

Note that forecast temperatures on meteograms are adjusted to the station height by using a 6.5K/km lapse rate applied across the difference between the station height and the ensemble orography.   Therefore departures from the atmosphere's lapse rate, particularly where there is an inversion at low altitude above the local surface, will result in errors in the meteogram 2m temperature.

Note urban and heat island effects are not currently represented in HTESSEL.   Users should assess the potential for deficiencies in low-level parameters and adjust as necessary.

Additional Sources of Information

(Note: In older material there may be references to issues that have subsequently been addressed)

• Read more on "Reducing surface temperature errors at coastlines".
• Read more on Bias in IFS.
• Watch a comprehensive lecture on Model Physics: Concepts, Practice and Products (36sec delay in start, screen temperatures discussed from 17min50sec to 26min09sec).
• Watch a webinar on statistical post processing of ENS products.