Comparison of
Four Methods
to Estimate Urban Heat Storage
Sarah M.
Roberts, T. R.
Oke, C. S. B. Grimmond, and J. A. Voogt
JOURNAL OF APPLIED
METEOROLOGY
AND CLIMATOLOGY, 45 (12) : 1766-1781, DEC 2006
The relative
performance of
four independent methods to estimate the magnitude and diurnal behavior
of net
heat storage fluxes (ΔQS)
in a city center is assessed. This heat flux is a significant but
understudied
component of the urban surface energy balance (SEB). Direct measurement
of this
SEB term at the local scale (horizontal length scale ~102–104
m) is practically unattainable primarily because of the complex array
of
materials and the three-dimensionality of urban systems. Results of an
8-day
summertime observational study at a site in the center of Marseille, France,
are presented. This locale is an ideal environment for such research
because of
the warm, dry climate (hence the SEB is dominated by sensible heat
exchanges)
and the high density of tall buildings with thick walls (hence it has a
large
thermal mass that favors heat storage as a component of the SEB).
Estimates of
ΔQS
derived as
residuals in the SEB, after the remaining terms are measured directly,
(termed
RES) are compared with those calculated from a parameterization scheme
[objective hysteresis model (OHM)], a local-scale numerical model [Town
Energy
Balance model (TEB)], and a bulk heat transfer method [thermal mass
scheme
(TMS)]. Inputs to the methods include observed meteorological data and
morphometric properties of the urban site. All approaches yield a
similar
diurnal course. The OHM and TEB methods tend to slightly overestimate
storage
uptake by day when compared with the RES, whereas TMS slightly
underestimates
it. All methods underestimate heat storage release at night when
compared with
RES and show some sensitivity to wind speed, especially above about 5 m
s−1.
OHM estimates perform satisfactorily in the mean but miss short-term
variability and are poor at night. TEB simulations show the best
agreement with
RES results, particularly at night. TMS values are comparable to those
from the
other methods, but its extensive input requirements render it almost
impractical. Overall, the convergence of results is reassuring but the
lack of
a standard for quantifying heat storage and the spread of results mean
this
term remains a source of imprecision in urban energy balance
measurement and
modeling.