notes 3.2.2005/LR

Working meeting on Nordic Temperature Problem
was held 31 January 2005 10:00-16:30, FMI lecture room, 3rd floor


Collecting the facts and problems = short presentations about the old and ongoing work

Stefan Gollvik: Parametrizations over snow and forest surfaces

Sander Tijm: Experiment results from 1D HIRLAM runs

Mihail Sofiev (presented by Pilvi Siljamo): The year 2000 problem of sensible and latent heat fluxes

Laura Rontu: A wintertime verification picture and Example of using scatterplots

Carl Fortelius (presented by Per Unden): A Sodankylä spring example

Sergei Zilitinkevich: How to parametrize the stable PBL turbulence

Simo Järvenoja: Additional comparisons of predicted and observed temperature and cloudiness. See also Simo's review in Norrköping workshop

Operative Sodankylä-model comparison Saturday 29 January 2005 and Sunday 30 January 2005

Definition of the problems and listing the suspected causes

Northern winter problem

In the model, too warm predicted near-surface temperatures in a stable arctic boundary layer. Differences between observation and forecast of the order of ten degrees are common. In reality, clear sky, no significant SW radiation, shallow surface layer with strong surface temperature inversion over snow covered surface.  Relative humidity may be large but not close to saturation. Observed latent and sensible heat fluxes are small, in the model generally somewhat larger.  Extra clouds/fog seem to form in HIRLAM.

Northern spring problem

Too cold daytime near-surface temperatures, supressed diurnal cycle. Differences between observation and forecast of the order of ten degrees are common. Clear sky cases, magnitude of measured SW radiation and heat fluxes increase towards summer. At least some snow on ground but typically not on trees, ice cover on lakes.  In the model heat fluxes are often large, compensate each other and the ratio of sensible (smaller)/latent (larger) heat is often incorrect. Excessive cloudiness and humidity are frequently produced by the model.

Further analysis: questions


Comparison of three-dimensional cloudiness: observed and predicted by the model?

Extra clouds seem to form in HIRLAM - why?

What is the observed height of planetary boundary layer?

Do capping inversions exist, what would be their role in nature/model?

What is the role of vertical resolution of the model in situations of very shallow boundary layer?


How does evaporation from vegetation influence in model? Snow and its melting on ground/on trees?

Heat fluxes in forest/low vegetation. Heating of forest, small scale convection.
(equilibrium within tiles - influence on neighbouring areas)

Role of soil moisture/ soil moisture assimilation?


- Correlation between cloud cover and low-level temperature bias

- Influence of the overprediction of  weak winds by HIRLAM


Which observational and model data we need to study ?

- cloud condensate content,cloud cover : time/height, maps (cloud, low cloud, fog)

- temperature profiles, moisture (q), wind

- fluxes: latf,senf,radf (all components)

- fog, cloud from SYNOP

- PBLH, capping inversion (from T profiles), types of PBL, relation to synoptic situation

- snow temperature profiles (heat conduction), canopy temperature in model/nature


What data are available for comparisons?

- Sodankylä, Cabauw: profiles, fluxes  - routinely compared with RCR at

- Sodankylä sounding and synop data

- RCR/MBE forecasts from Feb 2004

- Hyytiälä, Värriö, Nopex Uppsala tower data - need to contact university people

- Helsinki testbed boundary layer data - future: 2005-2007

What additional data we would need

Downwelling long-wave radiation at surface (SLWDN)
- available in Cabauw, Uppsala,  the Sodankylä possibilites to be checked (FMI)

To be added to the routine comparisons: SLWDN from model, clouds from SYNOP observation, absolute humidity along with the present relative humidity values?
- Markku

Plan for model experiments and comparisons

Observation - model comparisons

A study of factors controlling T2m in Nordic winter conditions. Systematic comparison of HIRLAM data (analyses, forecasts of RCR 6.3, RCR 6.4, MBE  with Sodankylä obsevations 
- started Timo, Zhenia, in contact with Sergei

Next step: the same for Nordic spring conditions
- Timo, Zhenia, in contact with Sergei

Single column experiments

Winter and spring studies related to condensation parametrization and  cloud-LW-radiation interactions, sources and sinks of  moisture in the model
- Sander to contact Bent, Javier, Karl-Ivar

Evaporation from vegetation in the reference HIRLAM (quick temporary fix?)
- Sander

Full model experiments to test suggested developments

Testing of the new surface scheme (snow/forest etc)
- Stefan et al. with support from FMI/RSHU

Testing of stable PBL flux modifications - HIRLAM 6.3.5 over Arctic
- ongoing, Veniamin

Surface analysis: snow cover analysis, soil moisture assimilation
- ongoing, Ernesto et al.

Basic studies of particular aspects

Study of wind in stable boundary conditions -> turbulence parametrization problem?
- Priit, Laura

Vertical resolution of the HIRLAM - later

Roughness - stability dependence - later

Possible problems in fluxes when used as boundary values for the (moist) turbulence scheme?

Parametrized cloud-radiation interactions, cloud LW emissivity - Ulrika, Nastia?
Comparisons with LES results
- in contact with Igor Esau, Nansen center 

What statistical - diagnostic methods do we need to handle obs/model data?

Standard station verification with advanced tools for analysis (bias classified according to observed variables etc).
- Tools to be added to reference HIRLAM, FMI

Mast verification - pictures, statistics
- ask Eric Bazile about Arpege - Sodankylä comparisons!

Time-height graphs - tools?

Accumulated diagnostics = tendencies, fluxes
- available in the reference HIRLAM


Verification problems

- Forest 2m temperatures in surface analysis - low vegetation tiles instead?

- Representativeness of (Sodankylä mast) observations?


Expected products

1. Report on comparison Sodankylä-RCR in stable winter conditions : Technical report/Newsletter/Journal

2. Web page to keep in contact, to see the plans and what people are doing to fulfill them
    Including references for relevant reports/articles

3. Short term HIRLAM code modifications, long-term also

4. 1D winter and spring cases + tools + data for comparison

5. Recommendations for further research


Sodankylä summer school on boundary layer modelling, 4-14.6.2005

Participants of the meeting at 31 Jan, 2005

Yevgeni Atlaskin
Kalle Eerola
Carl Fortelius
Stefan Gollvik
Sylvain Joffre
Simo Järvenoja
Marko Kaasik
Markku Kangas
Katherina Kourzeneva
Laura Rontu
Patrick Samuelsson
Hannu Savijärvi
Pilvi Siljamo
Sander Tijm
Priit Tisler
Per Unden
Timo Vihma
Sergei Zilitinkevich

contact: laura.rontu at