Background
For information on model, data, and statistical methods, click
here
Ensemble forecast
Under development. DMI's storm surge model, forced by 18 nwp ensembles produced by the DMI ensemble system COMEPS.
DKSS-COMEPS 2020
On-line validation
2 days of hindsight, 5 days ahead. Danish stations only.
Pre-operational
Pre-operational, DINI vs. NEA test (since Sept 8th 2023: DINEA) SHUT DOWN 2024-06-17
Quasi-synoptic North Sea surge
Modelled nrt surge, North-Baltic Seas
Modelled nrt surge, Transition Area
Modelled nrt surge, Wadden Sea
Modelled nrt surge, Limfjorden
Modelled nrt surge, Lillebelt
Modelled nrt surge, Roskilde/Isefjord
Results
Last year - operational
Detailed comparison of one calendar year of Danish coastal tide gauge data, and operational
storm surge model predictions. The general sea level forecast quality is evaluated for each station.
The storm surge (extreme high water) forecast quality is evaluated using only the 10 highest events at each station.
Stations are intercompared, and the quality dependency on forecast range is calculated.
Previous years
Operational storm surge quality as above, with reports since 1997.
Opr/test model intercomparison
A "test model" is a combination of weather/ocean model that differs from the operational one,
or a different tuning of the operational system. We compare opr/test models in terms of ordinary statistics
and in terms of Q. We only study the short-term (0-6h) forecast range, thereby putting the emphasis on the
performance of the storm surge model rather than the weather model.
NOOS models
Intercomparison of sea level forecasts from 7 institutes bordering the North Sea.
The forecast exchange are part of the NOOS co-operation.
3 danish west coast stations are part of this exchange.
British tide gauge data is missing for a large part of 2012. Statistics
is to be re-calc'd, when/if this data has been made available.
NOOS models - common station table
Resultatkontrakt
The quality Q is based on the 3 highest events at a fixed, small subset of tide gauge stations.
Q is re-calculated every month, using the preceding 12 months.
On annual basis, Q is calculated for test models also.
Please note that Q is an error measure, that ideally would be zero.
Q is used in reports to the Ministry of Climate and Energy (KILT).
Case Studies
Detailed description of selected storm surge events. In Danish.
Storm surge events since 1999
Nov 1st, 2006
Northerly storm over Kattegat and the Baltic Proper. All-time high sea level in parts
of the Belt Sea region.
Jan 8th, 2005
Westerly hurricane. A new high water record in the Limfjord. Detailed description
of synoptic wind and pressure, tide gauge data, weather forecats and storm surge model results.
Further tracking of the atmospheric low throughout the Baltic region.
Sept, 2004
Westerly storm. A brief storm surge model intercomparison.
Mar 20th, 2004
Westerly storm. Description of the synoptic and predicted weather,
and detailed storm surge model intercomparison.
Dec 6th, 2003
Northerly storm. Description of synoptic and predicted weather, and high water study.
Feb 21st, 2002
Easterly storm. Classification of this sea level high in the western Baltic.
Jan 28th, 2002
Westerly storm. Classification of this sea level high at the West Coast and in the Kattegat.