Henning Gisselø Danish Meteorological Institute in Copenhagen Operational forecaster for 17 years.

Henning Gisselø

• Danish Meteorological Institute in Copenhagen

• Operational forecaster for 17 years

S03 HIRLAM (NWP) has problems when predicting 2mTemp (max and min) in cases with no or just a few clouds.

S03 HIRLAM

• 3.3km * 3.3km * 40 layers * 90sek

• +54 h

• Hydrostatic

• Introduced spring 2009

In situations with only a few clouds

the model is predicting too low max temp in day time and too high min temp in night time.

CASE

The night between

the 13th and 14th October 2009

Analysis, 14th Oct. 00 UTC

NOAA, 14th Oct. 02:43 UTC

S03 2mT Min, wind, cloudcover

S03 2mT Min, wind, cloudcover

- 3

- 5

- 2 - 3

- 4

- 4

- 4

- 1

+ 3

Possible explainations from

The HIRLAM Developing Group

DAYTIMECLOUDCOVER

• MODEL: many clouds

• ACTUALLY: few clouds

• RESULT: higher temp observed than predicted by model

DAYTIMEEVAPORATION

• MODEL: high evaporation (moist soil)

• ACTUALLY: low evaporation (dry soil)

• RESULT: higher temp observed than predicted by model

NIGHT TIMECLOUDCOVER

• MODEL: many clouds

• ACTUALLY: few clouds

• RESULT: lower temp observed than predicted by model

NIGHT TIMEHEAT CONDUCTION FROM GROUND

• MODEL: high heat conduction from ground

• ACTUALLY: low heat conduction from ground

• RESULT: lower temp observed than predicted by model

NIGHT TIMEVEGETATION

• MODEL: low vegetation

• ACTUALLY: high vegetation

• High vegetation cools more efficient than low vegetation, producing a cooler surface layer.

• RESULT: lower temp observed than predicted by model

Growing through spring and summer

END OF STORY