Aus Wufiwiki
Zur Navigation springenZur Suche springen

The *.WET Format for Climate Data

You can use existing IBP weather files (file format *.WET) to perform WUFI calculations; you can also convert your own weather data to that format and use them.

IBP weather files contain 21 lines of header which explain the meaning of the different data columns and an arbitrary number of lines with hourly weather data. WUFI skips the header.



(for formatting reasons, the data lines cannot be reproduced here in their entirety. Please refer to the file IBP1991.WET you received with WUFI for a complete example).

The format of the numbers is free; the only requirement is that a PASCAL program (like WUFI) must be able to read them.
The file need not conform to a specific column format (as in FORTRAN), the number of decimal places does not matter, floating point numbers need not be in exponential format etc.

The number type real/integer has to be adhered to, however. The individual numbers must be separated by blanks, not tabulators.

The fact that the floating point numbers in the example all start with a zero is irrelevant (they were created with a FORTRAN program which likes to write numbers this way).

The data in detail:

  • Day, month, year, hour (integer):
    The IBP files contain hourly mean values of the measured data. These time stamps indicate the point in time when each measuring interval of one hour was finished and the measured mean values recorded. I.e., the data line starting with 01 01 91 04 contains the mean values for the hour from 03:00 to 04:00 on January 1, 1991.
    Since its data are real-life data, an IBP weather file may contain an intercalary day (February 29). WUFI however knows nothing about intercalary days (at least in the present version). These must be deleted from files you want to use for WUFI calculations.
    The time count always refers to winter time (CET). There is no daylight saving time. 

  • Exterior air temperature (real):
    the temperature of the exterior ambient air in [°C]. It is used without change as the exterior air temperature in the WUFI calculation.  

  • Temperature of black surface (real):
  • Temperature of white surface (real):
    Surface temperatures of west-facing test facade elements painted black (as ~ 0.9) and white (as ~ 0.4), respectively.
    These may optionally be read instead of the exterior air temperature. 

  • Temperature of Ground Surface (real):
  • Temperature 50 cm below Ground Surface (real):
  • Temperature 1 m below Ground Surface (real):
    Soil temperatures at various depths in [°C].
    These may optionally be read instead of the exterior air temperature. 

  • Global radiation (real):
    the sum of the direct and diffuse components of the solar radiation incident on a horizontal surface; in [W/m²]. 

  • Diffuse radiation (real):
    the diffuse component of the solar radiation incident on a horizontal surface; in [W/m²].
    Subtracting the diffuse radiation from the global radiation results in the direct radiation. From this, the direct normal radiation can be calculated, i.e., the amount of radiation incident on a surface that is orthogonal to the ray direction. This calculation requires knowledge of the time and geographical location of the measurement, since the corresponding position of the sun has to be determined [2].
    The direct normal radiation can then be used to calculate the direct radiation incident on a surface with arbitrary orientation and inclination. Addition of the appropriate fraction of diffuse radiation results in the global radiation incident on the building component.
    These calculations are automatically done by WUFI during a simulation run.
    WUFI uses a simplified radiation model which treats the diffuse radiation as isotropic and ignores the albedo of the ground. If you are more demanding or want to allow for local specifics (like shadows etc.), you can perform the conversion yourself and feed the results to WUFI via a *.KLI file.  

  • Western radiation (real):
    the global radiation measured with a west-facing solarimeter; in [W/m²].
    Optionally, the radiation component impinging on a (west-facing) facade may directly be read from this column. No conversion is then necessary as would be the case if global and diffuse radiation were used as data source.  

  • Relative humidity (real):
    the relative humidity of the ambient air (0..1). Used by WUFI without change as the exterior relative humidity. 

  • Barometric pressure (real):
    the ambient air pressure reduced to sea level; in [hPA].  

  • Wind velocity (real):
    the scalar mean of the wind velocity; in [m/s]. It is needed to compute the driving rain from the normal rain.
    The scalar mean of the wind velocity is the mean value of the readings of an anemometer without the directions taken into account. For example, if for an hour the wind is blowing at 2 m/s but at the same time steadily travelling through all the direcions, then the scalar mean is 2 m/s, whereas the vectorial mean is 0 m/s.
    In the vectorial mean, some vector components may cancel. Thus, the scalar mean is greater than (or at least equal to) the vectorial mean.
    The IBP weather files also contain information about the distribution of the wind directions (see below). This allows to ignore components that come from 'behind' (and don't contribute to the rain load) and to weight the remaining directions according to the angle they make with the surface.  

  • Normal rain / tipping bucket (real):
  • Normal rain / droplet counter (real):
    the normal rain (i.e., the rain incident on a horizontal surface in open area) during one hour, measured with a tipping bucket and a droplet counter rain gauge, respectively; in [Ltr/m²h] or, equivalently, [mm/h].
    WUFI uses the readings of the droplet counter, since it is in general more reliable.
    The rain load on a wall is determined by the driving rain rather than the normal rain. The driving rain can be estimated from the normal rain and the wind velocity:
    The rain coefficients R1 and R2 in the dialog orientation, inclination and height serve to estimate the driving rain load on a surface of arbitrary orientation and inclination from data on normal rain, wind velocity and wind direction distributions, using the relation
    rain load = rain · (R1 + R2 · wind velocity),
where 'rain' is the normal rain and 'wind velocity' is that component of the mean wind velocity (measured at a height of 10 m, in open area), which is orthogonal to the building surface. This component is determined from the scalar mean of the wind velocity (see above) and the wind direction distribution (see below).
R1 and R2 are strongly dependent on the specific location on the building facade. If calculations are to be compared to actual measurements, these coefficients should be experimentally determined for that location.
If they cannot be measured, only a very rough estimate is possible. For vertical surfaces, R1 is zero. R2 is about 0.2 s/m for free-standing locations without influence from surrounding buildings etc.; it is markedly less in the center of a facade (e.g. 0.07 s/m); it may even be greater at exposed locations of a building (near edges and corners)[1]. 

  • Driving rain / tipping bucket (real):
  • Driving rain / droplet counter (real):
    the driving rain load on a west-facing wall, measured at a height of 1.50 m with a tipping bucket and a droplet counter rain gauge, respectively. In [Ltr/m²h] or, equivalently, [mm/h].
    WUFI uses the readings of the droplet counter, since it is in general more reliable.
    The values in this column can optionally be read and used for the variable 'rain' in the formula for the rain load (see above: normal rain). Set R1 to 1 and R2 to zero; the resulting rain load will then directly be the measured driving rain.  

  • Wind direction distribution (17 * integer):
    the frequency of wind directions, binned into sectors of 22.5° each.
    The wind direction is averaged over 2 minutes and the counter of the corresponding sector then increased by one. A special counter is registering calms. The sum of all counts during one hour should thus be 30 if no interruption of the measurements has occured.
    The individual counters are assigned to the following directions:
For the calculation of the driving rain load, the contributions of the different directions are determined individually and then summed up. Wind coming 'from behind' does not contribute to the rain load and is ignored. 

A *.WET file supplied by the user must be accompanied by a *.AGD file which contains geographical information on the climate location.


[1]Künzel, H.M.: Bestimmung der Schlagregenbelastung von Fassadenflächen.
IBP-Mitteilung 21 (1994), Nr. 263
[2]VDI 3789 Umweltmeteorologie, Blatt 2: Wechselwirkungen zwischen Atmosphäre und Oberflächen; Berechnung der kurz- und der langwelligen Strahlung.
Oktober 1994