AbsoluteNaturalDisasterRiskIndex

1
Understanding of Fog Structure, Development and
Forecasting, COST722
COST 722 member states Austria, Bulgaria,
Cyprus, Finland, France, Germany, Hungary,
Norway, Poland, Spain, Sweden, Switzerland, UK
web page http//www.lcrs.de ? COST presented
by J. Bendix, Faculty of Geography, Laboratory
for Climatology Remote Sensing (LCRS),
University of Marburg, Email bendix_at_staff.uni-mar
burg.de
1. Motivation 2. Main goals and general outline
of COST722 3. Phase 1 results 4. Phase 2
activities within working groups 5. Outlook
14 October 2004, Cape Town, South Africa
2
1. Motivation
Pearson 2002
3
2.1 COST programme major goals
COST Intergovernmental Framework for
European Cooperation in the Field of
Scientific and Technical Research
Hosted by the European Science
Foundation (ESF)
http//cost.cordis.lu/src/whatiscost.cfm

• co-ordination of nationally funded research
  on a European level
• ensure that Europe holds a strong position
  in the field of
• scientific and technical
  research for peaceful purposes
• cover basic, pre-competitive research and
  activities of public utility
• increasing European co-operation and
  interaction
• solving environmental and cross-border
  problems

4
2.2 COST 722
www.lcrs.de
Chairman Dr. Wilfried Jacobs (DWD) Vice
Chairman Vesa Nietosvaara (FMI)
5
2.3 Main objectives

• Main objective
• To develop advanced methods for very short-range
  forecasts of fog, visibility
• and low clouds, adapted to characteristic areas
  and to user requirements
• This overall objective includes
• the development of pre-processing methods of
  the necessary input data
• the development of the appropriate forecast
  models and methods and
• the development of adaptable application
  software for the production of the forecasts.

6
2.4 Time table
Inventory phase
phase 1 report www.lcrs.de/cost/publications.html
COST-722 Very short range forecasting of fog
and low clouds Inventory phase on current
knowledge and requirements by users and
forecasters, 186 pp.
7
3. Phase 1 results

• Working Group I Existing forecast methods
  (chair Herbert Gmoser, ZAMG)
• Conclusions from inventories
• Intensified use of satellite data for
  nowcasting, better spatio-temporal (3 D) coverage
• NWP models only reliable for large scale
  forcing
• 1D-models suitable for local forecast but
  problems with advection
• Statistical techniques site dependent ? use
  for pre- and postprocessing of model results
• 3D-models to slow improvements required (e.g.
  microphysics) ? development
• Working Group II Requirements from the
  forecasters and from the customers
• (chair Silas
  Michaelides, MSC)
• Conclusions from questionairs (e.g. aviation
  customers, road authorities)
• improve short range forecasting quality
• forecasts should be valid not only for points
  but also for areas of about 20x20 km
• demand on probability forecasts
• increased temporal and spatial resolution of
  observation and measurement data is required
• satellite data with fog channel
• more information about the climatology of fog
  at more sites

8
4. phase 2 outline and goals

• Working Group 1 initial data (chair Joerg
  Bendix, PUM-LCRS)
• provide more sophisticated data on fog (?
  initialization, validation WG 2)
• provide more detailed climatologies (? tuning of
  statistical methods WG 3)
• Working Group 2 models (chair Andreas Bott,
  UB-IM)
• improve models (microphysics, turbulence etc.)
• Develop cost-effective 3D-models
• Working Group 3 statistical methods (chair
  Silas Michaelides, MSC)
• improve statistical methods
• Develop validation concepts

9
4.1.1 station climatologies and weather situation
A High pressure situation, C cyclonic, X no
advection
A. Glowacka
10
4.1.2 Conceptual models Madrid Airport
red warm advection black catabatic flow
Darío Cano and Enric Terradellas
11
4.1.3 assimilation of satellite data Madrid
Airport

• estimations of horizontal thermal gradients
  around the airport from Meteosat
• initialization of the H1D model ? improvement
  of model results by using the
• catabatic module

Darío Cano and Enric Terradellas
12
4.1.4 Combining SYNOP and satellit data
1961-1990 mean frequency VIS lt 1 km
1989-1999
intermediate maps
13
4.1.5 Combining SYNOP and satellit data
14
4.2.1 1D-model improvement
T. Bergot
15
4.2.2 1D-model performance
False Alarm
Forecast range h
T. Bergot
16
4.2.3 3D-model sensitivity UK Unified Model
12 km
4 km
1 km
Rachel Capon
17
4.2.4 3D-model sensitivity UK Unified Model
12 km
4 km
1 km
Rachel Capon
18
4.3.1 Statistical model for Zurich airport -
predictors

• Considered predictors
• Station Data
• Wind, Wind direction, Temperature, Dewpoint, rel.
  Humidity, Pressure, Net radiative budget
• On site or very close
• Precipitation, 5cm Temperature, soil temperature,
  PM10, sun elevation
• Calculated data (FOGSI fog stability index,
  Inversions, Gradients)
• Power transformed data (ff, rH, Spread, Q)
• aLMo NWP Output
• Several temperature-, wind-, moisture- and
  pressure variables

Meteo Swiss
19
4.3.2 MOS probablistic forecast performance
Good, but biased deterministic forecast
Dijon (winter 2002-2004) Predictors from the
ARPEC model
Y/n
Poor, but unbiased deterministic forecast
Frédéric Atger
YN/(YYYN)
20
5. Outlook

• improvement of satellite retrievals,
  climatologies and data sets
• improvement of 1D-models (microphysics,
  turbulence etc.)
• development of appropriate 3D-applications
  (based on LM, UM etc.)
• adaptation of statistical models (MOS-based,
  decision trees etc.)
• intercomparison experiment at Paris CDG
  airport winter 2004/05
• 1st workshop

http//www.lcrs.de ? COST722