IAP logo

Advanced Analytical Methods for Climate Research
Case study 4: Daily Maximum Windspeed in the Netherlands

 IAP logo
The final case study is in some ways similar to the Irish rainfall study (Case Study 2). The aim is to provide data and some specimen analyses, for a realistic climatological example (in contrast to Case Study 1, where a very small dataset was studied).

Contents
  • Up to main case studies page.

    Background to the study
    Large areas of the Netherlands are below sea level, and are protected by a system of dikes. These dikes are continually attacked by waves, and it is necessary to continually monitor the risk of dike failure. Since wind is the main source of waves, the Royal Netherlands Meteorological Institute (KNMI) are currently undertaking a thorough investigation of the wind climate of the Netherlands. This investigation is called the HYDRA project; details may be found here.

    As part of the HYDRA project, KNMI have provided online access to data from their observational windspeed network. A map of the stations is shown opposite. The data are of extremely high quality, and have been corrected for inhomogeneities due to, for example, changes in exposure. Details of the correction procedures may be found on the HYDRA project webpage.

    Since the primary concern here is wave damage, it is of particular interest to study extreme windspeeds. This is similar to Case Study 2, where the GLM approach allowed extreme rainfalls to be studied within the context of changing probability distributions.

    		Click the image to enlarge it
    Map provided by Royal Netherlands Meteorological Institute (KNMI)
    The data provided by KNMI are hourly. This case study focuses on data from 9 sites, each of which have continuous records from 1961-1998. The actual data provided below are at a daily timescale, and have been calculated as the largest of the 4 values at times 0600, 1200, 1800 and 2400 each day. These daily values are referred to as Daily Maximum Windspeeds (DMWS).

    An interesting feature of the DMWS series from these sites is that many of them show strong trends over the 40-year period of record. However, these trends are in opposite directions in different parts of the country. The figure opposite illustrates this. This has been produced by fitting straight lines through annual mean DMWS sequences at each of the 9 sites; the circles in the figure represent the slopes of these lines. The figure clearly shows windspeeds increasing near the coast but decreasing further inland. This pattern is not due to inhomogeneities in the station records (a comparison with NCEP windspeed data over Europe supports this). What is interesting here is the marked differences between DMWS regimes over a small area.

  • To contents of this page.

    Questions of interest

     Some of the questions that may be asked in this case study are:
  • To contents of this page.

    Data and Software

     For this case study, the dataset is provided below, together with some specimen model definition files. These enable Gamma distributions to be fitted to DMWS series using Dr Chandler's GLM fitting software. Details of how to obtain the software may be found on the workshop links page.
  • Up to main case studies page.
    Page last updated: 24th May 2001.