Basic HTML Version
NORDIC LIGHT & COLOUR
83
ing patterns in either quantity and their short-term variability.
Most obvious is the daily/seasonal pattern for both illumi-
nances: short periods of daylight in the winter months, longer
in summer. The hour-by-hour variation in the direct normal
illuminance is clearly visible, though it is also present to a
lesser degree in the diffuse horizontal illuminance (i.e. light
from the sky). The data is for Östersund in northern Sweden –
the location of which is identified in the map above the legend.
Local time is shown, i.e. summertime is local time plus one
hour. The start and end period of summertime are indicated by
vertical dashed lines in each of the figures. Recall the patterns
of indoor illumination for the four components of daylight given
in Figure 5. Each of the 4380 (i.e. the daylight hours) unique
combinations of sky and sun conditions in the weather file
(Figure 6) will result in a unique pattern of internal daylight
illumination.
Both diffuse and direct illuminances will, in reality, vary over
periods much shorter than an hour. Interpolation of the dataset
to a time-step shorter than one hour will provide a smoother
traversal of the sun, which may be necessary when using the
data for simulation of daylight. Interpolation alone however will
not introduce short-term variability into the values for diffuse
horizontal and direct normal illuminance. If required, this vari-
ability would have to be synthesised using stochastic models
(Skartveit and Olseth, 1992).
The illuminance data in the standardised weather files reveal
the true nature of the patterns in daylight illumination from the
sun and the sky. It is also evident from the visualisation of the
data (Figure 6) that any “snapshot” evaluation using just part of
the data would not be representative and could lead to highly
flawed conclusions regarding the daylighting performance of
the building. How these data might be used in their entirety to
better predict actual building performance is described in the
next Section.
Climate-Based Daylight Modelling
Climate-based daylight modelling is the prediction of various
radiant or luminous quantities (e.g. irradiance, illuminance,
radiance and luminance) using sun and sky conditions that
are derived from standardised annual meteorological datas-
ets. Climate-based modelling delivers predictions of absolute
quantities (e.g. illuminance) that are dependent both on the
locale (i.e. geographically-specific climate data is used) and
the fenestration orientation (i.e. accounting for solar position
and non-uniform sky conditions), in addition to the space’s
geometry and material properties. The operation of the space
can also be modelled to varying degrees of precision depending
on the type of device (e.g. luminaire, venetian blinds, etc.) and
its assumed control strategy (e.g. automatic, by occupant, or
some combination).
The term climate-based daylight modelling does not yet have a
formally accepted definition - it was first coined by the author
in the title of a paper given at the 2006 CIBSE National Confer-
ence (Mardaljevic, 2006). However it is generally taken to mean
any evaluation that is founded on the totality (i.e. sun and sky
components) of time-series daylight data appropriate to the
locale over the course of a year. In practice, this means sun
and sky parameters found in, or derived from, the standard
meteorological data files which contain 8760 hourly values for
a full year (Figure 6). Given the self-evident nature of the sea-
sonal pattern in sunlight availability, a function of both the sun
position and the seasonal patterns of cloudiness, an evaluation
period of twelve months is needed to capture all of the natu-
rally occurring variation in conditions that is represented in the
climate dataset. It is also possible to use real-time monitored
weather for a given time period, if calibration to actual moni-
tored conditions within a space is desired. In short, climate-
based daylight modelling is the ability to predict daylight
illumination (such as that shown in Figure 5) for all the hourly
(or sub-hourly) sky and sun conditions in a climate file.
There are a number of possible ways to use climate-based
daylight modelling. The two principal analysis methods are cu-
mulative and time-series. A cumulative analysis is the predic-
tion of some aggregate measure of daylight (e.g. total annual
illuminance) founded on the cumulative luminance effect of
(hourly) sky and the sun conditions derived from the climate
dataset. It is usually determined over a period of a full year,
or on a seasonal or monthly basis, i.e. predicting a cumulative
measure for each season or month in turn. Evaluating cumula-
tive measures for periods shorter than one month is not rec-
ommended since the output will tend to be more revealing of
the unique pattern in the climate dataset than of “typical” con-
ditions for that period. The cumulative method can be used for
predicting the micro-climate and solar access in urban envi-
ronments, the long-term exposure of art works to daylight, and
quick assessments of seasonal daylight availability and/or the
requirement for solar shading at the early design stage. Time-
series analysis involves predicting instantaneous measures
(e.g. illuminance) based on each of the hourly (or sub-hourly)
values in the annual climate dataset. These predictions are
used to evaluate, for example, the overall daylighting potential
of the building, the occurrence of excessive illuminances or
luminances, as inputs to behavioural models for light switching
and/or blinds usage, and the potential of daylight responsive
lighting controls to reduce building energy usage. Thus a day-
light performance metric would need to be based on a time-