Within KNMI, research is carried out in the fields of weather, climate and seismology. KNMI's meteorological (weather-related) research is primarily aimed at maintaining the quality and ease of access of meteorological observations and model data at a high level according to international standards. Wherever possible, R&D results are used for innovation of the operational production process.Research is being carried out in several areas: atmospherical models, oceanographic models, statistical forecast methods, ground-based observations, satellite remote sensing and climatological research.
For the period 2002-2006, the following main strategic objectives for meteorological research have been defined:
|
|
Fig. 1: Removal of noise and the velocity ambiguity from the Doppler data of the
radar of De Bilt, by applying the dual-PRF technique and the correction-algorithm developed
by Holleman and Beekhuis (2002). Shown are dual-PRF radial wind velocities for 6
November 2000, 14h54 UTC, as a function of distance to the radar and azimuth.
The azimuth scan has been taken at an elevation of 0.5 degrees, using
pulse frequencies of 750 and 1000Hz. Horizontally and vertically are displayed the
distance to the radar and azimuth, respectively. Red colours indicate
wind velocities in directions away from the radar, blue velocities towards the radar.
The figure on the left shows the raw Doppler radial wind velocities.
The image contains noise (particularly at short distances to the radar), and outliers in homogeneous velocity areas.
Reflection of the radar beam against a nearby high building is visible as the horizontal red line at azimuth
values around 245 degrees. The figure on the right shows the Doppler velocity image
after noise reduction, elimination of the blocked azimuth values and removal of the velocity ambiguity.The correction algorithm
of Holleman and Beekhuis is clearly able to remove most contaminations from the velocity
images, hereby makeing them useable for assimilation purposes.
Fig. 2: IWV columns as measured with GPS. Shown are data obtained in near-real-time
from GPS groundtstation Delft, compared with analysed water vapour column values
for that same location from the HIRLAM model. Also presented are IWV-values derived from
METEOSAT infrared (IR) and water vapour (WV) images.
In the spring of 2002, the operational atmospheric model HIRLAM will be renewed and
significantly increased in horizontal resolution. In the context of the HIRLAM-6 2003-2005
research project, the potential will be investigated of further improvements, by means of the
assimilation of detailed remote sensing data and the development of more realistic physical
parametrizations (in particular for clouds and convection). For the tidal model WAQUA and the
ocean and shallow water wave model NEDWAM, very fine-scale model versions (4 - <1km) are
also under development. Furthermore, physical post-processing methods have been set up to
allow analyses and predictions to be made for specific weather parameters (wind, in particular) at
very high horizontal resolution (500-1000m); these methods will be validated extensively and
implemented operationally.
To aid the determination and testing of safety levels for the Dutch coastal defense, the
wind climatology for the Netherlands will be refined further. The Dutch climatology for
precipitation and evaporation will be updated and regionalized as well.
Research on improved methodologies and tools for detecting and predicting severe
weather will be continued. The highest priority in the coming few years will be given to the
development of better observations and prediction methods for (extreme) precipitation and severe
convection. Probabilistic methods will increasingly be used as tools for the interpretation of
mesoscale phenomena, especially in cases of (locally) severe weather (thunderstorms, fog, road
icing).
For the operational control of shipping traffic and harbor access, Dutch coastal zone
authorities have expressed a wish for spatially more detailed weather information (wind in
particular), on scales of ~1km. So-called downscaling methods are being developed to enable the
derivation of such information from the operational mesoscale HIRLAM model (fig.3). The wind
climatology for the Netherlands is being refined to suit the needs of the authorities responsible
for the periodic testing of the required strength of Dutch coastal defense systems.
Fig.3: An example of downscaling of HIRLAM 10m wind analyses to a very fine
grid (mesh size 1km) over the Netherlands. At 28 mei 2000 a compact
storm system swept over the Netherlands, causing wind velocities of 10 Bft in the southern
part of the North Sea, and 9 Bft over Lake IJssel. From left to right are shown the analysed
wind fields for the area around Lake IJssel as obtained from the operational HIRLAM model
(55km mesh), the HIRLAM model at a grid of 11km mesh width, and the downscaling module
applied to the 11km HIRLAM model. The effect of spatial resolution on the level of detail in
the wind fields is obvious. Wind velocities obtained by the downscaling
module are the best in agreement with observations, and show sharp, realistic
gradients in the environment of land-sea transitions.
Hydrological authorities have great need for a better quantitative description of the actual
and predicted precipitation, and for a more detailed precipitation climatology for design purposes.
This will be a focal point for research for the next few years. Additionally, attention will be paid to
enhancing the use of uncertainty information for hydrological applications, and to improving the
coupling of meteorological and climatological data to hydrological decision support systems.
In support of aviation safety, more accurate methods for the prediction of poor visibility,
wind shear and gusts, and severe convection will be implemented. With the help of a downscaling
module, spatially detailed analyses and prognoses will be made available for the behavior of wind
in the neighborhood of the runways of Schiphol airport. This information can be used to optimize
the planning of runway use for take-offs and landings. The observational network at Schiphol
airport will be adapted and extended in view of the construction of a fifth runway and of the
planned automation of visual observations.
Finally, KNMI will attempt to better employ meteorological information to the benefit of
various environmental issues. Very detailed wind information can be used, for example, to
improve the description of the local dispersion of air pollution or of noise levels near airports. In
collaboration with institutes involved in energy research, it will be assessed to what extent
improved and customized meteorological information may lead to more accurate production
prognoses (and hence, lower costs) of wind and solar energy.
2. Improved tailoring to user needs
A major goal is the development of new, or the tailoring of existing, meteorological information to
better suit the needs of several important user groups: coastal zone and harbor authorities,
hydrological management, aviation and environmental agencies.
3. Increase ease of access to actual and historical meteorological data
KNMI aims to be the primary center issuing operational weather data and information to the
Dutch society. High priority will therefore be given to increasing as much as possible the ease of
access to meteorological data by external users from the government, the private sector and the
general public. It is intended to accomplish this by means of:
The infrastructure for obtaining and processing remote sensing image information will be renewed completely, in preparation for the reception of new types of data (MSG, radar Doppler data, METOP/EPS, etc.). In spring 2002, a significant increase in computational power will enable the introduction of improved and more detailed numerical prediction model. Also the systems for archiving historical observations will be updated and extended with a larger diversity of meteorological data.
Finally, continuous attention will be paid to the quality assurance of meteorological data. The quality control of real-time observations will be improved and become more automated. Increasingly, meta-data such as station and instrument changes are recorded and archived automatically. Best practices in how to equip and inspect observing stations and how to process observational data are being formulated and collected in a Handbook on Observations. More effort will also be put in enhancing the quality, reliability and homogeneity of climatological data.