There is an increased demand for hourly heat and moisture simulations for buildings. In simulation programs aimed at energy calculations, it has been enough to include the parameters: temperature, relative humidity, solar radiation, wind speed, wind direction, and cloud cover (Crawley et al. 1999; Wilcox and Marion 2008). These parameters are often a mix of measured and calculated data. Long-wave radiation from the sky and precipitation are sometimes used as parameters but are not as frequently measured. When focusing on horizontal or tilted surfaces, long-wave radiation can not be neglected. In most climate files, the long-wave radiation is not present, so the simulation programs have to make an educated guess based on available parameters. One important secondary parameter is cloudiness, which is measured or calculated from solar radiation compared to maximal theoretic solar radiation. Precipitation is necessary when making moisture calculations for the building envelope. The precipitation is often measured every 6th or 12th hour, so the simulation programs must distribute this over a period hours in between. This paper presents some of the techniques for these calculations and compares the results with real, hourly measured data for four locations in Sweden. General results are that the existing investigated models for long-wave radiation give a root-mean-square accuracy between 24–27 W/m2 and that the models for Sweden using parameters identified above give a root-mean-square error up to 23.2 W/m2. For precipitation, the optimal hourly limit value for precipitation was 88% RHc. INTRODUCTION When calculating temperature and moisture for buildings, some kind of hourly climate file must be used. The two most obvious possibilities are to use measured data from a specific location or constructed data as a base for the climate file. The constructed data can be, for example, hourly values for a typical year or a design (worst case) year for a location. In any case, the climate data must be based on measurements. Meteorological stations usually measure temperature, moisture content (measured as dew point or relative humidity), pressure, wind speed, and wind direction on an hourly basis. Precipitation and cloud cover are often measured more seldom; for example, cloud cover is measured every three and precipitation every twelve hours or daily (SMHI 1988). Global and diffuse solar radiation are measured in fewer locations. If the solar parameters are not measured, they must be modeled from latitude, cloud cover, etc. (Meyers and Dale 1983; Atwater et al. 1978). Note that there are, of course, large differences between countries and locations; the airports will always have climate stations but do not necessarily have automatic hourly stations. Temperature calculations for walls and windows can, in most cases, be made based on this data with reasonable accuracy. The sky temperature for these vertical cases typically is set equal to the air temperature minus around 10°C. For tilted or horizontal surfaces (roofs, glazed spaces, etc.) the calculations must also include the detailed long-wave thermal radiation from the sky to be accurate (Wall 1996). However, the long-wave radiation from the sky is often not measured at all. The reason for this is probably that the price of the measuring instrument has been higher than the expected use of the data. For moisture calculations, hourly values for precipitations are needed. This paper investigates © 2010 ASHRAE. P. Wallentén is a senior lecturer in the Department of Building Physics, Lund University, Lund, Sweden. some existing techniques for constructing hourly values for long-wave radiation and precipitation. LONG-WAVE RADIATION Long-wave radiation from the sky (Lw) is typically measured in W/m2 or W·h/m2·h. The measuring instrument is some kind of pyrgeometer, e.g., a Hukseflux IR02 with an accuracy of 10% for daily sums (Hukseflux manual). To make it clear that this radiation does not originate from the sun, it is sometimes called the atmospheric long-wave radiation. The long-wave radiation is in the order of 200–400 W/m2 (Flerchinger et al. 2009; Crawford and Duchon 1999) and varies on a daily and seasonal basis (see Figure 1). When formulating algorithms based on other meteorological data, it is natural to start with the Stefan-Boltzmann equation for thermal radiation from a surface with temperature T (K):
[1]
S. B. Idso,et al.
A model of thermal radiation from partly cloudy and overcast skies
,
1982
.
[2]
A. Louche,et al.
Determination of Ångström's turbidity coefficient from direct total solar irradiance measurements
,
1987
.
[3]
J. Fox.
Calculated Ångströim's Turbidity Coefficients for Fairbanks, Alaska
,
1994
.
[4]
T. Meyers,et al.
Predicting Daily Insolation with Hourly Cloud Height and Coverage
,
1983
.
[5]
Paul Berdahl,et al.
Emissivity of clear skies
,
1984
.
[6]
M. A. Atwater,et al.
A numerical solar radiation model based on standard meteorological observations
,
1978
.
[7]
Maria Wall.
Climate and Energy Use in Glazed Spaces
,
1996
.
[8]
A. Ångström,et al.
A study of the radiation of the atmosphere
,
2011
.
[9]
A. T. Young.
Air mass and refraction.
,
1994,
Applied optics.
[10]
Marc Aubinet,et al.
Longwave Sky Radiation Parametrizations
,
1994
.